Alexander Alekhin
efcf307b4c
ocl: cleanup dead code in case of disabled OpenCL
2020-08-31 11:30:42 +00:00
Alexander Alekhin
2c1f3487a4
Merge pull request #18037 from danielenricocahall:improve-brisk-init-perf
2020-08-18 20:06:17 +00:00
danielenricocahall
ac177b849c
Improve initialization performance of Brisk
...
reformatting
Improve initialization performance of Brisk
fix formatting
Improve initialization performance of Brisk
formatting
Improve initialization performance of Brisk
make a lookup table for ring
use cosine/sine lookup table for theta in brisk and utilize trig identity
fix ring lookup table
use cosine/sine lookup table for theta in brisk and utilize trig identity
formatting
use cosine/sine lookup table for theta in brisk and utilize trig identity
move scale radius product to ring loop to ensure it's not recomputed for each rot
revert change
move scale radius product to ring loop to ensure it's not recomputed for each rot
remove rings lookup table
move scale radius product to ring loop to ensure it's not recomputed for each rot
fix formatting of for loop
move scale radius product to ring loop to ensure it's not recomputed for each rot
use sine/cosine approximations for brisk lookup table.
add documentation for sine/cosine lookup tables
Improve initialization performance of BRISK
2020-08-18 07:11:21 -04:00
Yosshi999
1834eed809
Merge pull request #18001 from Yosshi999:sift-8bit-descr
...
* 8-bit SIFT descriptors
* use clearer parameter
* update docs
* propagate type info
* overload function for avoiding ABI-break
* bugfix: some values are undefined when CV_SIMD is absent
2020-08-17 10:28:44 +00:00
Yosshi999
922108060d
Merge pull request #17907 from Yosshi999:gsoc_asift-py2cpp
...
* Implement ASIFT in C++
* '>>' should be '> >' within a nested template
* add a sample for asift usage
* bugfix empty keypoints cause crash
* simpler initialization for mask
* suppress the number of lines
* correct tex document
* type casting
* add descriptorsize for asift
* smaller testdata for asift
* more smaller test data
* add OpenCV short license header
2020-08-03 14:11:55 +00:00
Alexander Alekhin
d41b20b268
Merge pull request #17830 from alalek:fix_17815
2020-07-13 18:52:34 +00:00
Alexander Alekhin
36da867caf
features2d: v_fma => v_muladd for integers
2020-07-13 17:31:08 +00:00
Alexander Alekhin
b96e9a4a4a
Merge pull request #17800 from Yosshi999:gsoc_sift-universal-intrinsic
2020-07-11 19:09:21 +00:00
Alexander Alekhin
e18df87a15
Merge pull request #17635 from jsxyhelu:3.4
2020-07-10 15:11:53 +00:00
jsxyhelu
476094ad5a
Use“ moms” replace "contourArea"
...
double area = moms.m00;
is same as
double area = contourArea(contours[contourIdx]);
Not to mention
"moms" already calculated here,"contourArea" should not apply
2020-07-10 11:05:24 +00:00
Yosshi999
0df8fb70b4
use bufferarea for allocating buffer
2020-07-09 16:50:20 +00:00
Maksim Shabunin
4dd9a36a3c
Added test for checkMasks with UMat train descs
2020-07-09 16:22:19 +03:00
Maksim Shabunin
41678fe3d3
Fixed checkMasks in DescriptorMatcher with train descs in UMats
2020-07-09 16:05:20 +03:00
Yosshi999
920c180052
use universal SIMD intrinsics for SIFT
2020-06-30 06:44:12 +00:00
Yosshi999
4064d4c7eb
Merge pull request #17618 from Yosshi999:gsoc_sift-better-test
...
Added/Fixed testcases for SIFT
* merge perf_sift into conventional perf tests
* Fix disabled SIFT scale invariance tests
allows trainIdx duplication in matching scaled keypoints
2020-06-25 11:34:31 +00:00
Alexander Smorkalov
821fac187c
Generate constructor with smart pointer, if it's expected.
2020-05-29 16:40:19 +03:00
Alexander Alekhin
ea0d1424d8
Merge pull request #17287 from tomoaki0705:fixAkazeTestFailureMali
2020-05-14 06:59:28 +00:00
Tomoaki Teshima
35245cb76e
fix test failure on Mali T760 and Mali T628
2020-05-14 05:44:14 +09:00
Alexander Alekhin
1f9713195b
features2d(sift): enable runtime dispatching
2020-05-03 11:59:27 +00:00
Alexander Alekhin
c05ac8729e
features2d: keep history of simd.cpp
2020-04-27 18:19:14 +00:00
Alexander Alekhin
9926a93a78
features2d: copy sift.dispatch.cpp
2020-04-27 18:18:16 +00:00
Alexander Alekhin
7093752cb5
features2d: copy sift.simd.hpp
2020-04-27 18:17:15 +00:00
Alexander Alekhin
a327418767
features2d(sift): move SIFT tests / headers / build fixes
2020-04-23 08:45:22 +00:00
Alexander Alekhin
ed58b5489f
[move sift.cpp] sift: avoid inplace calls of GaussianBlur
...
- should unlock IPP optimizations
original commit: ce7c8f2646
2020-04-16 17:58:45 +00:00
Alexander Alekhin
ef5fa498d4
[move sift.cpp] sift: perf tests and trace regions
...
original commit: a15e105db1
2020-04-16 17:58:45 +00:00
ab-dragon
2947877871
[move sift.cpp] Merge pull request opencv/opencv_contrib#2301 from ab-dragon:conditionally_compute_dog_pyramid
...
Build DoG Pyramid if useProvideKeypoints is false
The buildDoGPyramid operation need not be performed unconditionally. In cases where it is not needed, both memory and speed performance can be improved
original commit: e45887e1c0
2019-11-01 21:28:18 +03:00
Alexander Alekhin
fadb90c579
[move sift.cpp] xfeatures2d: use updated TLS API
...
original commit: 3e4fb8f415
2019-10-20 14:17:05 +00:00
Alexander Alekhin
fd46684bf8
[move sift.cpp] opencv: use cv::AutoBuffer<>::data()
...
original commit: fc69aa57bc
2018-06-13 19:11:18 +00:00
Alexander Alekhin
9badb0d903
[move sift.cpp] xfeatures2d: apply CV_OVERRIDE/CV_FINAL
...
original commit: ec65e5b29c
2018-03-28 16:50:00 +03:00
Vitaly Tuzov
c5f06814bc
[move sift.cpp] Updated internal calls to linear resize to use bit-exact version
...
original commit: 8c394a4f2e
2017-12-14 13:00:09 +03:00
woody.chow
69d43e2997
[move sift.cpp] Remove unnecessary _mm256_round_ps
...
original commit: b5340f6428
2017-09-26 10:12:30 +09:00
Woody Chow
0f0dea79fc
[move sift.cpp] Use TLS instead of mutex in SIFT
...
original commit: ab43a3b2d9
2017-05-31 15:08:32 +09:00
Woody Chow
b6d636214e
[move sift.cpp] Multithreading findScaleSpaceExtremaComputer. Sort the keypoints afterwards to make the output stable
...
original commit: 6be2945abb
2017-05-31 10:02:09 +09:00
Woody Chow
4b64955a12
[move sift.cpp] Parallelize calcDescriptors and buildDoGPyramid. Simplify 2 lines of AVX2 instructions
...
original commit: 443f68cb71
2017-03-24 16:31:18 +09:00
Woody Chow
546239a3a8
[move sift.cpp] Optimize SIFT with AVX2
...
original commit: c5e55dfde9
2017-03-08 10:08:50 +09:00
Suleyman TURKMEN
8be0a3452d
[move sift.cpp] Update sift.cpp
...
original commit: cb7b59f203
2016-12-23 13:21:30 +03:00
Martin Nowak
99d914ea3b
[move sift.cpp] fix overflow issue when computing diagonal
...
- with big images the int multiplication can overflow
original commit: d4df727d38
2016-06-11 17:51:46 +02:00
Maksim Shabunin
24284d3d17
[move sift.cpp] Fixed HAL headers location
...
original commit: f529a1df2b
2015-12-15 18:41:26 +03:00
Vadim Pisarevsky
9fc872b70b
[move sift.cpp] fixed contrib code to match the HAL
...
original commit: cdddcc8237
2015-04-16 22:52:05 +03:00
Vadim Pisarevsky
e3654d5416
[move sift.cpp] refactored xfeatures2d in the same style as features2d
...
original commit: 0cfd795303
2014-10-16 16:33:21 +04:00
Alexander Alekhin
d92bee821d
Merge upstream branch
2020-04-21 06:20:05 +00:00
Alexander Alekhin
df10411e05
features2d(sift): patent expiration note
2020-04-21 06:19:16 +00:00
Alexander Alekhin
44e9fb306d
features2d(sift): code from nonfree module
2020-04-21 04:14:18 +00:00
Alexander Alekhin
f0ffc52435
fix files permissions
2020-04-13 04:29:55 +00:00
Maksim Shabunin
d4d95bd70d
Fixs several problems found by static analysis
2020-03-16 19:15:26 +03:00
Maksim Shabunin
bf96d8239d
Use BufferArea in more places
2020-02-26 11:45:19 +03:00
Alexander Alekhin
e180cc050b
Merge pull request #16236 from alalek:fix_core_simd_emulator
...
* core: fix intrin_cpp, allow to build modules with SIMD emulator
* core(arithm): fix v_zero initialization
* core(simd): 'strict' types for binary/bitwise operations
* features2d: avoid aligned load issue in GCC 5.4 with emulated SIMD
* core(simd): alignment checks in SIMD emulator
2020-01-10 21:31:02 +03:00
Brian Wignall
f9c514b391
Fix spelling typos
...
backport commit 659ffaddb4
2019-12-27 12:46:53 +00:00
Alexander Alekhin
4dfa0a0383
bindings: basic support for #if preprocessor directives
...
- #if 0
- #ifdef __OPENCV_BUILD
2019-12-04 18:42:31 +03:00
Alexander Alekhin
f4d55d512f
imgproc: fix bit-exact GaussianBlur() / sepFilter2D() ( #15855 )
...
* imgproc: fix bit-exact GaussianBlur() / sepFilter2D()
- avoid kernels with bad approximation
- GaussiabBlur - apply error-diffusion approximation for kernel (8-bit fraction)
* java(test): update features2d ref data
* test: update test_facedetect
2019-11-18 01:39:27 +03:00