AMP's profile - activity

sprintf(s,"YOUR_DIRECTORY/frame%d.jpg",num);

Dilation and erosion work on white pixels(pixels with 1 value) so when you dilate above image you expand white area and when eroding reduce white area.

So there is not any bug!

I think you can't find them,but you can train any thing you want with sample images of that.

Cascade Classifier Training

you don't need to match histograms,if you find the color that maximize histogram you can change color of the other image to match the color of face!

First you should detect face.you can use OpenCV Face detection.

then you can compute histogram of detected face.the most frequent color is face's color and you can use it to change color of body.

You should write this part out of Loop

writer = cv2.VideoWriter(vidDir, fourcc, 30, (width,height))

because it construct a new video file at each iteration.

Squares.cpp is a sample code(inside your OpenCV folder) for detecting squares in c++,you should read it and change it to java.if you want to find another objects you should change few parameters to find triangle, circle,...

I don't think so...

You can get idea from "squares.cpp" in samples folder of OpenCV.

Check out DCRaw

Hi,

you should keep edge frame of last iteration of loop then compute " absdiff " between new edge frame and last edge frame.

if (c == 'f')

{

sprintf(s,"frame%d.jpg",num);

cvSaveImage(s,frame);

}

this part save image when you press f

if you want to process them you can type what you want in "for loop" and then save output. cvSaveImage(s,output)

Try this.(press "f" button to save frame)
it's your code with some changes!

#include<opencv\cv.h>
#include<opencv\highgui.h>
#include<opencv\ml.h>
#include<opencv\cxcore.h>
#include<fstream>



int main( int argc, char** argv ) {
cvNamedWindow( "DisplayVideo", CV_WINDOW_AUTOSIZE );
CvCapture* capture = cvCreateFileCapture( argv[1] );
IplImage* frame;
int num=0;
char s [20];
while(1) {
num++;
frame = cvQueryFrame( capture );
if( !frame ) break;
cvShowImage( "DisplayVideo", frame );
char c = cvWaitKey(33);
if( c == 27 ) break;
if (c == 'f')
{
sprintf(s,"frame%d.jpg",num);
cvSaveImage(s,frame);

}
}
cvReleaseCapture( &capture );
cvDestroyWindow("DisplayVideo" );
}

click on "Display properties window" button at top of the window or CTRL+P,now you can see your button.

Out put of capture.read() is a bool and an image you should separate them ; and you should use waitKey.

import cv2
capture=cv2.VideoCapture(0)
while True:
    b,frame=capture.read()
    cv2.imshow("image", frame)
    cv2.waitKey(1)

If they are fixed so you only need to map the outputs of these cameras to a surface and then Blend them together(you can do that manually only one time).

I think it's easier than panorama!

Are your webcams fixed?!

Sorry I don't know objectiveC programming and Xcode project...!

It's in your OpenCV folder! ...../OpenCV-2.4.3/samples/cpp/squares.cpp

I can mail it to you if you want.

You can use "squares.cpp" in samples folder to implement that for iOS.(you need to edit this code)

I think squares.cpp(or squares.py) in samples folder can help you.

you can change some parts of the code and detect any simple objects (like square,circle,triangle...)

yes,it does create a 3x1 matrix.

you should manually iterate over each pixel by changing i & j in a for loop.

I think this code can help you!

Mat x,y,z;
x=Mat(image.at<cv::Vec3b>(i,j));
x.convertTo(y,c.type()); //c is your 3x3 matrix
z=c*y;