generic-library/opencv
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

move image extraction logic from Octave to Dataset class

Browse Source
This commit is contained in:
marina.kolpakova 2013-01-09 15:21:04 +04:00
parent a3600b94db
commit a0e93d0482
3 changed files with 109 additions and 83 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

92
apps/sft/include/sft/octave.hpp
View File

@ -46,17 +46,70 @@
#include <opencv2/ml/ml.hpp>
#include <sft/common.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#include <opencv2/highgui/highgui.hpp>
namespace sft
{
class Dataset
class Preprocessor
{
public:
Dataset(const sft::string& path, const int octave);
Preprocessor() {}
// private:
svector pos;
svector neg;
void apply(const cv::Mat& frame, cv::Mat& integrals) const
{
CV_Assert(frame.type() == CV_8UC3);
int h = frame.rows;
int w = frame.cols;
cv::Mat channels, gray;
channels.create(h * BINS, w, CV_8UC1);
channels.setTo(0);
cvtColor(frame, gray, CV_BGR2GRAY);
cv::Mat df_dx, df_dy, mag, angle;
cv::Sobel(gray, df_dx, CV_32F, 1, 0);
cv::Sobel(gray, df_dy, CV_32F, 0, 1);
cv::cartToPolar(df_dx, df_dy, mag, angle, true);
mag *= (1.f / (8 * sqrt(2.f)));
cv::Mat nmag;
mag.convertTo(nmag, CV_8UC1);
angle *= 6 / 360.f;
for (int y = 0; y < h; ++y)
{
uchar* magnitude = nmag.ptr<uchar>(y);
float* ang = angle.ptr<float>(y);
for (int x = 0; x < w; ++x)
{
channels.ptr<uchar>(y + (h * (int)ang[x]))[x] = magnitude[x];
}
}
cv::Mat luv, shrunk;
cv::cvtColor(frame, luv, CV_BGR2Luv);
std::vector<cv::Mat> splited;
for (int i = 0; i < 3; ++i)
splited.push_back(channels(cv::Rect(0, h * (7 + i), w, h)));
split(luv, splited);
float shrinkage = static_cast<float>(integrals.cols - 1) / channels.cols;
CV_Assert(shrinkage == 0.25);
cv::resize(channels, shrunk, cv::Size(), shrinkage, shrinkage, CV_INTER_AREA);
cv::integral(shrunk, integrals, cv::noArray(), CV_32S);
}
enum {BINS = 10};
};
struct ICF
@ -74,7 +127,7 @@ struct ICF
}
float operator() (const Mat& integrals, const cv::Size& model) const
float operator() (const cv::Mat& integrals, const cv::Size& model) const
{
int step = model.width + 1;
@ -95,11 +148,11 @@ private:
cv::Rect bb;
int channel;
friend void write(cv::FileStorage& fs, const string&, const ICF& f);
friend void write(cv::FileStorage& fs, const std::string&, const ICF& f);
friend std::ostream& operator<<(std::ostream& out, const ICF& f);
};
void write(cv::FileStorage& fs, const string&, const ICF& f);
void write(cv::FileStorage& fs, const std::string&, const ICF& f);
std::ostream& operator<<(std::ostream& out, const ICF& m);
class ICFFeaturePool : public cv::FeaturePool
@ -108,7 +161,8 @@ public:
ICFFeaturePool(cv::Size model, int nfeatures);
virtual int size() const { return (int)pool.size(); }
virtual float apply(int fi, int si, const Mat& integrals) const;
virtual float apply(int fi, int si, const cv::Mat& integrals) const;
virtual void preprocess(const cv::Mat& frame, cv::Mat& integrals) const;
virtual void write( cv::FileStorage& fs, int index) const;
virtual ~ICFFeaturePool();
@ -124,12 +178,30 @@ private:
static const unsigned int seed = 0;
Preprocessor preprocessor;
enum { N_CHANNELS = 10 };
};
using cv::FeaturePool;
class Dataset
{
public:
typedef enum {POSITIVE = 1, NEGATIVE = 2} SampleType;
Dataset(const sft::string& path, const int octave);
cv::Mat get(SampleType type, int idx) const;
int available(SampleType type) const;
private:
svector pos;
svector neg;
};
// used for traning single octave scale
class Octave : cv::Boost
{
@ -163,7 +235,7 @@ protected:
const cv::Mat& sampleIdx=cv::Mat(), const cv::Mat& varType=cv::Mat(), const cv::Mat& missingDataMask=cv::Mat());
void processPositives(const Dataset& dataset, const FeaturePool* pool);
void generateNegatives(const Dataset& dataset);
void generateNegatives(const Dataset& dataset, const FeaturePool* pool);
float predict( const Mat& _sample, const cv::Range range) const;
private:

98
apps/sft/octave.cpp
View File

@ -44,9 +44,6 @@
#include <sft/random.hpp>
#include <glob.h>
#include <opencv2/imgproc/imgproc.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <queue>
// ============ Octave ============ //
@ -138,85 +135,26 @@ void sft::Octave::setRejectThresholds(cv::Mat& thresholds)
namespace {
using namespace sft;
class Preprocessor
{
public:
Preprocessor(int shr) : shrinkage(shr) {}
void apply(const Mat& frame, Mat& integrals)
{
CV_Assert(frame.type() == CV_8UC3);
int h = frame.rows;
int w = frame.cols;
cv::Mat channels, gray;
channels.create(h * BINS, w, CV_8UC1);
channels.setTo(0);
cvtColor(frame, gray, CV_BGR2GRAY);
cv::Mat df_dx, df_dy, mag, angle;
cv::Sobel(gray, df_dx, CV_32F, 1, 0);
cv::Sobel(gray, df_dy, CV_32F, 0, 1);
cv::cartToPolar(df_dx, df_dy, mag, angle, true);
mag *= (1.f / (8 * sqrt(2.f)));
cv::Mat nmag;
mag.convertTo(nmag, CV_8UC1);
angle *= 6 / 360.f;
for (int y = 0; y < h; ++y)
{
uchar* magnitude = nmag.ptr<uchar>(y);
float* ang = angle.ptr<float>(y);
for (int x = 0; x < w; ++x)
{
channels.ptr<uchar>(y + (h * (int)ang[x]))[x] = magnitude[x];
}
}
cv::Mat luv, shrunk;
cv::cvtColor(frame, luv, CV_BGR2Luv);
std::vector<cv::Mat> splited;
for (int i = 0; i < 3; ++i)
splited.push_back(channels(cv::Rect(0, h * (7 + i), w, h)));
split(luv, splited);
cv::resize(channels, shrunk, cv::Size(), 1.0 / shrinkage, 1.0 / shrinkage, CV_INTER_AREA);
cv::integral(shrunk, integrals, cv::noArray(), CV_32S);
}
int shrinkage;
enum {BINS = 10};
};
}
void sft::Octave::processPositives(const Dataset& dataset, const FeaturePool* pool)
{
Preprocessor prepocessor(shrinkage);
int w = boundingBox.width;
int h = boundingBox.height;
integrals.create(pool->size(), (w / shrinkage + 1) * (h / shrinkage * 10 + 1), CV_32SC1);
int total = 0;
for (svector::const_iterator it = dataset.pos.begin(); it != dataset.pos.end(); ++it)
// for (svector::const_iterator it = dataset.pos.begin(); it != dataset.pos.end(); ++it)
for (int curr = 0; curr < dataset.available( Dataset::POSITIVE); ++curr)
{
const string& curr = *it;
cv::Mat sample = cv::imread(curr);
cv::Mat sample = dataset.get( Dataset::POSITIVE, curr);
cv::Mat channels = integrals.row(total).reshape(0, h / shrinkage * 10 + 1);
sample = sample(boundingBox);
prepocessor.apply(sample, channels);
pool->preprocess(sample, channels);
responses.ptr<float>(total)[0] = 1.f;
if (++total >= npositives) break;
@ -228,7 +166,7 @@ void sft::Octave::processPositives(const Dataset& dataset, const FeaturePool* po
nnegatives = cvRound(nnegatives * total / (double)npositives);
}
void sft::Octave::generateNegatives(const Dataset& dataset)
void sft::Octave::generateNegatives(const Dataset& dataset, const FeaturePool* pool)
{
// ToDo: set seed, use offsets
sft::Random::engine eng(65633343L);
@ -237,9 +175,7 @@ void sft::Octave::generateNegatives(const Dataset& dataset)
// int w = boundingBox.width;
int h = boundingBox.height;
Preprocessor prepocessor(shrinkage);
int nimages = (int)dataset.neg.size();
int nimages = dataset.available(Dataset::NEGATIVE);
sft::Random::uniform iRand(0, nimages - 1);
int total = 0;
@ -248,7 +184,7 @@ void sft::Octave::generateNegatives(const Dataset& dataset)
{
int curr = iRand(idxEng);
Mat frame = cv::imread(dataset.neg[curr]);
Mat frame = dataset.get(Dataset::NEGATIVE, curr);
int maxW = frame.cols - 2 * boundingBox.x - boundingBox.width;
int maxH = frame.rows - 2 * boundingBox.y - boundingBox.height;
@ -262,7 +198,7 @@ void sft::Octave::generateNegatives(const Dataset& dataset)
frame = frame(cv::Rect(dx, dy, boundingBox.width, boundingBox.height));
cv::Mat channels = integrals.row(i).reshape(0, h / shrinkage * 10 + 1);
prepocessor.apply(frame, channels);
pool->preprocess(frame, channels);
dprintf("generated %d %d\n", dx, dy);
@ -386,7 +322,7 @@ bool sft::Octave::train(const Dataset& dataset, const FeaturePool* pool, int wea
// 1. fill integrals and classes
processPositives(dataset, pool);
generateNegatives(dataset);
generateNegatives(dataset, pool);
// 2. only sumple case (all features used)
int nfeatures = pool->size();
@ -455,6 +391,11 @@ sft::ICFFeaturePool::ICFFeaturePool(cv::Size m, int n) : FeaturePool(), model(m)
fill(nfeatures);
}
void sft::ICFFeaturePool::preprocess(const Mat& frame, Mat& integrals) const
{
preprocessor.apply(frame, integrals);
}
float sft::ICFFeaturePool::apply(int fi, int si, const Mat& integrals) const
{
return pool[fi](integrals.row(si), model);
@ -571,4 +512,15 @@ Dataset::Dataset(const string& path, const int oct)
// Check: files not empty
CV_Assert(pos.size() != size_t(0));
CV_Assert(neg.size() != size_t(0));
}
cv::Mat Dataset::get(SampleType type, int idx) const
{
const std::string& src = (type == POSITIVE)? pos[idx]: neg[idx];
return cv::imread(src);
}
int Dataset::available(SampleType type) const
{
return (int)((type == POSITIVE)? pos.size():neg.size());
}

2
modules/ml/include/opencv2/ml/ml.hpp
View File

@ -2140,6 +2140,8 @@ public:
virtual float apply(int fi, int si, const Mat& integrals) const = 0;
virtual void write( cv::FileStorage& fs, int index) const = 0;
virtual void preprocess(const Mat& frame, Mat& integrals) const = 0;
virtual ~FeaturePool() = 0;
};