isra60's profile - activity

Video Capture read() does not maintain average fps in live source (rtsp) I have a Videcapture from a gstreamer pipeline

Maybe a little late... but the conversion is like this.. Y = (0.257 * R) + (0.504 * G) + (0.098 * B) + 16 Cr = V

c++ alternative to this kind of python cv::Mat operations Hi i have this code self.L = (static == 255) * (self.L+1) + (

Differences between Gray to YUV conversion (CvtColor VS Create a Mat with Y component) AFAIK , A YUV image from Gray8 im

This is for a gray 8 camera source and pass to an encoder which needs yuv420 format

How I can convert from GRAY8 to YUV420? I found this code colors https://docs.opencv.org/3.1.0/d7/d1b/group__imgproc__mi

Prototxt and caffemodel files for faster_rcnn.cpp example I dont' find the correct files to make works this sample (new

I have a Cv::Mat calle Prob

prob = deepLearningModel.forward();

This mat have 4 dimensions , and the size of each dimensions are tipically 1-1-nDetections-detectionsParameter

So I want is a 2D Mat based on the last dimensions , I mean nDetections*detectionsPameteter How can I achieve it??

Thanks.

Great!! Thank you .

In my case is not a nDetections columns. my model is a object detection, so it classifies and also give the bounding box

In my case is not a nDetections columns. my model is a object detection, so it classifies and also give the bounding box

How to extract 2D Mat from a 4D Mat I have a Cv::Mat calle Prob prob = deepLearningModel.forward(); This mat have 4 di

Hi. I'm trying to use a track method that does not lost the track when the camera moves (a pan or tilt movement). I have used trackers from tracking module of opencv_extras, For example KCF works perfect with a static video, but when the camera moves the object is lost.

Also I try with Camshift with the same results... Maybe I need to tweek this standar tracking methods?? Or is better use feature detection methods, in that case which method should be better??

I have tried face recognition with the opencv tutorial

It works, but the problem is that i have very poor video performance when a face appear on video stream. I'm trying with a rtsp stream and the video seems very slow and with a poor quality. The video stream works fine when no face appears.

The example I'm using is http://docs.opencv.org/doc/tutorials/objdetect/cascade_classifier/cascade_classifier.html#cascade-classifier

I have an IP camera which streams a rtsp The follow code allows me to show the rtsp stream

int main(int, char**) {
     cv::VideoCapture vcap;
     cv::Mat image;

     const std::string videoStreamAddress = "rtsp://my_rtspstream";

     //open the video stream and make sure it's opened
     if(!vcap.open(videoStreamAddress)) {
        std::cout << "Error opening video stream or file" << std::endl;
        return -1;
    }

    for(;;) {
        if(!vcap.read(image)) {
            std::cout << "No frame" << std::endl;
            cv::waitKey();
        }
        cv::imshow("Output Window", image);
        if(cv::waitKey(1) >= 0) break;
    }
}

The problem is that i have about 450-600 ms of delay difference between what is the camera seeing and what i have o the window of opencv, and i need less becouse is camera with pt control.

Do you have any aproach? I should use another thing instead opencv??

Thanks