How do I compute the brightness histogram aggregat

问题:

I want to segment car plate to get separate characters. I found some article, where such segmentation performed using brightness histograms (as i understand - sum of all non-zero pixels).

How can i calculate such histogram? I would really appreciate for any help!

回答1:

std::vector<int> computeColumnHistogram(const cv::Mat& in) {

  std::vector<int> histogram(in.cols,0); //Create a zeroed histogram of the necessary size
  for (int y = 0; y < in.rows; y++) {
    p_row = in.ptr(y); ///Get a pointer to the y-th row of the image
    for (int x = 0; x < in.cols; x++)
      histogram[x] += p_row[x]; ///Update histogram value for this image column
  }

  //Normalize if you want (you'll get the average value per column): 
  //  for (int x = 0; x < in.cols; x++)
  //    histogram[x] /= in.rows;

  return histogram;

}

Or use reduce as suggested by Berak, either calling

cv::reduce(in, out, 0, CV_REDUCE_AVG);

or

cv::reduce(in, out, 0, CV_REDUCE_SUM, CV_32S);

out is a cv::Mat, and it will have a single row.