Hi, I'm trying to find whether the image contains robs attached on the door or not. I used Ada-boost's algorithm for training using the area of rod and also used Hough transform to find out the position of rods. For the Ada-boost training part, the result is not acceptable, as the area extracted is much larger than the expected. For the Hough Transform part, there are too many false positive results, which sometimes even missed the actual area of robs.
My purpose is to count the number of robs attached on the door. I am so confused about the result at present and don't know how to deal with the results obtained by the Hough Transform. Thanks for your suggestion in advance.
