potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-30 07:43:03 +04:00
Code Packages Releases Wiki Activity

Merge pull request #9820 from sovrasov:text_detector_dnn

Browse Source
This commit is contained in:
Vadim Pisarevsky
2017-10-11 13:31:46 +00:00
parent 5e82c98a9f f7175f5050
commit 0873ebb9b0
4 changed files with 558 additions and 413 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/src/caffe/caffe.proto
+2
View File
@@ -145,6 +145,8 @@ message PriorBoxParameter {
optional float step_w = 12;
// Offset to the top left corner of each cell.
optional float offset = 13 [default = 0.5];
// If true, two additional boxes for each center will be generated. Their centers will be shifted by y coordinate.
optional bool additional_y_offset = 14 [default = false];
}
// Message that store parameters used by DetectionOutputLayer
modules/dnn/src/layers/prior_box_layer.cpp
+50 -2
View File
@@ -216,6 +216,14 @@ public:
_stepY = 0;
_stepX = 0;
}
if(params.has("additional_y_offset"))
{
_additional_y_offset = getParameter<bool>(params, "additional_y_offset");
if(_additional_y_offset)
_numPriors *= 2;
}
else
_additional_y_offset = false;
}
bool getMemoryShapes(const std::vector<MatShape> &inputs,
@@ -246,10 +254,11 @@ public:
CV_TRACE_FUNCTION();
CV_TRACE_ARG_VALUE(name, "name", name.c_str());
size_t real_numPriors = _additional_y_offset ? _numPriors / 2 : _numPriors;
if (_scales.empty())
_scales.resize(_numPriors, 1.0f);
_scales.resize(real_numPriors, 1.0f);
else
CV_Assert(_scales.size() == _numPriors);
CV_Assert(_scales.size() == real_numPriors);
int _layerWidth = inputs[0]->size[3];
int _layerHeight = inputs[0]->size[2];
@@ -289,6 +298,19 @@ public:
// ymax
outputPtr[idx++] = (center_y + _boxHeight / 2.) / _imageHeight;
if(_additional_y_offset)
{
float center_y_offset_1 = (h + 1.0) * stepY;
// xmin
outputPtr[idx++] = (center_x - _boxWidth / 2.) / _imageWidth;
// ymin
outputPtr[idx++] = (center_y_offset_1 - _boxHeight / 2.) / _imageHeight;
// xmax
outputPtr[idx++] = (center_x + _boxWidth / 2.) / _imageWidth;
// ymax
outputPtr[idx++] = (center_y_offset_1 + _boxHeight / 2.) / _imageHeight;
}
if (_maxSize > 0)
{
// second prior: aspect_ratio = 1, size = sqrt(min_size * max_size)
@@ -301,6 +323,19 @@ public:
outputPtr[idx++] = (center_x + _boxWidth / 2.) / _imageWidth;
// ymax
outputPtr[idx++] = (center_y + _boxHeight / 2.) / _imageHeight;
if(_additional_y_offset)
{
float center_y_offset_1 = (h + 1.0) * stepY;
// xmin
outputPtr[idx++] = (center_x - _boxWidth / 2.) / _imageWidth;
// ymin
outputPtr[idx++] = (center_y_offset_1 - _boxHeight / 2.) / _imageHeight;
// xmax
outputPtr[idx++] = (center_x + _boxWidth / 2.) / _imageWidth;
// ymax
outputPtr[idx++] = (center_y_offset_1 + _boxHeight / 2.) / _imageHeight;
}
}
// rest of priors
@@ -319,6 +354,18 @@ public:
outputPtr[idx++] = (center_x + _boxWidth / 2.) / _imageWidth;
// ymax
outputPtr[idx++] = (center_y + _boxHeight / 2.) / _imageHeight;
if(_additional_y_offset)
{
float center_y_offset_1 = (h + 1.0) * stepY;
// xmin
outputPtr[idx++] = (center_x - _boxWidth / 2.) / _imageWidth;
// ymin
outputPtr[idx++] = (center_y_offset_1 - _boxHeight / 2.) / _imageHeight;
// xmax
outputPtr[idx++] = (center_x + _boxWidth / 2.) / _imageWidth;
// ymax
outputPtr[idx++] = (center_y_offset_1 + _boxHeight / 2.) / _imageHeight;
}
}
}
}
@@ -385,6 +432,7 @@ public:
bool _flip;
bool _clip;
bool _additional_y_offset;
size_t _numPriors;