Reputation: 2552
I am running the following code to calculated the Running Average of a collection of images read from a video on OpenCV.
EDIT: (Code updated)
#include <iostream>
#include <string>
#include <iomanip>
#include <sstream>
#include <cstdio>
#include "opencv2/core/core.hpp"
#include "opencv2/opencv.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/video/video.hpp"
using namespace std;
using namespace cv;
int main(int argc, char *argv[]) {
if(argc < 2) {
printf("Quitting. Insufficient parameters\n");
return 0;
}
char c;
int frameNum = -1;
const char* WIN_MAIN = "Main Window";
namedWindow(WIN_MAIN, CV_WINDOW_AUTOSIZE);
VideoCapture capture;
capture.open(argv[1]);
Mat acc, img;
capture.retrieve(img, 3);
acc = Mat::zeros(img.size(), CV_32FC3);
for(;;) {
if(!capture.grab()) {
printf("End of frame\n");
break;
}
capture.retrieve(img, 3);
Mat floating;
img.convertTo(floating, CV_32FC3);
accumulateWeighted(floating, acc, 0.01);
imshow(WIN_MAIN, img);
waitKey(10);
}
return 0;
}
On running the code with a sample video the following error pops up
OpenCV Error: Assertion failed (dst.size == src.size && dst.channels() == cn) in accumulateWeighted, file /usr/lib/opencv/modules/imgproc/src/accum.cpp, line 430
terminate called after throwing an instance of 'cv::Exception'
what(): /usr/lib/opencv/modules/imgproc/src/accum.cpp:430: error: (-215) dst.size == src.size && dst.channels() == cn in function accumulateWeighted
Aborted (core dumped)
What could be the possible reason for the error? Could you please guide me in the right direction?
Compiler used : g++ OpenCV version : 2.4.5
Thanks!
Upvotes: 3
Views: 3457
Reputation: 139
I had the same problem problem and here's my solution.
Mat frame, acc;
// little hack, read the first frame
capture >> frame;
acc = Mat::zeros(frame.size(), CV_32FC3);
for(;;) {
...
Upvotes: 0
Reputation: 39796
from the refman:
src – Input image as 1- or 3-channel, 8-bit or 32-bit floating point.
dst – Accumulator image with the same number of channels as input image, 32-bit or 64-bit
floating-point.
so, your camera input img CV_8UC3, and your acc img is (currently) CV_32F. that's a misfit.
you want 3channel floating point for acc, so that's :
acc = Mat::zeros(img.size(), CV_8UC3);`
for more precision, you want to change your img to float type, too, so it would be:
acc = Mat::zeros(img.size(), CV_32FC3); // note: 3channel now
for (;;) {
if(!capture.grab()) {
printf("End of frame\n");
break;
}
capture.retrieve(img); // video probably has 1 stream only
Mat floatimg;
img.convertTo(floatimg, CV_32FC3);
accumulateWeighted(floatimg, acc, 0.01);
EDIT:
try to replace your grab/retrieve sequence by:
for(;;) {
capture >> img;
if ( img.empty() )
break;
Upvotes: 3
Reputation: 1405
As I can read in the OpenCV documentation of retrieve
C++: bool VideoCapture::retrieve(Mat& image, int channel=0);
second argument is the channel and in the documentation of accumulateWeighted it says :
C++: void accumulateWeighted(InputArray src, InputOutputArray dst, double alpha, InputArray mask=noArray() )
Parameters:src – Input image as 1- or 3-channel, 8-bit or 32-bit floating point.
But in your code :
capture.retrieve(img, 2);
I guess you have the wrong channel parameter
Upvotes: 0