Geek asks and answers

How to match the optical field of a stream (float) with pixel data (char) for image distortion?

I played with optical stream features in OpenCV and got stuck. I have successfully created the fields and maps of the optical flux X and Y using the Farneback method, but I do not know how to apply this to the coordinates of the input image to warp the images. The resulting X and Y fields are 32-bit float (0-1.0), but how does this translate to the coordinates of the input and output images? For example, 1.0 of what? Image width? The difference between the two?

Also, I'm not sure what my loop will look like to apply the / warp transform. I have done many cycles to change the color, but the pixels always remain in one place. Moving pixels around is a new territory for me!

Update: I got this to work, but the resulting image is randomly:

//make a float copy of 8 bit grayscale source image
IplImage *src_img = cvCreateImage(img_sz, IPL_DEPTH_32F, 1);
cvConvertScale(input_img,src_img,1/255.0); //convert 8 bit to float

//create destination image
IplImage *dst_img = cvCreateImage(img_sz, IPL_DEPTH_32F, 1);

for(y = 0; y < flow->height; y++){

    //grab flow maps for X and Y
    float* vx = (float*)(velx->imageData + velx->widthStep*y);
    float* vy = (float*)(vely->imageData + vely->widthStep*y);

    //coords for source and dest image
    const float *srcpx = (const float*)(src_img->imageData+(src_img->widthStep*y));
    float *dstpx = (float*)(dst_img->imageData+(dst_img->widthStep*y));

    for(x=0; x < flow->width; x++)
    {
        int newx = x+(vx[x]);
        int newy = (int)(vy[x])*flow->width;
        dstpx[newx+newy] = srcpx[x];
    }
}

I could not get this to work. The output was just distorted noise:

cvRemap(src_img,dst_img,velx,vely,CV_INTER_CUBIC,cvScalarAll(0));
+3
source share
2 answers

Stream vectors are velocity values. If the pixel in image 1 in position (x, y)has a stream vector (vx, vy), then it is evaluated in position (x+vx, y+vy)(therefore, the values ​​are not in the range [0, 1]- they can be larger and be negative too). The easiest way to do the warp is to create floating point images with these values ​​( x+vxfor the x direction, similar to y), and then use cv::remap.

+5
source

Using OpenCV

https://github.com/opencv/opencv/blob/master/samples/python/opt_flow.py

def warp_flow(img, flow):
    h, w = flow.shape[:2]
    flow = -flow
    flow[:,:,0] += np.arange(w)
    flow[:,:,1] += np.arange(h)[:,np.newaxis]
    res = cv2.remap(img, flow, None, cv2.INTER_LINEAR)
    return res

Result Example

0

More articles:


All Articles