Edge Detection and Symmetry Method for Cleft Lip and Palate Children Using Image Processing

摘要

To evaluate the treatment result of the cleft lip and palate children, it needed the specialists to evaluate it on the basis of the nasolabial method. The vital principle of this method was the result of operation of the lip shape to see whether it was symmetrical after being operated. The problem was that it needed at least five specialists to evaluate the treatment result through the nasolabial method. In case there was any disagreement between the specialists, it needed to hold the meeting to summarize the result of the lip operation. According to the found problem, this study proposed the suitable method for the edge detection, and symmetry for the cleft lip and palate children using the image processing based on the criteria of the shape of the vermilion border. For the procedures of analysis, the 5 tested images of the child patient who has been operated to modify the cleft lip and palate condition were examined to measure the efficiency. The comparison of the methods to finding out the suitable edge detection was carried out by Roberts, Prewitt, Sobel, Canny, and Laplacian of Gaussian (LoG) methods. The results of the study revealed that the average of LoG was the suitable method of finding out the shape of the vermilion border as compared to other methods. It had the least pixel difference of all as compared to the actual edge image. Moreover, it was found that the left lip and the right lip had the least distance pixel difference when finding out the symmetry by clipping the image. Thus, this proposed method could be the guideline of evaluating the treatment more accurately and efficiently.