EAR IMAGE SEGMENTATION WITH EDGE DETECTION METHOD ON CANNY AND LAPLACE ALGORITHMS

Abstract

Technology in identifying ear shapes is the most important step in an automatic ear shape identification system. The purpose of this paper is to introduce an approach to image segmentation, color by determining the pixel values in the database and scanning results, the similarity results are formed, the error value of each image. The method used is color segmentation based on RGB(red, green, blue) values, edge detection with the canny and laplace methods and the results of the segmentation. The results obtained are that the program that has been created can identify the shape of the ear image in the database compared to the scanning results using the segmentation method and calculate the number of image pixels between the database image and the scanned image where the minimum number of pixels for the ear shape image in the database is 452 pixels, while the total the maximum pixels is 3028 pixels. For the image of the shape of the ear the result of scanning the minimum number of pixels is 419 pixels and the maximum number of pixels is 2742 pixels. The percentage of identification results for the shape of the ear has an average similarity level: 92%, the results of this study show a very high level of accuracy. The percentage of error in identifying the shape of the ear has an average error rate of: 8%, the results of this study indicate a very low error rate. a conclusion that comparing one image with another image will get a very high level of accuracy in the canny image results are better because the edge detection is clearer and the noise is less. While image laplace is worse because there is a lot of noise