SEGMENTASI CITRA MENGGUNAKAN LEVEL SET UNTUK ACTIVE CONTOUR BERBASIS PARALLEL GPU CUDA

Image segmentation is an important process of image processing in the biomedical field. The method proposed in this study, the method of Level Set Active Contour Local Image Fitting developed by Zhang et al (2009). Zhang et al stated that the level set function can detect the image with intensity inhomogen. In this method, the computing process takes a long time. To address the need for a long time it is proposed to use the GPU CUDA-based parallel computing. CUDA is a programming model that can improve performance by utilizing GPU computing. In images with 256x256 pixel size variation, 512x512 pixels, and 1024x1024 pixels on the NVIDIA GTX 660 GPUs can speed up the process of computing 34-42x faster and NVIDIA GT 635m GPU can accelerate the process of computing the 17-34x faster than using the CPU. With an error rate of 3.76% - 12.62% on the GPU code and error rate of 8.13% -15.83% on the CPU than the matlab code Zhang et