The original image
After conversion to grey scale, and convolution with a smoothing kernel
After convolution with a Laplacian edge detection kernel
The above image equivalent in a 150x150 Hough space showing 15 selected peaks, each of which should correspond to lines in the image
The 15 selected Hough transform peaks represented as lines in the original image, with the original alongside