mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2025-08-06 14:36:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Update obj. detect sample

Browse Source
This commit is contained in:
Alexander Nesterov 2019-01-22 15:01:48 -01:00 committed by Alexander Nesterov
parent eb00dce78e
commit cca2c4b314
2 changed files with 35 additions and 57 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

66
samples/dnn/object_detection.cpp
View File

@ -153,51 +153,39 @@ void postprocess(Mat& frame, const std::vector<Mat>& outs, Net& net)
std::vector<int> classIds; std::vector<int> classIds;
std::vector<float> confidences; std::vector<float> confidences;
std::vector<Rect> boxes; std::vector<Rect> boxes;
if (net.getLayer(0)->outputNameToIndex("im_info") != -1) // Faster-RCNN or R-FCN if (outLayerType == "DetectionOutput")
{ {
// Network produces output blob with a shape 1x1xNx7 where N is a number of // Network produces output blob with a shape 1x1xNx7 where N is a number of
// detections and an every detection is a vector of values // detections and an every detection is a vector of values
// [batchId, classId, confidence, left, top, right, bottom] // [batchId, classId, confidence, left, top, right, bottom]
CV_Assert(outs.size() == 1); CV_Assert(outs.size() > 0);
float* data = (float*)outs[0].data; for (size_t k = 0; k < outs.size(); k++)
for (size_t i = 0; i < outs[0].total(); i += 7)
{ {
float confidence = data[i + 2]; float* data = (float*)outs[k].data;
if (confidence > confThreshold) for (size_t i = 0; i < outs[k].total(); i += 7)
{ {
int left = (int)data[i + 3]; float confidence = data[i + 2];
int top = (int)data[i + 4]; if (confidence > confThreshold)
int right = (int)data[i + 5]; {
int bottom = (int)data[i + 6]; int left = (int)data[i + 3];
int width = right - left + 1; int top = (int)data[i + 4];
int height = bottom - top + 1; int right = (int)data[i + 5];
classIds.push_back((int)(data[i + 1]) - 1); // Skip 0th background class id. int bottom = (int)data[i + 6];
boxes.push_back(Rect(left, top, width, height)); int width = right - left + 1;
confidences.push_back(confidence); int height = bottom - top + 1;
} if (width * height <= 1)
} {
} left = (int)(data[i + 3] * frame.cols);
else if (outLayerType == "DetectionOutput") top = (int)(data[i + 4] * frame.rows);
{ right = (int)(data[i + 5] * frame.cols);
// Network produces output blob with a shape 1x1xNx7 where N is a number of bottom = (int)(data[i + 6] * frame.rows);
// detections and an every detection is a vector of values width = right - left + 1;
// [batchId, classId, confidence, left, top, right, bottom] height = bottom - top + 1;
CV_Assert(outs.size() == 1); }
float* data = (float*)outs[0].data; classIds.push_back((int)(data[i + 1]) - 1); // Skip 0th background class id.
for (size_t i = 0; i < outs[0].total(); i += 7) boxes.push_back(Rect(left, top, width, height));
{ confidences.push_back(confidence);
float confidence = data[i + 2]; }
if (confidence > confThreshold)
{
int left = (int)(data[i + 3] * frame.cols);
int top = (int)(data[i + 4] * frame.rows);
int right = (int)(data[i + 5] * frame.cols);
int bottom = (int)(data[i + 6] * frame.rows);
int width = right - left + 1;
int height = bottom - top + 1;
classIds.push_back((int)(data[i + 1]) - 1); // Skip 0th background class id.
boxes.push_back(Rect(left, top, width, height));
confidences.push_back(confidence);
} }
} }
} }

26
samples/dnn/object_detection.py
View File

@ -102,7 +102,7 @@ def postprocess(frame, outs):
classIds = [] classIds = []
confidences = [] confidences = []
boxes = [] boxes = []
if net.getLayer(0).outputNameToIndex('im_info') != -1: # Faster-RCNN or R-FCN if lastLayer.type == 'DetectionOutput':
# Network produces output blob with a shape 1x1xNx7 where N is a number of # Network produces output blob with a shape 1x1xNx7 where N is a number of
# detections and an every detection is a vector of values # detections and an every detection is a vector of values
# [batchId, classId, confidence, left, top, right, bottom] # [batchId, classId, confidence, left, top, right, bottom]
@ -116,23 +116,13 @@ def postprocess(frame, outs):
bottom = int(detection[6]) bottom = int(detection[6])
width = right - left + 1 width = right - left + 1
height = bottom - top + 1 height = bottom - top + 1
classIds.append(int(detection[1]) - 1) # Skip background label if width * height <= 1:
confidences.append(float(confidence)) left = int(detection[3] * frameWidth)
boxes.append([left, top, width, height]) top = int(detection[4] * frameHeight)
elif lastLayer.type == 'DetectionOutput': right = int(detection[5] * frameWidth)
# Network produces output blob with a shape 1x1xNx7 where N is a number of bottom = int(detection[6] * frameHeight)
# detections and an every detection is a vector of values width = right - left + 1
# [batchId, classId, confidence, left, top, right, bottom] height = bottom - top + 1
for out in outs:
for detection in out[0, 0]:
confidence = detection[2]
if confidence > confThreshold:
left = int(detection[3] * frameWidth)
top = int(detection[4] * frameHeight)
right = int(detection[5] * frameWidth)
bottom = int(detection[6] * frameHeight)
width = right - left + 1
height = bottom - top + 1
classIds.append(int(detection[1]) - 1) # Skip background label classIds.append(int(detection[1]) - 1) # Skip background label
confidences.append(float(confidence)) confidences.append(float(confidence))
boxes.append([left, top, width, height]) boxes.append([left, top, width, height])