Cost Aggregation with Guided Image Filter and Superpixel for Stereo Matching

Cost aggregation is one of the popular method for stereo matching due to efficiency and effectiveness. Their limitation is a high complexity and some error near the contour, which makes them not to implement in real time. Furthermore, the weakness makes them unattractive for many applications which require the accurate depth information. In this paper, we present a cost aggregation method using the superpixel-based edge-preserving filter and the guided image filter for stereo matching. First, we combine cost using a census transform and truncated absolute difference of gradients. The guided filter and the super pixel based smooth filter are exploited for the cost aggregation in order. In order to refine depth information, we apply occlusion handling and median filter. Consequently, the proposed method increases the accuracy of the depth map, and experimental results show that the proposed method generates more robust depth maps compared to the conventional methods.