SEGMENTASI CITRA LAHAN BEKAS TAMBANG TERBUKA MENGGUNAKAN METODE DETEKSI TEPI CANNY

Abstract

Edge detection in the image of mining land become an important  because this process is the first step improve the appearance of the boundary line of an object image. Canny edge detection method is popular and has the advantages of detecting and localizing well, as well as a clear response. Selection of parameters greatly affect the outcome Canny edge. Parameters include standard deviation Gaussian (σ), the size of the Gaussian filter matrix and a threshold value. Canny method has 5 process consists a smoothing, finding gradient, non-maxima suppression, double thresholding and edge tracking by hysteresis. In this study determined the size of the matrix of 5x5, 7x7 and 9x9 while the value of σ of 1, 2.5 and 5.  Thresholds are adjusted using Otsu method. The program is created using programming language C#. The result showed that the greater size of the matrix and value of σ then produce blurred image. For the same matrix size, the greater the value of σ produce the smaller threshold and edge image obtained more detail.