Description
Consider the gray-scale image, Car on Mountain Road.tif below, (a) apply the Marr-Hildreth edge detection algorithm to obtain the edge image. Plot all the images generated during the entire step-by-step procedure of applying the algorithm. Assume two thresholds: 0% and 4% the maximum gray level of the LoG image (refer to Example 10.7 and Figure 10.22). (b) Edge linking by Hough transform: based on the edge map obtained in (a) using 4% max{Log} as the threshold, use the Hough transform to perform edge linking. Firstly, determine the Hough parameter space using 1 increments for and 1-pixel increments for . Make the same plot as Figure 10.31(c) (Example 10.12). From your results of Hough parameter space, determine the possible cells for the license plate to make the same plots as Figures 10.31(d) and (e).
Your report (Word or pdf format) should contain:
- Source codes (30%)
- Figures of the LoG image (10%), binary images by zero-crossings with threshold of 0 and 4% of max(LoG) (30%)
- Figure of Hough parameter space, refer to Fig 10.31(c) (15%)
- Figures of linked edges alone and overlapped on the original image, refer to Fig 10.31(d),(e) (15%)
Note: Images must be plotted with good resolution
