Skip to Main Content
This paper proposes a real-time stereo matching algorithm using GPU programming. The likelihood model is implemented using GPU programming for real-time operation. And the prior model is proposed to improve the accuracy of disparity estimation. First, the likelihood matching based on rank transform is implemented in GPU programming. The shared memory handling in graphic hardware is introduced in calculating the likelihood model. The prior model considers the smoothness of disparity map and is defined as a pixel-wise energy function using adaptive interaction among neighboring disparities. The disparity is determined by minimizing the joint energy function which combines the likelihood model with prior model. These processes are performed in the multi-resolution approach. The disparity map is interpolated using the reliability of likelihood model and color-based similarity in the neighborhood. This paper evaluates the proposed approach with the Middlebury stereo images. According to the experiments, the proposed algorithm shows good estimation accuracy over 30 frames/second for 640×480 image and 60 disparity range. The proposed disparity estimation algorithm is applied to real-time stereo camera system such as 3-D image display, depth-based object extraction, 3-D rendering, and so on.