www.pudn.com > firev0.01.rar > extractlfv2.cpp
/*
This file is part of the FIRE -- Flexible Image Retrieval System
FIRE is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
FIRE is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with FIRE; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#include "image.hpp"
#include "rgbimage.hpp"
#include "getpot.hpp"
#include "imagetools.hpp"
#include
using namespace std;
using namespace img;
typedef pair Coordinates;
typedef pair* > ColorLF;
typedef pair* > GrayLF;
void USAGE() {
cout << "NYI" << endl
<< endl;
}
///-------------- methods for color images
double localVar(const RGBImage & img, const int y, const int x, const int winsize) {
ColorPixel mean(0), variance(0);
for(int i=y-winsize;i<=y+winsize;++i) {
for(int j=x-winsize;j<=x+winsize;++j) {
for(int c=0;c<3;++c) {
mean[c]+=img(i,j,c);
variance[c]+=img(i,j,c)*img(i,j,c);
}
}
}
double divid=2*winsize+1;
divid*=divid;
for(int c=0;c<3;++c) {
variance[c]/=divid;
mean[c]/=divid;
}
for(int c=0;c<3;++c) {
variance[c]-=mean[c]*mean[c];
}
return ((variance[0]+variance[1]+variance[2])/3.0);
}
RGBImage* extractFeature(const RGBImage & img, const pair pos, const int winsize) {
RGBImage* result=new RGBImage(2*winsize+1,2*winsize+1);
int ri=0;
int rj=0;
for(int i=pos.first-winsize;i<=pos.first+winsize;++i) {
rj=0;
for(int j=pos.second-winsize;j<=pos.second+winsize;++j) {
(*result)(ri,rj)=img(i,j);
++rj;
}
++ri;
}
return result;
}
vector< ColorLF > extractlf(const RGBImage &img,
int winsize,
double threshold,
int nOfFeatures,
int subsampling,
int) {
vector< pair > localVarVector;
vector result;
if(nOfFeatures!= -1) {
for(int i=winsize;i<=int(img.yDim())-winsize;i+=subsampling) {
for(int j=winsize;j<=int(img.xDim())-winsize;j+=subsampling) {
localVarVector.push_back(pair(localVar(img,i,j,winsize),pair(i,j)));
}
}
DBG(DBG_MESSAGE) << "Got local Variances" << endl;
sort(localVarVector.rbegin(), localVarVector.rend());
DBG(DBG_MESSAGE) << "Sorted" << endl;
for(int i=0;i threshold) {
result.push_back(ColorLF(Coordinates(i,j),extractFeature(img, pair(i,j), winsize)));
}
}
}
}
DBG(DBG_MESSAGE) << "Got "<< result.size() << " features." << endl;
return result;
}
//----------- methods for gray images
double localVar(const BaseImage & img, const int y, const int x, const int winsize) {
double mean=0.0, variance=0.0;
for(int i=y-winsize;i<=y+winsize;++i) {
for(int j=x-winsize;j<=x+winsize;++j) {
mean+=img(i,j);
variance+=img(i,j)*img(i,j);
}
}
double divid=2*winsize+1;
divid*=divid;
variance/=divid;
mean/=divid;
variance-=mean*mean;
return variance;
}
BaseImage* extractFeature(const BaseImage & img, const pair pos, const int winsize) {
BaseImage* result=new BaseImage(2*winsize+1,2*winsize+1);
int ri=0;
int rj=0;
for(int i=pos.first-winsize;i<=pos.first+winsize;++i) {
rj=0;
for(int j=pos.second-winsize;j<=pos.second+winsize;++j) {
(*result)(ri,rj)=img(i,j);
++rj;
}
++ri;
}
return result;
}
vector< GrayLF > extractlf(const BaseImage &img,
int winsize,
double threshold,
int nOfFeatures,
int subsampling,
int) {
vector< pair > localVarVector;
vector result;
if(nOfFeatures!= -1) {
for(int i=winsize;i<=int(img.yDim())-winsize;i+=subsampling) {
for(int j=winsize;j<=int(img.xDim())-winsize;j+=subsampling) {
localVarVector.push_back(pair(localVar(img,i,j,winsize),pair(i,j)));
}
}
DBG(DBG_MESSAGE) << "Got local Variances" << endl;
sort(localVarVector.rbegin(), localVarVector.rend());
DBG(DBG_MESSAGE) << "Sorted" << endl;
for(int i=0;i threshold) {
result.push_back(GrayLF(Coordinates(i,j),extractFeature(img, pair(i,j), winsize)));
}
}
}
}
DBG(DBG_MESSAGE) << "Got "<< result.size() << " features." << endl;
return result;
}
////--------------- main program
int main(int argc, char **argv) {
GetPot cl(argc, argv);
string cmdline="";
for(int i=0;i img;
img.load(filename);
if(usePadding) {
img=padding(img, img.yDim()+2*winsize, img.xDim()+2*winsize);
}
vector< ColorLF > vec=extractlf(img,winsize,threshold, nOfFeatures, subsampling, saveImages);
if(cl.search("-images")) {
ostringstream fn;
for(int i=0;iwritePNG(fn.str().c_str());
}
}
ofstream ofs(string(filename+suffix).c_str());
if(!ofs) {
ERR << "Cannot open '" << filename << suffix << "' for writing." << endl;
} else {
ofs << "# local features for color image "<< filename << endl
<< "# cmdline " << cmdline << endl
<< "# format: posy posx [data1 ... datan]" << endl
<< "winsize " << winsize << endl
<< "threshold " << threshold << endl
<< "subsampling " << subsampling << endl
<< "pixeldim 3" << endl;
for(int i=0;iyDim());++y) {
for(int x=0;xxDim());++x) {
ofs << int((*(vec[i].second))(y,x,c)) << " ";;
}
}
ofs << " ";
}
ofs << endl;
}
}
ofs.close();
} else if(cl.search("-gray")) {
DBG(DBG_MESSAGE) << "Processing gray images" << endl;
BaseImage img;
img.load(filename);
if(usePadding) {
img=padding(img, img.yDim()+2*winsize, img.xDim()+2*winsize);
}
vector< GrayLF > vec=extractlf(img,winsize,threshold, nOfFeatures, subsampling, saveImages);
if(cl.search("-images")) {
ostringstream fn;
for(int i=0;iwritePNG(fn.str().c_str());
}
}
ofstream ofs(string(filename+suffix).c_str());
if(!ofs) {
ERR << "Cannot open '" << filename << suffix << "' for writing." << endl;
} else {
ofs << "# local features for gray image "<< filename << endl
<< "# cmdline: " << cmdline << endl
<< "# format: posy posx [data1 ... datan]" << endl
<< "winsize " << winsize << endl
<< "threshold " << threshold << endl
<< "subsampling " << subsampling << endl
<< "pixeldim 1" << endl;
for(int i=0;iyDim());++y) {
for(int x=0;xxDim());++x) {
ofs << int((*(vec[i].second))(y,x)) << " ";;
}
}
ofs << endl;
}
}
ofs.close();
} else {
USAGE();
}
}