Perbandingan Metode Sobel, Prewitt, Robert dan Canny pada Deteksi Tepi Objek Bergerak

• Main Author: Wahyu Supriyatin
• Format: Article
• Language: English
• Published: Fakultas Ilmu Komputer UMI 2020-08-01
• Series: Ilkom Jurnal Ilmiah
• Subjects: canny; edge detection; prewitt; robert; sobel
• Online Access: http://jurnal.fikom.umi.ac.id/index.php/ILKOM/article/view/541

Computer vision is one of field of image processing. To be able to recognize a shape, it requires the initial stages in image processing, namely as edge detection. The object used in tracking in computer vision is a moving object (video). Edge detection is used to recognize edges of objects and reduce existing noise. Edge detection algorithms used for this research are using Sobel, Prewitt, Robert and Canny. Tests were carried out on three videos taken from the Matlab library. Testing is done using Simulik Matlab tools. The edge and overlay test results show that the Prewitt algorithm has better edge detection results compared to other algorithms. The Prewitt algorithm produces edges whose level of accuracy is smoother and clearer like the original object. The Canny algorithm failed to produce an edge on the video object. The Sobel and Robert algorithm can detect edges, but it is not clear as Prewitt does, because there are some missing edges.