Stereo matching is one of the methods in computer vision and image processing. There have numerous algorithms that have been found associated between disparity maps and ground truth data. Stereo Matching Algorithms were applied to obtain high accuracy of the depth as well as reducing the computational cost of the stereo image or video. The smoother the disparity depth map, the better results of triangulation can be achieved. The selection of an appropriate set of stereo data is very important because these stereo pairs have different characteristics. This paper discussed the performance analysis on stereo matching algorithm through Peak Signal to Noise Ratio (PSNR in dB), Structural Similarity (SSIM), the effect of window size and execution time for different type of techniques such as Sum Absolute Differences (SAD), Sum Square Differences (SSD), Normalized Cross Correlation (NCC), Block Matching (BM), Global Error Energy Minimization by Smoothing Functions, Adapting BP and Dynamic Programming (DP). The dataset of stereo images that used for the experimental purpose is obtained from Middlebury Stereo Datasets.