Parallel implementation of Sobel filter using CUDA
Hana Ben Fredj, Mouna Ltaif, Anis Ammar, Chokri Souani · 2017
Efficient solutions must be considered, in order to solve the problem of intensive computing of the image processing applications and to achieve high real-time performance. The graphics processing unit (GPU) is an effective and the most recent method used for accelerating extensive calculation algorithms to reduce the execution time by exploiting the power of parallel programming techniques and to obtain the highest performance. In this paper, we present a parallel GPU implementation of an edge detection algorithm with a Sobel operator using CUDA (Compute Unifies Architecture) environment. Furthermore, we analyze and prove the high performance of GPU implementation, by testing the algorithm on a standard central processing unit (CPU) to compare the computational efficiency of these systems. Our experimental results show that the effectiveness of the GPU implementation by its higher performances compared to sequential calculation.