Implementasi Algoritma Region Growing untuk Segmentasi Retakan Bidang Batuan

Abstract

Crack is a factor that caused rock discontinuity. Rock discontinuity is an important input variable in slope stability analysis in open pit mining area. Slope stability analysis should be conducted periodically to prevent landslide which can cause damages and fatalities. The data acquisition of rock discontinuity commonly is taking place directly in the field on the slope to be analyzed, which is time consuming and expensive. The objective of this research is to review the image processing application to identify rock discontinuity in particular rock crack. The image processing approach which is applied in this research is segmentation using region growing algorithm. This algorithm was chosen because of its flexibility to reduce noise with tuning the criteria of threshold value. The algorithm was implemented in C++ programming language. 35 image samples were tested to rate the segmentation result. The result shows that this algorithm was quite successful to detect rock cracks. Furthermore the result can be optimized with increase the threshold value to minimize the isolated pixel which is noise.