mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-24 03:00:14 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'upstream/3.4' into merge-3.4

Browse Source
This commit is contained in:
Alexander Alekhin 2020-01-12 09:34:34 +00:00
commit fb61f88b9c
29 changed files with 1065 additions and 312 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
CMakeLists.txt
View File

@ -68,6 +68,9 @@ if(POLICY CMP0075)
cmake_policy(SET CMP0075 NEW) # CMake 3.12+: Include file check macros honor `CMAKE_REQUIRED_LIBRARIES`
endif()
if(POLICY CMP0077)
cmake_policy(SET CMP0077 NEW) # CMake 3.13+: option() honors normal variables.
endif()
#
# Configure OpenCV CMake hooks

231
doc/tutorials/core/file_input_output_with_xml_yml/file_input_output_with_xml_yml.markdown
View File

@ -17,7 +17,7 @@ You'll find answers for the following questions:
Source code
-----------
@add_toggle_cpp
You can [download this from here
](https://github.com/opencv/opencv/tree/master/samples/cpp/tutorial_code/core/file_input_output/file_input_output.cpp) or find it in the
`samples/cpp/tutorial_code/core/file_input_output/file_input_output.cpp` of the OpenCV source code
@ -26,13 +26,25 @@ library.
Here's a sample code of how to achieve all the stuff enumerated at the goal list.
@include cpp/tutorial_code/core/file_input_output/file_input_output.cpp
@end_toggle
@add_toggle_python
You can [download this from here
](https://github.com/opencv/opencv/tree/master/samples/python/tutorial_code/core/file_input_output/file_input_output.py) or find it in the
`samples/python/tutorial_code/core/file_input_output/file_input_output.py` of the OpenCV source code
library.
Here's a sample code of how to achieve all the stuff enumerated at the goal list.
@include python/tutorial_code/core/file_input_output/file_input_output.py
@end_toggle
Explanation
-----------
Here we talk only about XML and YAML file inputs. Your output (and its respective input) file may
have only one of these extensions and the structure coming from this. They are two kinds of data
structures you may serialize: *mappings* (like the STL map) and *element sequence* (like the STL
structures you may serialize: *mappings* (like the STL map and the Python dictionary) and *element sequence* (like the STL
vector). The difference between these is that in a map every element has a unique name through what
you may access it. For sequences you need to go through them to query a specific item.
@ -40,12 +52,12 @@ you may access it. For sequences you need to go through them to query a specific
and at the end to close it. The XML/YAML data structure in OpenCV is @ref cv::FileStorage . To
specify that this structure to which file binds on your hard drive you can use either its
constructor or the *open()* function of this:
@code{.cpp}
string filename = "I.xml";
FileStorage fs(filename, FileStorage::WRITE);
//...
fs.open(filename, FileStorage::READ);
@endcode
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp open
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py open
@end_toggle
Either one of this you use the second argument is a constant specifying the type of operations
you'll be able to on them: WRITE, READ or APPEND. The extension specified in the file name also
determinates the output format that will be used. The output may be even compressed if you
@ -53,75 +65,83 @@ you may access it. For sequences you need to go through them to query a specific
The file automatically closes when the @ref cv::FileStorage objects is destroyed. However, you
may explicitly call for this by using the *release* function:
@code{.cpp}
fs.release(); // explicit close
@endcode
-# **Input and Output of text and numbers.** The data structure uses the same \<\< output operator
that the STL library. For outputting any type of data structure we need first to specify its
name. We do this by just simply printing out the name of this. For basic types you may follow
this with the print of the value :
@code{.cpp}
fs << "iterationNr" << 100;
@endcode
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp close
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py close
@end_toggle
-# **Input and Output of text and numbers.** In C++, the data structure uses the \<\< output
operator in the STL library. In Python, @ref cv::FileStorage.write() is used instead. For
outputting any type of data structure we need first to specify its name. We do this by just
simply pushing the name of this to the stream in C++. In Python, the first parameter for the
write function is the name. For basic types you may follow this with the print of the value :
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp writeNum
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py writeNum
@end_toggle
Reading in is a simple addressing (via the [] operator) and casting operation or a read via
the \>\> operator :
@code{.cpp}
int itNr;
fs["iterationNr"] >> itNr;
itNr = (int) fs["iterationNr"];
@endcode
the \>\> operator. In Python, we address with getNode() and use real() :
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp readNum
@end_toggle
@add_toggle_python
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp readNum
@end_toggle
-# **Input/Output of OpenCV Data structures.** Well these behave exactly just as the basic C++
types:
@code{.cpp}
Mat R = Mat_<uchar >::eye (3, 3),
T = Mat_<double>::zeros(3, 1);
fs << "R" << R; // Write cv::Mat
fs << "T" << T;
fs["R"] >> R; // Read cv::Mat
fs["T"] >> T;
@endcode
and Python types:
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp iomati
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp iomatw
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp iomat
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py iomati
@snippet python/tutorial_code/core/file_input_output/file_input_output.py iomatw
@snippet python/tutorial_code/core/file_input_output/file_input_output.py iomat
@end_toggle
-# **Input/Output of vectors (arrays) and associative maps.** As I mentioned beforehand, we can
output maps and sequences (array, vector) too. Again we first print the name of the variable and
then we have to specify if our output is either a sequence or map.
For sequence before the first element print the "[" character and after the last one the "]"
character:
@code{.cpp}
fs << "strings" << "["; // text - string sequence
fs << "image1.jpg" << "Awesomeness" << "baboon.jpg";
fs << "]"; // close sequence
@endcode
character. With Python, the "]" character could be written with the name of the sequence or
the last element of the sequence depending on the number of elements:
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp writeStr
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py writeStr
@end_toggle
For maps the drill is the same however now we use the "{" and "}" delimiter characters:
@code{.cpp}
fs << "Mapping"; // text - mapping
fs << "{" << "One" << 1;
fs << "Two" << 2 << "}";
@endcode
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp writeMap
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py writeMap
@end_toggle
To read from these we use the @ref cv::FileNode and the @ref cv::FileNodeIterator data
structures. The [] operator of the @ref cv::FileStorage class returns a @ref cv::FileNode data
structures. The [] operator of the @ref cv::FileStorage class (or the getNode() function in Python) returns a @ref cv::FileNode data
type. If the node is sequential we can use the @ref cv::FileNodeIterator to iterate through the
items:
@code{.cpp}
FileNode n = fs["strings"]; // Read string sequence - Get node
if (n.type() != FileNode::SEQ)
{
cerr << "strings is not a sequence! FAIL" << endl;
return 1;
}
FileNodeIterator it = n.begin(), it_end = n.end(); // Go through the node
for (; it != it_end; ++it)
cout << (string)*it << endl;
@endcode
For maps you can use the [] operator again to access the given item (or the \>\> operator too):
@code{.cpp}
n = fs["Mapping"]; // Read mappings from a sequence
cout << "Two " << (int)(n["Two"]) << "; ";
cout << "One " << (int)(n["One"]) << endl << endl;
@endcode
items. In Python, the at() function can be used to address elements of the sequence and the
size() function returns the length of the sequence:
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp readStr
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py readStr
@end_toggle
For maps you can use the [] operator (at() function in Python) again to access the given item (or the \>\> operator too):
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp readMap
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py readMap
@end_toggle
-# **Read and write your own data structures.** Suppose you have a data structure such as:
@add_toggle_cpp
@code{.cpp}
class MyData
{
@ -133,53 +153,52 @@ you may access it. For sequences you need to go through them to query a specific
string id;
};
@endcode
It's possible to serialize this through the OpenCV I/O XML/YAML interface (just as in case of
the OpenCV data structures) by adding a read and a write function inside and outside of your
class. For the inside part:
@code{.cpp}
void write(FileStorage& fs) const //Write serialization for this class
{
fs << "{" << "A" << A << "X" << X << "id" << id << "}";
}
void read(const FileNode& node) //Read serialization for this class
{
A = (int)node["A"];
X = (double)node["X"];
id = (string)node["id"];
}
@endcode
Then you need to add the following functions definitions outside the class:
@code{.cpp}
void write(FileStorage& fs, const std::string&, const MyData& x)
{
x.write(fs);
}
void read(const FileNode& node, MyData& x, const MyData& default_value = MyData())
{
if(node.empty())
x = default_value;
else
x.read(node);
}
@end_toggle
@add_toggle_python
@code{.py}
class MyData:
def __init__(self):
self.A = self.X = 0
self.name = ''
@endcode
@end_toggle
In C++, it's possible to serialize this through the OpenCV I/O XML/YAML interface (just as
in case of the OpenCV data structures) by adding a read and a write function inside and outside of your
class. In Python, you can get close to this by implementing a read and write function inside
the class. For the inside part:
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp inside
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py inside
@end_toggle
@add_toggle_cpp
In C++, you need to add the following functions definitions outside the class:
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp outside
@end_toggle
Here you can observe that in the read section we defined what happens if the user tries to read
a non-existing node. In this case we just return the default initialization value, however a
more verbose solution would be to return for instance a minus one value for an object ID.
Once you added these four functions use the \>\> operator for write and the \<\< operator for
read:
@code{.cpp}
MyData m(1);
fs << "MyData" << m; // your own data structures
fs["MyData"] >> m; // Read your own structure_
@endcode
read (or the defined input/output functions for Python):
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp customIOi
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp customIOw
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp customIO
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py customIOi
@snippet python/tutorial_code/core/file_input_output/file_input_output.py customIOw
@snippet python/tutorial_code/core/file_input_output/file_input_output.py customIO
@end_toggle
Or to try out reading a non-existing read:
@code{.cpp}
fs["NonExisting"] >> m; // Do not add a fs << "NonExisting" << m command for this to work
cout << endl << "NonExisting = " << endl << m << endl;
@endcode
@add_toggle_cpp
@snippet cpp/tutorial_code/core/file_input_output/file_input_output.cpp nonexist
@end_toggle
@add_toggle_python
@snippet python/tutorial_code/core/file_input_output/file_input_output.py nonexist
@end_toggle
Result
------

16
doc/tutorials/video/meanshift/meanshift.markdown
View File

@ -57,6 +57,14 @@ low light, low light values are discarded using **cv.inRange()** function.
@include samples/python/tutorial_code/video/meanshift/meanshift.py
@end_toggle
@add_toggle_java
- **Downloadable code**: Click
[here](https://github.com/opencv/opencv/tree/master/samples/java/tutorial_code/video/meanshift/MeanshiftDemo.java)
- **Code at glance:**
@include samples/java/tutorial_code/video/meanshift/MeanshiftDemo.java
@end_toggle
Three frames in a video I used is given below:
![image](images/meanshift_result.jpg)
@ -98,6 +106,14 @@ parameters (used to be passed as search window in next iteration). See the code
@include samples/python/tutorial_code/video/meanshift/camshift.py
@end_toggle
@add_toggle_java
- **Downloadable code**: Click
[here](https://github.com/opencv/opencv/tree/master/samples/java/tutorial_code/video/meanshift/CamshiftDemo.java)
- **Code at glance:**
@include samples/java/tutorial_code/video/meanshift/CamshiftDemo.java
@end_toggle
Three frames of the result is shown below:
![image](images/camshift_result.jpg)

18
doc/tutorials/video/optical_flow/optical_flow.markdown
View File

@ -109,6 +109,15 @@ below:
@include samples/python/tutorial_code/video/optical_flow/optical_flow.py
@end_toggle
@add_toggle_java
- **Downloadable code**: Click
[here](https://github.com/opencv/opencv/tree/master/samples/java/tutorial_code/video/optical_flow/OpticalFlowDemo.java)
- **Code at glance:**
@include samples/java/tutorial_code/video/optical_flow/OpticalFlowDemo.java
@end_toggle
(This code doesn't check how correct are the next keypoints. So even if any feature point disappears
in image, there is a chance that optical flow finds the next point which may look close to it. So
actually for a robust tracking, corner points should be detected in particular intervals. OpenCV
@ -151,6 +160,15 @@ corresponds to Value plane. See the code below:
@end_toggle
@add_toggle_java
- **Downloadable code**: Click
[here](https://github.com/opencv/opencv/tree/master/samples/java/tutorial_code/video/optical_flow/OpticalFlowDenseDemo.java)
- **Code at glance:**
@include samples/java/tutorial_code/video/optical_flow/OpticalFlowDenseDemo.java
@end_toggle
See the result below:
![image](images/opticalfb.jpg)

4
doc/tutorials/video/table_of_content_video.markdown
View File

@ -17,12 +17,12 @@ tracking and foreground extractions.
- @subpage tutorial_meanshift
*Languages:* C++, Python
*Languages:* C++, Java, Python
Learn how to use the Meanshift and Camshift algorithms to track objects in videos.
- @subpage tutorial_optical_flow
*Languages:* C++, Python
*Languages:* C++, Java, Python
We will learn how to use optical flow methods to track sparse features or to create a dense representation.

56
modules/core/include/opencv2/core/hal/intrin.hpp
View File

@ -99,6 +99,7 @@ enum StoreMode
}
// TODO FIXIT: Don't use "God" traits. Split on separate cases.
template<typename _Tp> struct V_TypeTraits
{
};
@ -130,21 +131,51 @@ template<typename _Tp> struct V_TypeTraits
} \
}
#define CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(type, int_type_, uint_type_, abs_type_, w_type_, sum_type_, nlanes128_) \
template<> struct V_TypeTraits<type> \
{ \
typedef type value_type; \
typedef int_type_ int_type; \
typedef abs_type_ abs_type; \
typedef uint_type_ uint_type; \
typedef w_type_ w_type; \
typedef sum_type_ sum_type; \
enum { nlanes128 = nlanes128_ }; \
\
static inline int_type reinterpret_int(type x) \
{ \
union { type l; int_type i; } v; \
v.l = x; \
return v.i; \
} \
\
static inline type reinterpret_from_int(int_type x) \
{ \
union { type l; int_type i; } v; \
v.i = x; \
return v.l; \
} \
}
CV_INTRIN_DEF_TYPE_TRAITS(uchar, schar, uchar, uchar, ushort, unsigned, unsigned, 16);
CV_INTRIN_DEF_TYPE_TRAITS(schar, schar, uchar, uchar, short, int, int, 16);
CV_INTRIN_DEF_TYPE_TRAITS(ushort, short, ushort, ushort, unsigned, uint64, unsigned, 8);
CV_INTRIN_DEF_TYPE_TRAITS(short, short, ushort, ushort, int, int64, int, 8);
CV_INTRIN_DEF_TYPE_TRAITS(unsigned, int, unsigned, unsigned, uint64, void, unsigned, 4);
CV_INTRIN_DEF_TYPE_TRAITS(int, int, unsigned, unsigned, int64, void, int, 4);
CV_INTRIN_DEF_TYPE_TRAITS(float, int, unsigned, float, double, void, float, 4);
CV_INTRIN_DEF_TYPE_TRAITS(uint64, int64, uint64, uint64, void, void, uint64, 2);
CV_INTRIN_DEF_TYPE_TRAITS(int64, int64, uint64, uint64, void, void, int64, 2);
CV_INTRIN_DEF_TYPE_TRAITS(double, int64, uint64, double, void, void, double, 2);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(unsigned, int, unsigned, unsigned, uint64, unsigned, 4);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(int, int, unsigned, unsigned, int64, int, 4);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(float, int, unsigned, float, double, float, 4);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(uint64, int64, uint64, uint64, void, uint64, 2);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(int64, int64, uint64, uint64, void, int64, 2);
CV_INTRIN_DEF_TYPE_TRAITS_NO_Q_TYPE(double, int64, uint64, double, void, double, 2);
#ifndef CV_DOXYGEN
#ifndef CV_CPU_OPTIMIZATION_HAL_NAMESPACE
#ifdef CV_CPU_DISPATCH_MODE
#ifdef CV_FORCE_SIMD128_CPP
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE hal_EMULATOR_CPP
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_BEGIN namespace hal_EMULATOR_CPP {
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_END }
#elif defined(CV_CPU_DISPATCH_MODE)
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE __CV_CAT(hal_, CV_CPU_DISPATCH_MODE)
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_BEGIN namespace __CV_CAT(hal_, CV_CPU_DISPATCH_MODE) {
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_END }
@ -197,7 +228,6 @@ using namespace CV_CPU_OPTIMIZATION_HAL_NAMESPACE;
#else
#define CV_SIMD128_CPP 1
#include "opencv2/core/hal/intrin_cpp.hpp"
#endif
@ -242,6 +272,10 @@ CV_CPU_OPTIMIZATION_HAL_NAMESPACE_BEGIN
#define CV_SIMD128 0
#endif
#ifndef CV_SIMD128_CPP
#define CV_SIMD128_CPP 0
#endif
#ifndef CV_SIMD128_64F
#define CV_SIMD128_64F 0
#endif
@ -346,7 +380,7 @@ template<typename _Tp> struct V_RegTraits
CV_DEF_REG_TRAITS(v, v_int16x8, short, s16, v_uint16x8, v_int32x4, v_int64x2, v_int16x8, void);
CV_DEF_REG_TRAITS(v, v_uint32x4, unsigned, u32, v_uint32x4, v_uint64x2, void, v_int32x4, void);
CV_DEF_REG_TRAITS(v, v_int32x4, int, s32, v_uint32x4, v_int64x2, void, v_int32x4, void);
#if CV_SIMD128_64F
#if CV_SIMD128_64F || CV_SIMD128_CPP
CV_DEF_REG_TRAITS(v, v_float32x4, float, f32, v_float32x4, v_float64x2, void, v_int32x4, v_int32x4);
#else
CV_DEF_REG_TRAITS(v, v_float32x4, float, f32, v_float32x4, void, void, v_int32x4, v_int32x4);
@ -433,7 +467,11 @@ namespace CV__SIMD_NAMESPACE {
} // namespace
using namespace CV__SIMD_NAMESPACE;
#elif (CV_SIMD128 || CV_SIMD128_CPP) && (!defined(CV__SIMD_FORCE_WIDTH) || CV__SIMD_FORCE_WIDTH == 128)
#if defined CV_SIMD128_CPP
#define CV__SIMD_NAMESPACE simd128_cpp
#else
#define CV__SIMD_NAMESPACE simd128
#endif
namespace CV__SIMD_NAMESPACE {
#define CV_SIMD CV_SIMD128
#define CV_SIMD_64F CV_SIMD128_64F

320
modules/core/include/opencv2/core/hal/intrin_cpp.hpp
View File

@ -50,6 +50,14 @@
#include <algorithm>
#include "opencv2/core/saturate.hpp"
//! @cond IGNORED
#define CV_SIMD128_CPP 1
#if defined(CV_FORCE_SIMD128_CPP) || defined(CV_DOXYGEN)
#define CV_SIMD128 1
#define CV_SIMD128_64F 1
#endif
//! @endcond
namespace cv
{
@ -402,50 +410,102 @@ typedef v_reg<uint64, 2> v_uint64x2;
/** @brief Two 64-bit signed integer values */
typedef v_reg<int64, 2> v_int64x2;
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
#define OPENCV_HAL_IMPL_BIN_OP(bin_op) \
template<typename _Tp, int n> inline v_reg<_Tp, n> \
operator bin_op (const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
/** @brief Add values
For all types. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator+(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator+=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Subtract values
For all types. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator-(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator-=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Multiply values
For 16- and 32-bit integer types and floating types. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator*(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator*=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Divide values
For floating types only. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator/(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator/=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Bitwise AND
Only for integer types. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator&(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator&=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Bitwise OR
Only for integer types. */
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator|(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator|=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Bitwise XOR
Only for integer types.*/
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator^(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n>& operator^=(v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b);
/** @brief Bitwise NOT
Only for integer types.*/
template<typename _Tp, int n> CV_INLINE v_reg<_Tp, n> operator~(const v_reg<_Tp, n>& a);
#ifndef CV_DOXYGEN
#define CV__HAL_INTRIN_EXPAND_WITH_INTEGER_TYPES(macro_name, ...) \
__CV_EXPAND(macro_name(uchar, __VA_ARGS__)) \
__CV_EXPAND(macro_name(schar, __VA_ARGS__)) \
__CV_EXPAND(macro_name(ushort, __VA_ARGS__)) \
__CV_EXPAND(macro_name(short, __VA_ARGS__)) \
__CV_EXPAND(macro_name(unsigned, __VA_ARGS__)) \
__CV_EXPAND(macro_name(int, __VA_ARGS__)) \
__CV_EXPAND(macro_name(uint64, __VA_ARGS__)) \
__CV_EXPAND(macro_name(int64, __VA_ARGS__)) \
#define CV__HAL_INTRIN_EXPAND_WITH_FP_TYPES(macro_name, ...) \
__CV_EXPAND(macro_name(float, __VA_ARGS__)) \
__CV_EXPAND(macro_name(double, __VA_ARGS__)) \
#define CV__HAL_INTRIN_EXPAND_WITH_ALL_TYPES(macro_name, ...) \
CV__HAL_INTRIN_EXPAND_WITH_INTEGER_TYPES(macro_name, __VA_ARGS__) \
CV__HAL_INTRIN_EXPAND_WITH_FP_TYPES(macro_name, __VA_ARGS__) \
#define CV__HAL_INTRIN_IMPL_BIN_OP_(_Tp, bin_op) \
template<int n> inline \
v_reg<_Tp, n> operator bin_op (const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
{ \
v_reg<_Tp, n> c; \
for( int i = 0; i < n; i++ ) \
c.s[i] = saturate_cast<_Tp>(a.s[i] bin_op b.s[i]); \
return c; \
} \
template<typename _Tp, int n> inline v_reg<_Tp, n>& \
operator bin_op##= (v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
template<int n> inline \
v_reg<_Tp, n>& operator bin_op##= (v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
{ \
for( int i = 0; i < n; i++ ) \
a.s[i] = saturate_cast<_Tp>(a.s[i] bin_op b.s[i]); \
return a; \
}
/** @brief Add values
#define CV__HAL_INTRIN_IMPL_BIN_OP(bin_op) CV__HAL_INTRIN_EXPAND_WITH_ALL_TYPES(CV__HAL_INTRIN_IMPL_BIN_OP_, bin_op)
For all types. */
OPENCV_HAL_IMPL_BIN_OP(+)
CV__HAL_INTRIN_IMPL_BIN_OP(+)
CV__HAL_INTRIN_IMPL_BIN_OP(-)
CV__HAL_INTRIN_IMPL_BIN_OP(*)
CV__HAL_INTRIN_EXPAND_WITH_FP_TYPES(CV__HAL_INTRIN_IMPL_BIN_OP_, /)
/** @brief Subtract values
For all types. */
OPENCV_HAL_IMPL_BIN_OP(-)
/** @brief Multiply values
For 16- and 32-bit integer types and floating types. */
OPENCV_HAL_IMPL_BIN_OP(*)
/** @brief Divide values
For floating types only. */
OPENCV_HAL_IMPL_BIN_OP(/)
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
#define OPENCV_HAL_IMPL_BIT_OP(bit_op) \
template<typename _Tp, int n> inline v_reg<_Tp, n> operator bit_op \
(const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
#define CV__HAL_INTRIN_IMPL_BIT_OP_(_Tp, bit_op) \
template<int n> CV_INLINE \
v_reg<_Tp, n> operator bit_op (const v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
{ \
v_reg<_Tp, n> c; \
typedef typename V_TypeTraits<_Tp>::int_type itype; \
@ -454,8 +514,8 @@ template<typename _Tp, int n> inline v_reg<_Tp, n> operator bit_op \
V_TypeTraits<_Tp>::reinterpret_int(b.s[i]))); \
return c; \
} \
template<typename _Tp, int n> inline v_reg<_Tp, n>& operator \
bit_op##= (v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
template<int n> CV_INLINE \
v_reg<_Tp, n>& operator bit_op##= (v_reg<_Tp, n>& a, const v_reg<_Tp, n>& b) \
{ \
typedef typename V_TypeTraits<_Tp>::int_type itype; \
for( int i = 0; i < n; i++ ) \
@ -464,33 +524,29 @@ template<typename _Tp, int n> inline v_reg<_Tp, n>& operator \
return a; \
}
/** @brief Bitwise AND
#define CV__HAL_INTRIN_IMPL_BIT_OP(bit_op) \
CV__HAL_INTRIN_EXPAND_WITH_INTEGER_TYPES(CV__HAL_INTRIN_IMPL_BIT_OP_, bit_op) \
CV__HAL_INTRIN_EXPAND_WITH_FP_TYPES(CV__HAL_INTRIN_IMPL_BIT_OP_, bit_op) /* TODO: FIXIT remove this after masks refactoring */
Only for integer types. */
OPENCV_HAL_IMPL_BIT_OP(&)
/** @brief Bitwise OR
CV__HAL_INTRIN_IMPL_BIT_OP(&)
CV__HAL_INTRIN_IMPL_BIT_OP(|)
CV__HAL_INTRIN_IMPL_BIT_OP(^)
Only for integer types. */
OPENCV_HAL_IMPL_BIT_OP(|)
#define CV__HAL_INTRIN_IMPL_BITWISE_NOT_(_Tp, dummy) \
template<int n> CV_INLINE \
v_reg<_Tp, n> operator ~ (const v_reg<_Tp, n>& a) \
{ \
v_reg<_Tp, n> c; \
for( int i = 0; i < n; i++ ) \
c.s[i] = V_TypeTraits<_Tp>::reinterpret_from_int(~V_TypeTraits<_Tp>::reinterpret_int(a.s[i])); \
return c; \
} \
/** @brief Bitwise XOR
CV__HAL_INTRIN_EXPAND_WITH_INTEGER_TYPES(CV__HAL_INTRIN_IMPL_BITWISE_NOT_, ~)
Only for integer types.*/
OPENCV_HAL_IMPL_BIT_OP(^)
#endif // !CV_DOXYGEN
/** @brief Bitwise NOT
Only for integer types.*/
template<typename _Tp, int n> inline v_reg<_Tp, n> operator ~ (const v_reg<_Tp, n>& a)
{
v_reg<_Tp, n> c;
for( int i = 0; i < n; i++ )
{
c.s[i] = V_TypeTraits<_Tp>::reinterpret_from_int(~V_TypeTraits<_Tp>::reinterpret_int(a.s[i]));
}
return c;
}
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
@ -503,6 +559,27 @@ template<typename _Tp, int n> inline v_reg<_Tp2, n> func(const v_reg<_Tp, n>& a)
return c; \
}
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
#define OPENCV_HAL_IMPL_MATH_FUNC_FLOAT(func, cfunc) \
inline v_reg<int, 4> func(const v_reg<float, 4>& a) \
{ \
v_reg<int, 4> c; \
for( int i = 0; i < 4; i++ ) \
c.s[i] = cfunc(a.s[i]); \
return c; \
} \
inline v_reg<int, 4> func(const v_reg<double, 2>& a) \
{ \
v_reg<int, 4> c; \
for( int i = 0; i < 2; i++ ) \
{ \
c.s[i] = cfunc(a.s[i]); \
c.s[i + 2] = 0; \
} \
return c; \
}
/** @brief Square root of elements
Only for floating point types.*/
@ -524,22 +601,22 @@ OPENCV_HAL_IMPL_MATH_FUNC(v_abs, (typename V_TypeTraits<_Tp>::abs_type)std::abs,
/** @brief Round elements
Only for floating point types.*/
OPENCV_HAL_IMPL_MATH_FUNC(v_round, cvRound, int)
OPENCV_HAL_IMPL_MATH_FUNC_FLOAT(v_round, cvRound)
/** @brief Floor elements
Only for floating point types.*/
OPENCV_HAL_IMPL_MATH_FUNC(v_floor, cvFloor, int)
OPENCV_HAL_IMPL_MATH_FUNC_FLOAT(v_floor, cvFloor)
/** @brief Ceil elements
Only for floating point types.*/
OPENCV_HAL_IMPL_MATH_FUNC(v_ceil, cvCeil, int)
OPENCV_HAL_IMPL_MATH_FUNC_FLOAT(v_ceil, cvCeil)
/** @brief Truncate elements
Only for floating point types.*/
OPENCV_HAL_IMPL_MATH_FUNC(v_trunc, int, int)
OPENCV_HAL_IMPL_MATH_FUNC_FLOAT(v_trunc, int)
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
@ -1083,9 +1160,8 @@ OPENCV_HAL_IMPL_SHIFT_OP(<< )
For 16-, 32- and 64-bit integer values. */
OPENCV_HAL_IMPL_SHIFT_OP(>> )
/** @brief Element shift left among vector
For all type */
//! @brief Helper macro
//! @ingroup core_hal_intrin_impl
#define OPENCV_HAL_IMPL_ROTATE_SHIFT_OP(suffix,opA,opB) \
template<int imm, typename _Tp, int n> inline v_reg<_Tp, n> v_rotate_##suffix(const v_reg<_Tp, n>& a) \
{ \
@ -1127,7 +1203,14 @@ template<int imm, typename _Tp, int n> inline v_reg<_Tp, n> v_rotate_##suffix(co
return c; \
}
/** @brief Element shift left among vector
For all type */
OPENCV_HAL_IMPL_ROTATE_SHIFT_OP(left, -, +)
/** @brief Element shift right among vector
For all type */
OPENCV_HAL_IMPL_ROTATE_SHIFT_OP(right, +, -)
/** @brief Sum packed values
@ -1389,6 +1472,7 @@ similar to cv::v_load, but source memory block should be aligned (to 16-byte bou
template<typename _Tp>
inline v_reg<_Tp, V_TypeTraits<_Tp>::nlanes128> v_load_aligned(const _Tp* ptr)
{
CV_Assert(isAligned<sizeof(v_reg<_Tp, V_TypeTraits<_Tp>::nlanes128>)>(ptr));
return v_reg<_Tp, V_TypeTraits<_Tp>::nlanes128>(ptr);
}
@ -1620,6 +1704,12 @@ inline void v_store(_Tp* ptr, const v_reg<_Tp, n>& a)
ptr[i] = a.s[i];
}
template<typename _Tp, int n>
inline void v_store(_Tp* ptr, const v_reg<_Tp, n>& a, hal::StoreMode /*mode*/)
{
v_store(ptr, a);
}
/** @brief Store data to memory (lower half)
Store lower half of register contents to memory.
@ -1659,22 +1749,22 @@ Pointer __should__ be aligned by 16-byte boundary. */
template<typename _Tp, int n>
inline void v_store_aligned(_Tp* ptr, const v_reg<_Tp, n>& a)
{
for( int i = 0; i < n; i++ )
ptr[i] = a.s[i];
CV_Assert(isAligned<sizeof(v_reg<_Tp, n>)>(ptr));
v_store(ptr, a);
}
template<typename _Tp, int n>
inline void v_store_aligned_nocache(_Tp* ptr, const v_reg<_Tp, n>& a)
{
for( int i = 0; i < n; i++ )
ptr[i] = a.s[i];
CV_Assert(isAligned<sizeof(v_reg<_Tp, n>)>(ptr));
v_store(ptr, a);
}
template<typename _Tp, int n>
inline void v_store_aligned(_Tp* ptr, const v_reg<_Tp, n>& a, hal::StoreMode /*mode*/)
{
for( int i = 0; i < n; i++ )
ptr[i] = a.s[i];
CV_Assert(isAligned<sizeof(v_reg<_Tp, n>)>(ptr));
v_store(ptr, a);
}
/** @brief Combine vector from first elements of two vectors
@ -1940,6 +2030,17 @@ template<int n> inline v_reg<float, n> v_cvt_f32(const v_reg<int, n>& a)
return c;
}
template<int n> inline v_reg<float, n*2> v_cvt_f32(const v_reg<double, n>& a)
{
v_reg<float, n*2> c;
for( int i = 0; i < n; i++ )
{
c.s[i] = (float)a.s[i];
c.s[i+n] = 0;
}
return c;
}
template<int n> inline v_reg<float, n*2> v_cvt_f32(const v_reg<double, n>& a, const v_reg<double, n>& b)
{
v_reg<float, n*2> c;
@ -1954,36 +2055,76 @@ template<int n> inline v_reg<float, n*2> v_cvt_f32(const v_reg<double, n>& a, co
/** @brief Convert to double
Supported input type is cv::v_int32x4. */
template<int n> inline v_reg<double, n> v_cvt_f64(const v_reg<int, n*2>& a)
CV_INLINE v_reg<double, 2> v_cvt_f64(const v_reg<int, 4>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i];
return c;
}
/** @brief Convert to double high part of vector
Supported input type is cv::v_int32x4. */
CV_INLINE v_reg<double, 2> v_cvt_f64_high(const v_reg<int, 4>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i + 2];
return c;
}
/** @brief Convert to double
Supported input type is cv::v_float32x4. */
template<int n> inline v_reg<double, n> v_cvt_f64(const v_reg<float, n*2>& a)
CV_INLINE v_reg<double, 2> v_cvt_f64(const v_reg<float, 4>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i];
return c;
}
/** @brief Convert to double high part of vector
Supported input type is cv::v_float32x4. */
CV_INLINE v_reg<double, 2> v_cvt_f64_high(const v_reg<float, 4>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i + 2];
return c;
}
/** @brief Convert to double
Supported input type is cv::v_int64x2. */
template<int n> inline v_reg<double, n> v_cvt_f64(const v_reg<int64, n>& a)
CV_INLINE v_reg<double, 2> v_cvt_f64(const v_reg<int64, 2>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i];
return c;
}
/** @brief Convert to double high part of vector
Supported input type is cv::v_int64x2. */
CV_INLINE v_reg<double, 2> v_cvt_f64_high(const v_reg<int64, 2>& a)
{
enum { n = 2 };
v_reg<double, n> c;
for( int i = 0; i < n; i++ )
c.s[i] = (double)a.s[i];
return c;
}
template<typename _Tp> inline v_reg<_Tp, V_TypeTraits<_Tp>::nlanes128> v_lut(const _Tp* tab, const int* idx)
{
v_reg<_Tp, V_TypeTraits<_Tp>::nlanes128> c;
@ -2038,6 +2179,28 @@ template<int n> inline v_reg<double, n> v_lut(const double* tab, const v_reg<int
return c;
}
inline v_int32x4 v_lut(const int* tab, const v_int32x4& idxvec)
{
return v_lut(tab, idxvec.s);
}
inline v_uint32x4 v_lut(const unsigned* tab, const v_int32x4& idxvec)
{
return v_lut(tab, idxvec.s);
}
inline v_float32x4 v_lut(const float* tab, const v_int32x4& idxvec)
{
return v_lut(tab, idxvec.s);
}
inline v_float64x2 v_lut(const double* tab, const v_int32x4& idxvec)
{
return v_lut(tab, idxvec.s);
}
template<int n> inline void v_lut_deinterleave(const float* tab, const v_reg<int, n>& idx,
v_reg<float, n>& x, v_reg<float, n>& y)
{
@ -2062,7 +2225,7 @@ template<int n> inline void v_lut_deinterleave(const double* tab, const v_reg<in
template<typename _Tp, int n> inline v_reg<_Tp, n> v_interleave_pairs(const v_reg<_Tp, n>& vec)
{
v_reg<float, n> c;
v_reg<_Tp, n> c;
for (int i = 0; i < n/4; i++)
{
c.s[4*i ] = vec.s[4*i ];
@ -2075,7 +2238,7 @@ template<typename _Tp, int n> inline v_reg<_Tp, n> v_interleave_pairs(const v_re
template<typename _Tp, int n> inline v_reg<_Tp, n> v_interleave_quads(const v_reg<_Tp, n>& vec)
{
v_reg<float, n> c;
v_reg<_Tp, n> c;
for (int i = 0; i < n/8; i++)
{
c.s[8*i ] = vec.s[8*i ];
@ -2092,7 +2255,7 @@ template<typename _Tp, int n> inline v_reg<_Tp, n> v_interleave_quads(const v_re
template<typename _Tp, int n> inline v_reg<_Tp, n> v_pack_triplets(const v_reg<_Tp, n>& vec)
{
v_reg<float, n> c;
v_reg<_Tp, n> c;
for (int i = 0; i < n/4; i++)
{
c.s[3*i ] = vec.s[4*i ];
@ -2523,6 +2686,17 @@ inline v_float32x4 v_matmuladd(const v_float32x4& v, const v_float32x4& m0,
v.s[0]*m0.s[3] + v.s[1]*m1.s[3] + v.s[2]*m2.s[3] + m3.s[3]);
}
inline v_float64x2 v_dotprod_expand(const v_int32x4& a, const v_int32x4& b)
{ return v_fma(v_cvt_f64(a), v_cvt_f64(b), v_cvt_f64_high(a) * v_cvt_f64_high(b)); }
inline v_float64x2 v_dotprod_expand(const v_int32x4& a, const v_int32x4& b, const v_float64x2& c)
{ return v_fma(v_cvt_f64(a), v_cvt_f64(b), v_fma(v_cvt_f64_high(a), v_cvt_f64_high(b), c)); }
inline v_float64x2 v_dotprod_expand_fast(const v_int32x4& a, const v_int32x4& b)
{ return v_dotprod_expand(a, b); }
inline v_float64x2 v_dotprod_expand_fast(const v_int32x4& a, const v_int32x4& b, const v_float64x2& c)
{ return v_dotprod_expand(a, b, c); }
////// FP16 support ///////
inline v_reg<float, V_TypeTraits<float>::nlanes128>
@ -2537,7 +2711,7 @@ v_load_expand(const float16_t* ptr)
}
inline void
v_pack_store(float16_t* ptr, v_reg<float, V_TypeTraits<float>::nlanes128>& v)
v_pack_store(float16_t* ptr, const v_reg<float, V_TypeTraits<float>::nlanes128>& v)
{
for( int i = 0; i < v.nlanes; i++ )
{

3
modules/core/src/arithm.cpp
View File

@ -1492,7 +1492,8 @@ struct InRange_SIMD<float>
v_float32 low2 = vx_load(src2 + x + v_float32::nlanes);
v_float32 high2 = vx_load(src3 + x + v_float32::nlanes);
v_pack_store(dst + x, v_pack(v_reinterpret_as_u32((values1 >= low1) & (high1 >= values1)), v_reinterpret_as_u32((values2 >= low2) & (high2 >= values2))));
v_pack_store(dst + x, v_pack(v_reinterpret_as_u32(values1 >= low1) & v_reinterpret_as_u32(high1 >= values1),
v_reinterpret_as_u32(values2 >= low2) & v_reinterpret_as_u32(high2 >= values2)));
}
vx_cleanup();
return x;

4
modules/core/src/arithm.simd.hpp
View File

@ -1576,7 +1576,7 @@ struct op_div_scale
}
static inline Tvec pre(const Tvec& denom, const Tvec& res)
{
const Tvec v_zero = Tvec();
const Tvec v_zero = vx_setall<typename Tvec::lane_type>(0);
return v_select(denom == v_zero, v_zero, res);
}
static inline T1 r(T1 a, T1 denom, const T2* scalar)
@ -1826,7 +1826,7 @@ struct op_recip
}
static inline Tvec pre(const Tvec& denom, const Tvec& res)
{
const Tvec v_zero = Tvec();
const Tvec v_zero = vx_setall<typename Tvec::lane_type>(0);
return v_select(denom == v_zero, v_zero, res);
}
static inline T1 r(T1 denom, const T2* scalar)

7
modules/core/src/lapack.cpp
View File

@ -916,8 +916,9 @@ double cv::invert( InputArray _src, OutputArray _dst, int method )
result = true;
d = 1./d;
#if CV_SIMD128
static const float CV_DECL_ALIGNED(16) inv[4] = { 0.f,-0.f,-0.f,0.f };
v_float32x4 s0 = (v_load_halves((const float*)srcdata, (const float*)(srcdata + srcstep)) * v_setall_f32((float)d)) ^ v_load((const float *)inv);//0123//3120
const float d_32f = (float)d;
const v_float32x4 d_vec(d_32f, -d_32f, -d_32f, d_32f);
v_float32x4 s0 = v_load_halves((const float*)srcdata, (const float*)(srcdata + srcstep)) * d_vec;//0123//3120
s0 = v_extract<3>(s0, v_combine_low(v_rotate_right<1>(s0), s0));
v_store_low((float*)dstdata, s0);
v_store_high((float*)(dstdata + dststep), s0);
@ -946,7 +947,7 @@ double cv::invert( InputArray _src, OutputArray _dst, int method )
v_float64x2 s0 = v_load((const double*)srcdata) * det;
v_float64x2 s1 = v_load((const double*)(srcdata+srcstep)) * det;
v_float64x2 sm = v_extract<1>(s1, s0);//30
v_float64x2 ss = v_extract<1>(s0, s1) ^ v_setall_f64(-0.);//12
v_float64x2 ss = v_setall<double>(0) - v_extract<1>(s0, s1);//12
v_store((double*)dstdata, v_combine_low(sm, ss));//31
v_store((double*)(dstdata + dststep), v_combine_high(ss, sm));//20
#else

2
modules/core/src/mathfuncs_core.simd.hpp
View File

@ -725,7 +725,7 @@ void log32f( const float *_x, float *y, int n )
yf0 = v_fma(v_cvt_f32(yi0), vln2, yf0);
v_float32 delta = v_reinterpret_as_f32(h0 == vx_setall_s32(510)) & vshift;
v_float32 delta = v_select(v_reinterpret_as_f32(h0 == vx_setall_s32(510)), vshift, vx_setall<float>(0));
xf0 = v_fma((v_reinterpret_as_f32(xi0) - v1), xf0, delta);
v_float32 zf0 = v_fma(xf0, vA0, vA1);

1
modules/core/src/ovx.cpp
View File

@ -8,6 +8,7 @@
// OpenVX related functions
#include "precomp.hpp"
#include "opencv2/core/utils/tls.hpp"
#include "opencv2/core/ovx.hpp"
#include "opencv2/core/openvx/ovx_defs.hpp"

12
modules/core/test/test_intrin_emulator.cpp
View File

@ -3,22 +3,14 @@
// of this distribution and at http://opencv.org/license.html.
#include "test_precomp.hpp"
// see "opencv2/core/hal/intrin.hpp"
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE hal_EMULATOR_CPP
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_BEGIN namespace hal_EMULATOR_CPP {
#define CV_CPU_OPTIMIZATION_HAL_NAMESPACE_END }
// see "opencv2/core/private/cv_cpu_include_simd_declarations.hpp"
//#define CV_CPU_OPTIMIZATION_DECLARATIONS_ONLY
#define CV_FORCE_SIMD128_CPP
#undef CV_FORCE_SIMD128_CPP
#define CV_FORCE_SIMD128_CPP 1
#undef CV_CPU_OPTIMIZATION_NAMESPACE_BEGIN
#undef CV_CPU_OPTIMIZATION_NAMESPACE_END
#define CV_CPU_OPTIMIZATION_NAMESPACE_BEGIN namespace opt_EMULATOR_CPP {
#define CV_CPU_OPTIMIZATION_NAMESPACE_END }
#include "test_intrin128.simd.hpp"
#undef CV_CPU_OPTIMIZATION_NAMESPACE_BEGIN
#undef CV_CPU_OPTIMIZATION_NAMESPACE_END
#undef CV_CPU_DISPATCH_MODE
#undef CV_FORCE_SIMD128_CPP
// tests implementation is in test_intrin_utils.hpp

11
modules/core/test/test_intrin_utils.hpp
View File

@ -222,7 +222,10 @@ template <typename R> std::ostream & operator<<(std::ostream & out, const Data<R
return out;
}
template<typename T> static inline void EXPECT_COMPARE_EQ_(const T a, const T b);
template<typename T> static inline void EXPECT_COMPARE_EQ_(const T a, const T b)
{
EXPECT_EQ(a, b);
}
template<> inline void EXPECT_COMPARE_EQ_<float>(const float a, const float b)
{
EXPECT_FLOAT_EQ( a, b );
@ -742,11 +745,11 @@ template<typename R> struct TheTest
for (int i = 0; i < n; ++i)
{
SCOPED_TRACE(cv::format("i=%d", i));
EXPECT_EQ((double)dataA[i*2] * (double)dataA[i*2] +
EXPECT_COMPARE_EQ((double)dataA[i*2] * (double)dataA[i*2] +
(double)dataA[i*2 + 1] * (double)dataA[i*2 + 1], resA[i]);
EXPECT_EQ((double)dataB[i*2] * (double)dataB[i*2] +
EXPECT_COMPARE_EQ((double)dataB[i*2] * (double)dataB[i*2] +
(double)dataB[i*2 + 1] * (double)dataB[i*2 + 1], resB[i]);
EXPECT_EQ((double)dataA[i*2] * (double)dataB[i*2] +
EXPECT_COMPARE_EQ((double)dataA[i*2] * (double)dataB[i*2] +
(double)dataA[i*2 + 1] * (double)dataB[i*2 + 1] + dataC[i], resC[i]);
}
#endif

12
modules/dnn/src/tensorflow/tf_graph_simplifier.cpp
View File

@ -950,6 +950,7 @@ void sortByExecutionOrder(tensorflow::GraphDef& net)
for (int i = 0; i < net.node_size(); ++i)
{
const tensorflow::NodeDef& node = net.node(i);
int numInputsInGraph = 0;
for (int j = 0; j < node.input_size(); ++j)
{
std::string inpName = node.input(j);
@ -957,22 +958,25 @@ void sortByExecutionOrder(tensorflow::GraphDef& net)
inpName = inpName.substr(inpName.find('^') + 1);
nodesMapIt = nodesMap.find(inpName);
CV_Assert(nodesMapIt != nodesMap.end());
if (nodesMapIt != nodesMap.end())
{
edges[nodesMapIt->second].push_back(i);
numInputsInGraph += 1;
}
if (node.input_size() == 0)
}
if (numInputsInGraph == 0)
nodesToAdd.push_back(i);
else
{
if (node.op() == "Merge" || node.op() == "RefMerge")
{
int numControlEdges = 0;
for (int j = 0; j < node.input_size(); ++j)
for (int j = 0; j < numInputsInGraph; ++j)
numControlEdges += node.input(j)[0] == '^';
numRefsToAdd[i] = numControlEdges + 1;
}
else
numRefsToAdd[i] = node.input_size();
numRefsToAdd[i] = numInputsInGraph;
}
}

4
modules/dnn/src/tensorflow/tf_importer.cpp
View File

@ -715,6 +715,10 @@ void TFImporter::populateNet(Net dstNet)
simplifySubgraphs(netBin);
sortByExecutionOrder(netBin);
}
else
{
sortByExecutionOrder(netTxt);
}
std::set<String> layers_to_ignore;

3
modules/features2d/src/fast_score.cpp
View File

@ -303,7 +303,8 @@ int cornerScore<8>(const uchar* ptr, const int pixel[], int threshold)
for (k = 0; k < N; k++)
d[k] = (short)(v - ptr[pixel[k]]);
#if CV_SIMD128
#if CV_SIMD128 \
&& (!defined(CV_SIMD128_CPP) || (!defined(__GNUC__) || __GNUC__ != 5)) // "movdqa" bug on "v_load(d + 1)" line (Ubuntu 16.04 + GCC 5.4)
if (true)
{
v_int16x8 v0 = v_load(d + 1);

42
modules/imgcodecs/src/utils.cpp
View File

@ -56,65 +56,65 @@ int validateToInt(size_t sz)
#define cG (int)(0.587*(1 << SCALE) + 0.5)
#define cB ((1 << SCALE) - cR - cG)
void icvCvt_BGR2Gray_8u_C3C1R( const uchar* rgb, int rgb_step,
void icvCvt_BGR2Gray_8u_C3C1R( const uchar* bgr, int bgr_step,
uchar* gray, int gray_step,
Size size, int _swap_rb )
{
int i;
for( ; size.height--; gray += gray_step )
{
short cRGB0 = cR;
short cRGB2 = cB;
if (_swap_rb) std::swap(cRGB0, cRGB2);
for( i = 0; i < size.width; i++, rgb += 3 )
short cBGR0 = cB;
short cBGR2 = cR;
if (_swap_rb) std::swap(cBGR0, cBGR2);
for( i = 0; i < size.width; i++, bgr += 3 )
{
int t = descale( rgb[0]*cRGB0 + rgb[1]*cG + rgb[2]*cRGB2, SCALE );
int t = descale( bgr[0]*cBGR0 + bgr[1]*cG + bgr[2]*cBGR2, SCALE );
gray[i] = (uchar)t;
}
rgb += rgb_step - size.width*3;
bgr += bgr_step - size.width*3;
}
}
void icvCvt_BGRA2Gray_16u_CnC1R( const ushort* rgb, int rgb_step,
void icvCvt_BGRA2Gray_16u_CnC1R( const ushort* bgr, int bgr_step,
ushort* gray, int gray_step,
Size size, int ncn, int _swap_rb )
{
int i;
for( ; size.height--; gray += gray_step )
{
short cRGB0 = cR;
short cRGB2 = cB;
if (_swap_rb) std::swap(cRGB0, cRGB2);
for( i = 0; i < size.width; i++, rgb += ncn )
short cBGR0 = cB;
short cBGR2 = cR;
if (_swap_rb) std::swap(cBGR0, cBGR2);
for( i = 0; i < size.width; i++, bgr += ncn )
{
int t = descale( rgb[0]*cRGB0 + rgb[1]*cG + rgb[2]*cRGB2, SCALE );
int t = descale( bgr[0]*cBGR0 + bgr[1]*cG + bgr[2]*cBGR2, SCALE );
gray[i] = (ushort)t;
}
rgb += rgb_step - size.width*ncn;
bgr += bgr_step - size.width*ncn;
}
}
void icvCvt_BGRA2Gray_8u_C4C1R( const uchar* rgba, int rgba_step,
void icvCvt_BGRA2Gray_8u_C4C1R( const uchar* bgra, int rgba_step,
uchar* gray, int gray_step,
Size size, int _swap_rb )
{
int i;
for( ; size.height--; gray += gray_step )
{
short cRGB0 = cR;
short cRGB2 = cB;
if (_swap_rb) std::swap(cRGB0, cRGB2);
for( i = 0; i < size.width; i++, rgba += 4 )
short cBGR0 = cB;
short cBGR2 = cR;
if (_swap_rb) std::swap(cBGR0, cBGR2);
for( i = 0; i < size.width; i++, bgra += 4 )
{
int t = descale( rgba[0]*cRGB0 + rgba[1]*cG + rgba[2]*cRGB2, SCALE );
int t = descale( bgra[0]*cBGR0 + bgra[1]*cG + bgra[2]*cBGR2, SCALE );
gray[i] = (uchar)t;
}
rgba += rgba_step - size.width*4;
bgra += rgba_step - size.width*4;
}
}

1
modules/imgproc/src/corner.avx.cpp
View File

@ -42,6 +42,7 @@
//M*/
#include "precomp.hpp"
#undef CV_FORCE_SIMD128_CPP // expected AVX implementation only
#include "opencv2/core/hal/intrin.hpp"
#include "corner.hpp"

100
modules/imgproc/src/resize.cpp
View File

@ -1109,23 +1109,29 @@ resizeNN( const Mat& src, Mat& dst, double fx, double fy )
struct VResizeNoVec
{
int operator()(const uchar**, uchar*, const uchar*, int ) const { return 0; }
template<typename WT, typename T, typename BT>
int operator()(const WT**, T*, const BT*, int ) const
{
return 0;
}
};
struct HResizeNoVec
{
int operator()(const uchar**, uchar**, int, const int*,
const uchar*, int, int, int, int, int) const { return 0; }
template<typename T, typename WT, typename AT> inline
int operator()(const T**, WT**, int, const int*,
const AT*, int, int, int, int, int) const
{
return 0;
}
};
#if CV_SIMD
struct VResizeLinearVec_32s8u
{
int operator()(const uchar** _src, uchar* dst, const uchar* _beta, int width ) const
int operator()(const int** src, uchar* dst, const short* beta, int width) const
{
const int** src = (const int**)_src;
const short* beta = (const short*)_beta;
const int *S0 = src[0], *S1 = src[1];
int x = 0;
v_int16 b0 = vx_setall_s16(beta[0]), b1 = vx_setall_s16(beta[1]);
@ -1153,12 +1159,9 @@ struct VResizeLinearVec_32s8u
struct VResizeLinearVec_32f16u
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, ushort* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1];
ushort* dst = (ushort*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]);
@ -1183,12 +1186,9 @@ struct VResizeLinearVec_32f16u
struct VResizeLinearVec_32f16s
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, short* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1];
short* dst = (short*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]);
@ -1213,12 +1213,9 @@ struct VResizeLinearVec_32f16s
struct VResizeLinearVec_32f
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, float* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1];
float* dst = (float*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]);
@ -1237,10 +1234,8 @@ struct VResizeLinearVec_32f
struct VResizeCubicVec_32s8u
{
int operator()(const uchar** _src, uchar* dst, const uchar* _beta, int width ) const
int operator()(const int** src, uchar* dst, const short* beta, int width) const
{
const int** src = (const int**)_src;
const short* beta = (const short*)_beta;
const int *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3];
int x = 0;
float scale = 1.f/(INTER_RESIZE_COEF_SCALE*INTER_RESIZE_COEF_SCALE);
@ -1274,12 +1269,9 @@ struct VResizeCubicVec_32s8u
struct VResizeCubicVec_32f16u
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, ushort* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3];
ushort* dst = (ushort*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
b2 = vx_setall_f32(beta[2]), b3 = vx_setall_f32(beta[3]);
@ -1300,12 +1292,9 @@ struct VResizeCubicVec_32f16u
struct VResizeCubicVec_32f16s
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, short* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3];
short* dst = (short*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
b2 = vx_setall_f32(beta[2]), b3 = vx_setall_f32(beta[3]);
@ -1326,12 +1315,9 @@ struct VResizeCubicVec_32f16s
struct VResizeCubicVec_32f
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, float* dst, const float* beta, int width) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3];
float* dst = (float*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
b2 = vx_setall_f32(beta[2]), b3 = vx_setall_f32(beta[3]);
@ -1351,10 +1337,12 @@ struct VResizeCubicVec_32f
struct VResizeLanczos4Vec_32f16u
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, ushort* dst, const float* beta, int width) const
{
if (CV_CPU_HAS_SUPPORT_SSE4_1) return opt_SSE4_1::VResizeLanczos4Vec_32f16u_SSE41(_src, _dst, _beta, width);
else return 0;
if (CV_CPU_HAS_SUPPORT_SSE4_1)
return opt_SSE4_1::VResizeLanczos4Vec_32f16u_SSE41(src, dst, beta, width);
else
return 0;
}
};
@ -1362,13 +1350,10 @@ struct VResizeLanczos4Vec_32f16u
struct VResizeLanczos4Vec_32f16u
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, ushort* dst, const float* beta, int width ) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3],
*S4 = src[4], *S5 = src[5], *S6 = src[6], *S7 = src[7];
ushort * dst = (ushort*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
b2 = vx_setall_f32(beta[2]), b3 = vx_setall_f32(beta[3]),
@ -1401,13 +1386,10 @@ struct VResizeLanczos4Vec_32f16u
struct VResizeLanczos4Vec_32f16s
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, short* dst, const float* beta, int width ) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3],
*S4 = src[4], *S5 = src[5], *S6 = src[6], *S7 = src[7];
short * dst = (short*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
b2 = vx_setall_f32(beta[2]), b3 = vx_setall_f32(beta[3]),
@ -1438,13 +1420,10 @@ struct VResizeLanczos4Vec_32f16s
struct VResizeLanczos4Vec_32f
{
int operator()(const uchar** _src, uchar* _dst, const uchar* _beta, int width ) const
int operator()(const float** src, float* dst, const float* beta, int width ) const
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3],
*S4 = src[4], *S5 = src[5], *S6 = src[6], *S7 = src[7];
float* dst = (float*)_dst;
int x = 0;
v_float32 b0 = vx_setall_f32(beta[0]), b1 = vx_setall_f32(beta[1]),
@ -1489,12 +1468,9 @@ typedef VResizeNoVec VResizeLanczos4Vec_32f;
template<typename ST, typename DT, typename AT, typename DVT>
struct HResizeLinearVec_X4
{
int operator()(const uchar** _src, uchar** _dst, int count, const int* xofs,
const uchar* _alpha, int, int, int cn, int, int xmax) const
int operator()(const ST** src, DT** dst, int count, const int* xofs,
const AT* alpha, int, int, int cn, int, int xmax) const
{
const ST **src = (const ST**)_src;
const AT *alpha = (const AT*)_alpha;
DT **dst = (DT**)_dst;
const int nlanes = 4;
const int len0 = xmax & -nlanes;
int dx = 0, k = 0;
@ -1549,11 +1525,9 @@ struct HResizeLinearVec_X4
struct HResizeLinearVecU8_X4
{
int operator()(const uchar** src, uchar** _dst, int count, const int* xofs,
const uchar* _alpha, int smax, int, int cn, int, int xmax) const
int operator()(const uchar** src, int** dst, int count, const int* xofs,
const short* alpha/*[xmax]*/, int smax, int /*dmax*/, int cn, int /*xmin*/, int xmax) const
{
const short *alpha = (const short*)_alpha;
int **dst = (int**)_dst;
int dx = 0, k = 0;
if(cn == 1)
@ -1827,8 +1801,8 @@ struct HResizeLinear
int dx, k;
VecOp vecOp;
int dx0 = vecOp((const uchar**)src, (uchar**)dst, count,
xofs, (const uchar*)alpha, swidth, dwidth, cn, xmin, xmax );
int dx0 = vecOp(src, dst, count,
xofs, alpha, swidth, dwidth, cn, xmin, xmax );
for( k = 0; k <= count - 2; k+=2 )
{
@ -1881,7 +1855,7 @@ struct VResizeLinear
CastOp castOp;
VecOp vecOp;
int x = vecOp((const uchar**)src, (uchar*)dst, (const uchar*)beta, width);
int x = vecOp(src, dst, beta, width);
#if CV_ENABLE_UNROLLED
for( ; x <= width - 4; x += 4 )
{
@ -1912,7 +1886,7 @@ struct VResizeLinear<uchar, int, short, FixedPtCast<int, uchar, INTER_RESIZE_COE
const buf_type *S0 = src[0], *S1 = src[1];
VResizeLinearVec_32s8u vecOp;
int x = vecOp((const uchar**)src, (uchar*)dst, (const uchar*)beta, width);
int x = vecOp(src, dst, beta, width);
#if CV_ENABLE_UNROLLED
for( ; x <= width - 4; x += 4 )
{
@ -1994,7 +1968,7 @@ struct VResizeCubic
CastOp castOp;
VecOp vecOp;
int x = vecOp((const uchar**)src, (uchar*)dst, (const uchar*)beta, width);
int x = vecOp(src, dst, beta, width);
for( ; x < width; x++ )
dst[x] = castOp(S0[x]*b0 + S1[x]*b1 + S2[x]*b2 + S3[x]*b3);
}
@ -2066,7 +2040,7 @@ struct VResizeLanczos4
{
CastOp castOp;
VecOp vecOp;
int x = vecOp((const uchar**)src, (uchar*)dst, (const uchar*)beta, width);
int x = vecOp(src, dst, beta, width);
#if CV_ENABLE_UNROLLED
for( ; x <= width - 4; x += 4 )
{

2
modules/imgproc/src/resize.hpp
View File

@ -67,7 +67,7 @@ namespace opt_SSE4_1
void resizeNN2_SSE4_1(const Range&, const Mat&, Mat&, int*, int, double);
void resizeNN4_SSE4_1(const Range&, const Mat&, Mat&, int*, int, double);
int VResizeLanczos4Vec_32f16u_SSE41(const uchar** _src, uchar* _dst, const uchar* _beta, int width);
int VResizeLanczos4Vec_32f16u_SSE41(const float** src, ushort* dst, const float* beta, int width);
#endif
}
}

5
modules/imgproc/src/resize.sse4_1.cpp
View File

@ -186,13 +186,10 @@ void resizeNN4_SSE4_1(const Range& range, const Mat& src, Mat &dst, int *x_ofs,
parallel_for_(range, invoker, dst.total() / (double)(1 << 16));
}
int VResizeLanczos4Vec_32f16u_SSE41(const uchar** _src, uchar* _dst, const uchar* _beta, int width)
int VResizeLanczos4Vec_32f16u_SSE41(const float** src, ushort* dst, const float* beta, int width)
{
const float** src = (const float**)_src;
const float* beta = (const float*)_beta;
const float *S0 = src[0], *S1 = src[1], *S2 = src[2], *S3 = src[3],
*S4 = src[4], *S5 = src[5], *S6 = src[6], *S7 = src[7];
short * dst = (short*)_dst;
int x = 0;
__m128 v_b0 = _mm_set1_ps(beta[0]), v_b1 = _mm_set1_ps(beta[1]),
v_b2 = _mm_set1_ps(beta[2]), v_b3 = _mm_set1_ps(beta[3]),

4
modules/objdetect/misc/python/test/test_qrcode_detect.py
View File

@ -4,7 +4,7 @@
QR code detect and decode pipeline.
===============================================================================
'''
import os
import numpy as np
import cv2 as cv
@ -12,7 +12,7 @@ from tests_common import NewOpenCVTests
class qrcode_detector_test(NewOpenCVTests):
def test_detect_and_decode(self):
img = cv.imread(self.extraTestDataPath + '/cv/qrcode/link_ocv.jpg')
img = cv.imread(os.path.join(self.extraTestDataPath, 'cv/qrcode/link_ocv.jpg'))
detector = cv.QRCodeDetector()
retval, points, straight_qrcode = detector.detectAndDecode(img)
self.assertEqual(retval, "https://opencv.org/");

37
samples/cpp/tutorial_code/core/file_input_output/file_input_output.cpp
View File

@ -25,6 +25,7 @@ public:
{}
explicit MyData(int) : A(97), X(CV_PI), id("mydata1234") // explicit to avoid implicit conversion
{}
//! [inside]
void write(FileStorage& fs) const //Write serialization for this class
{
fs << "{" << "A" << A << "X" << X << "id" << id << "}";
@ -35,6 +36,7 @@ public:
X = (double)node["X"];
id = (string)node["id"];
}
//! [inside]
public: // Data Members
int A;
double X;
@ -42,6 +44,7 @@ public: // Data Members
};
//These write and read functions must be defined for the serialization in FileStorage to work
//! [outside]
static void write(FileStorage& fs, const std::string&, const MyData& x)
{
x.write(fs);
@ -52,6 +55,7 @@ static void read(const FileNode& node, MyData& x, const MyData& default_value =
else
x.read(node);
}
//! [outside]
// This function will print our custom class to the console
static ostream& operator<<(ostream& out, const MyData& m)
@ -72,27 +76,48 @@ int main(int ac, char** av)
string filename = av[1];
{ //write
//! [iomati]
Mat R = Mat_<uchar>::eye(3, 3),
T = Mat_<double>::zeros(3, 1);
//! [iomati]
//! [customIOi]
MyData m(1);
//! [customIOi]
//! [open]
FileStorage fs(filename, FileStorage::WRITE);
// or:
// FileStorage fs;
// fs.open(filename, FileStorage::WRITE);
//! [open]
//! [writeNum]
fs << "iterationNr" << 100;
//! [writeNum]
//! [writeStr]
fs << "strings" << "["; // text - string sequence
fs << "image1.jpg" << "Awesomeness" << "../data/baboon.jpg";
fs << "]"; // close sequence
//! [writeStr]
//! [writeMap]
fs << "Mapping"; // text - mapping
fs << "{" << "One" << 1;
fs << "Two" << 2 << "}";
//! [writeMap]
//! [iomatw]
fs << "R" << R; // cv::Mat
fs << "T" << T;
//! [iomatw]
//! [customIOw]
fs << "MyData" << m; // your own data structures
//! [customIOw]
//! [close]
fs.release(); // explicit close
//! [close]
cout << "Write Done." << endl;
}
@ -101,9 +126,11 @@ int main(int ac, char** av)
FileStorage fs;
fs.open(filename, FileStorage::READ);
//! [readNum]
int itNr;
//fs["iterationNr"] >> itNr;
itNr = (int) fs["iterationNr"];
//! [readNum]
cout << itNr;
if (!fs.isOpened())
{
@ -112,6 +139,7 @@ int main(int ac, char** av)
return 1;
}
//! [readStr]
FileNode n = fs["strings"]; // Read string sequence - Get node
if (n.type() != FileNode::SEQ)
{
@ -122,19 +150,26 @@ int main(int ac, char** av)
FileNodeIterator it = n.begin(), it_end = n.end(); // Go through the node
for (; it != it_end; ++it)
cout << (string)*it << endl;
//! [readStr]
//! [readMap]
n = fs["Mapping"]; // Read mappings from a sequence
cout << "Two " << (int)(n["Two"]) << "; ";
cout << "One " << (int)(n["One"]) << endl << endl;
//! [readMap]
MyData m;
Mat R, T;
//! [iomat]
fs["R"] >> R; // Read cv::Mat
fs["T"] >> T;
//! [iomat]
//! [customIO]
fs["MyData"] >> m; // Read your own structure_
//! [customIO]
cout << endl
<< "R = " << R << endl;
@ -142,9 +177,11 @@ int main(int ac, char** av)
cout << "MyData = " << endl << m << endl << endl;
//Show default behavior for non existing nodes
//! [nonexist]
cout << "Attempt to read NonExisting (should initialize the data structure with its default).";
fs["NonExisting"] >> m;
cout << endl << "NonExisting = " << endl << m << endl;
//! [nonexist]
}
cout << endl

75
samples/java/tutorial_code/video/meanshift/CamshiftDemo.java Normal file
View File

@ -0,0 +1,75 @@
import java.util.Arrays;
import org.opencv.core.*;
import org.opencv.highgui.HighGui;
import org.opencv.imgproc.Imgproc;
import org.opencv.video.Video;
import org.opencv.videoio.VideoCapture;
class Camshift {
public void run(String[] args) {
String filename = args[0];
VideoCapture capture = new VideoCapture(filename);
if (!capture.isOpened()) {
System.out.println("Unable to open file!");
System.exit(-1);
}
Mat frame = new Mat(), hsv_roi = new Mat(), mask = new Mat(), roi;
// take the first frame of the video
capture.read(frame);
//setup initial location of window
Rect track_window = new Rect(300, 200, 100, 50);
// set up the ROI for tracking
roi = new Mat(frame, track_window);
Imgproc.cvtColor(roi, hsv_roi, Imgproc.COLOR_BGR2HSV);
Core.inRange(hsv_roi, new Scalar(0, 60, 32), new Scalar(180, 255, 255), mask);
MatOfFloat range = new MatOfFloat(0, 256);
Mat roi_hist = new Mat();
MatOfInt histSize = new MatOfInt(180);
MatOfInt channels = new MatOfInt(0);
Imgproc.calcHist(Arrays.asList(hsv_roi), channels, mask, roi_hist, histSize, range);
Core.normalize(roi_hist, roi_hist, 0, 255, Core.NORM_MINMAX);
// Setup the termination criteria, either 10 iteration or move by atleast 1 pt
TermCriteria term_crit = new TermCriteria(TermCriteria.EPS | TermCriteria.COUNT, 10, 1);
while (true) {
Mat hsv = new Mat() , dst = new Mat();
capture.read(frame);
if (frame.empty()) {
break;
}
Imgproc.cvtColor(frame, hsv, Imgproc.COLOR_BGR2HSV);
Imgproc.calcBackProject(Arrays.asList(hsv), channels, roi_hist, dst, range, 1);
// apply camshift to get the new location
RotatedRect rot_rect = Video.CamShift(dst, track_window, term_crit);
// Draw it on image
Point[] points = new Point[4];
rot_rect.points(points);
for (int i = 0; i < 4 ;i++) {
Imgproc.line(frame, points[i], points[(i+1)%4], new Scalar(255, 0, 0),2);
}
HighGui.imshow("img2", frame);
int keyboard = HighGui.waitKey(30);
if (keyboard == 'q'|| keyboard == 27) {
break;
}
}
System.exit(0);
}
}
public class CamshiftDemo {
public static void main(String[] args) {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
new Camshift().run(args);
}
}

70
samples/java/tutorial_code/video/meanshift/MeanshiftDemo.java Normal file
View File

@ -0,0 +1,70 @@
import java.util.Arrays;
import org.opencv.core.*;
import org.opencv.highgui.HighGui;
import org.opencv.imgproc.Imgproc;
import org.opencv.video.Video;
import org.opencv.videoio.VideoCapture;
class Meanshift {
public void run(String[] args) {
String filename = args[0];
VideoCapture capture = new VideoCapture(filename);
if (!capture.isOpened()) {
System.out.println("Unable to open file!");
System.exit(-1);
}
Mat frame = new Mat(), hsv_roi = new Mat(), mask = new Mat(), roi;
// take the first frame of the video
capture.read(frame);
//setup initial location of window
Rect track_window = new Rect(300, 200, 100, 50);
// setup initial location of window
roi = new Mat(frame, track_window);
Imgproc.cvtColor(roi, hsv_roi, Imgproc.COLOR_BGR2HSV);
Core.inRange(hsv_roi, new Scalar(0, 60, 32), new Scalar(180, 255, 255), mask);
MatOfFloat range = new MatOfFloat(0, 256);
Mat roi_hist = new Mat();
MatOfInt histSize = new MatOfInt(180);
MatOfInt channels = new MatOfInt(0);
Imgproc.calcHist(Arrays.asList(hsv_roi), channels, mask, roi_hist, histSize, range);
Core.normalize(roi_hist, roi_hist, 0, 255, Core.NORM_MINMAX);
// Setup the termination criteria, either 10 iteration or move by atleast 1 pt
TermCriteria term_crit = new TermCriteria(TermCriteria.EPS | TermCriteria.COUNT, 10, 1);
while (true) {
Mat hsv = new Mat() , dst = new Mat();
capture.read(frame);
if (frame.empty()) {
break;
}
Imgproc.cvtColor(frame, hsv, Imgproc.COLOR_BGR2HSV);
Imgproc.calcBackProject(Arrays.asList(hsv), channels, roi_hist, dst, range, 1);
// apply meanshift to get the new location
Video.meanShift(dst, track_window, term_crit);
// Draw it on image
Imgproc.rectangle(frame, track_window, new Scalar(255, 0, 0), 2);
HighGui.imshow("img2", frame);
int keyboard = HighGui.waitKey(30);
if (keyboard == 'q' || keyboard == 27) {
break;
}
}
System.exit(0);
}
}
public class MeanshiftDemo {
public static void main(String[] args) {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
new Meanshift().run(args);
}
}

96
samples/java/tutorial_code/video/optical_flow/OpticalFlowDemo.java Normal file
View File

@ -0,0 +1,96 @@
import java.util.ArrayList;
import java.util.Random;
import org.opencv.core.*;
import org.opencv.highgui.HighGui;
import org.opencv.imgproc.Imgproc;
import org.opencv.video.Video;
import org.opencv.videoio.VideoCapture;
class OptFlow {
public void run(String[] args) {
String filename = args[0];
VideoCapture capture = new VideoCapture(filename);
if (!capture.isOpened()) {
System.out.println("Unable to open this file");
System.exit(-1);
}
// Create some random colors
Scalar[] colors = new Scalar[100];
Random rng = new Random();
for (int i = 0 ; i < 100 ; i++) {
int r = rng.nextInt(256);
int g = rng.nextInt(256);
int b = rng.nextInt(256);
colors[i] = new Scalar(r, g, b);
}
Mat old_frame = new Mat() , old_gray = new Mat();
// Since the function Imgproc.goodFeaturesToTrack requires MatofPoint
// therefore first p0MatofPoint is passed to the function and then converted to MatOfPoint2f
MatOfPoint p0MatofPoint = new MatOfPoint();
capture.read(old_frame);
Imgproc.cvtColor(old_frame, old_gray, Imgproc.COLOR_BGR2GRAY);
Imgproc.goodFeaturesToTrack(old_gray, p0MatofPoint,100,0.3,7, new Mat(),7,false,0.04);
MatOfPoint2f p0 = new MatOfPoint2f(p0MatofPoint.toArray()) , p1 = new MatOfPoint2f();
// Create a mask image for drawing purposes
Mat mask = Mat.zeros(old_frame.size(), old_frame.type());
while (true) {
Mat frame = new Mat(), frame_gray = new Mat();
capture.read(frame);
if (frame.empty()) {
break;
}
Imgproc.cvtColor(frame, frame_gray, Imgproc.COLOR_BGR2GRAY);
// calculate optical flow
MatOfByte status = new MatOfByte();
MatOfFloat err = new MatOfFloat();
TermCriteria criteria = new TermCriteria(TermCriteria.COUNT + TermCriteria.EPS,10,0.03);
Video.calcOpticalFlowPyrLK(old_gray, frame_gray, p0, p1, status, err, new Size(15,15),2, criteria);
byte StatusArr[] = status.toArray();
Point p0Arr[] = p0.toArray();
Point p1Arr[] = p1.toArray();
ArrayList<Point> good_new = new ArrayList<>();
for (int i = 0; i<StatusArr.length ; i++ ) {
if (StatusArr[i] == 1) {
good_new.add(p1Arr[i]);
Imgproc.line(mask, p1Arr[i], p0Arr[i], colors[i],2);
Imgproc.circle(frame, p1Arr[i],5, colors[i],-1);
}
}
Mat img = new Mat();
Core.add(frame, mask, img);
HighGui.imshow("Frame", img);
int keyboard = HighGui.waitKey(30);
if (keyboard == 'q' || keyboard == 27) {
break;
}
// Now update the previous frame and previous points
old_gray = frame_gray.clone();
Point[] good_new_arr = new Point[good_new.size()];
good_new_arr = good_new.toArray(good_new_arr);
p0 = new MatOfPoint2f(good_new_arr);
}
System.exit(0);
}
}
public class OpticalFlowDemo {
public static void main(String[] args) {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
new OptFlow().run(args);
}
}

72
samples/java/tutorial_code/video/optical_flow/OpticalFlowDenseDemo.java Normal file
View File

@ -0,0 +1,72 @@
import java.util.ArrayList;
import org.opencv.core.*;
import org.opencv.highgui.HighGui;
import org.opencv.imgproc.Imgproc;
import org.opencv.video.Video;
import org.opencv.videoio.VideoCapture;
class OptFlowDense {
public void run(String[] args) {
String filename = args[0];
VideoCapture capture = new VideoCapture(filename);
if (!capture.isOpened()) {
//error in opening the video input
System.out.println("Unable to open file!");
System.exit(-1);
}
Mat frame1 = new Mat() , prvs = new Mat();
capture.read(frame1);
Imgproc.cvtColor(frame1, prvs, Imgproc.COLOR_BGR2GRAY);
while (true) {
Mat frame2 = new Mat(), next = new Mat();
capture.read(frame2);
if (frame2.empty()) {
break;
}
Imgproc.cvtColor(frame2, next, Imgproc.COLOR_BGR2GRAY);
Mat flow = new Mat(prvs.size(), CvType.CV_32FC2);
Video.calcOpticalFlowFarneback(prvs, next, flow,0.5,3,15,3,5,1.2,0);
// visualization
ArrayList<Mat> flow_parts = new ArrayList<>(2);
Core.split(flow, flow_parts);
Mat magnitude = new Mat(), angle = new Mat(), magn_norm = new Mat();
Core.cartToPolar(flow_parts.get(0), flow_parts.get(1), magnitude, angle,true);
Core.normalize(magnitude, magn_norm,0.0,1.0, Core.NORM_MINMAX);
float factor = (float) ((1.0/360.0)*(180.0/255.0));
Mat new_angle = new Mat();
Core.multiply(angle, new Scalar(factor), new_angle);
//build hsv image
ArrayList<Mat> _hsv = new ArrayList<>() ;
Mat hsv = new Mat(), hsv8 = new Mat(), bgr = new Mat();
_hsv.add(new_angle);
_hsv.add(Mat.ones(angle.size(), CvType.CV_32F));
_hsv.add(magn_norm);
Core.merge(_hsv, hsv);
hsv.convertTo(hsv8, CvType.CV_8U, 255.0);
Imgproc.cvtColor(hsv8, bgr, Imgproc.COLOR_HSV2BGR);
HighGui.imshow("frame2", bgr);
int keyboard = HighGui.waitKey(30);
if (keyboard == 'q' || keyboard == 27) {
break;
}
prvs = next;
}
System.exit(0);
}
}
public class OpticalFlowDenseDemo {
public static void main(String[] args) {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
new OptFlowDense().run(args);
}
}

156
samples/python/tutorial_code/core/file_input_output/file_input_output.py Normal file
View File

@ -0,0 +1,156 @@
from __future__ import print_function
import numpy as np
import cv2 as cv
import sys
def help(filename):
print (
'''
{0} shows the usage of the OpenCV serialization functionality. \n\n
usage:\n
python3 {0} outputfile.yml.gz\n\n
The output file may be either in XML, YAML or JSON. You can even compress it\n
by specifying this in its extension like xml.gz yaml.gz etc... With\n
FileStorage you can serialize objects in OpenCV.\n\n
For example: - create a class and have it serialized\n
- use it to read and write matrices.\n
'''.format(filename)
)
class MyData:
A = 97
X = np.pi
name = 'mydata1234'
def __repr__(self):
s = '{ name = ' + self.name + ', X = ' + str(self.X)
s = s + ', A = ' + str(self.A) + '}'
return s
## [inside]
def write(self, fs):
fs.write('MyData','{')
fs.write('A', self.A)
fs.write('X', self.X)
fs.write('name', self.name)
fs.write('MyData','}')
def read(self, node):
if (not node.empty()):
self.A = int(node.getNode('A').real())
self.X = node.getNode('X').real()
self.name = node.getNode('name').string()
else:
self.A = self.X = 0
self.name = ''
## [inside]
def main(argv):
if len(argv) != 2:
help(argv[0])
exit(1)
# write
## [iomati]
R = np.eye(3,3)
T = np.zeros((3,1))
## [iomati]
## [customIOi]
m = MyData()
## [customIOi]
filename = argv[1]
## [open]
s = cv.FileStorage(filename, cv.FileStorage_WRITE)
# or:
# s = cv.FileStorage()
# s.open(filename, cv.FileStorage_WRITE)
## [open]
## [writeNum]
s.write('iterationNr', 100)
## [writeNum]
## [writeStr]
s.write('strings', '[')
s.write('image1.jpg','Awesomeness')
s.write('../data/baboon.jpg',']')
## [writeStr]
## [writeMap]
s.write ('Mapping', '{')
s.write ('One', 1)
s.write ('Two', 2)
s.write ('Mapping', '}')
## [writeMap]
## [iomatw]
s.write ('R_MAT', R)
s.write ('T_MAT', T)
## [iomatw]
## [customIOw]
m.write(s)
## [customIOw]
## [close]
s.release()
## [close]
print ('Write Done.')
# read
print ('\nReading: ')
s = cv.FileStorage()
s.open(filename, cv.FileStorage_READ)
## [readNum]
n = s.getNode('iterationNr')
itNr = int(n.real())
## [readNum]
print (itNr)
if (not s.isOpened()):
print ('Failed to open ', filename, file=sys.stderr)
help(argv[0])
exit(1)
## [readStr]
n = s.getNode('strings')
if (not n.isSeq()):
print ('strings is not a sequence! FAIL', file=sys.stderr)
exit(1)
for i in range(n.size()):
print (n.at(i).string())
## [readStr]
## [readMap]
n = s.getNode('Mapping')
print ('Two',int(n.getNode('Two').real()),'; ')
print ('One',int(n.getNode('One').real()),'\n')
## [readMap]
## [iomat]
R = s.getNode('R_MAT').mat()
T = s.getNode('T_MAT').mat()
## [iomat]
## [customIO]
m.read(s.getNode('MyData'))
## [customIO]
print ('\nR =',R)
print ('T =',T,'\n')
print ('MyData =','\n',m,'\n')
## [nonexist]
print ('Attempt to read NonExisting (should initialize the data structure',
'with its default).')
m.read(s.getNode('NonExisting'))
print ('\nNonExisting =','\n',m)
## [nonexist]
print ('\nTip: Open up',filename,'with a text editor to see the serialized data.')
if __name__ == '__main__':
main(sys.argv)