An efficient method for automatic stereoscopic conversion

In this paper, we present an efficient method for automatic conversion from a monoscopic video to a stereoscopic video. Our method consists of two main steps: object segmentation and depth calculation. The proposed object segmentation algorithm is based on both image motion information and colour information, which can yield visually satisfactory results. The proposed depth calculation algorithm also considers the image motion information, along with an efficient algorithm for depth value correction. KLT tracker algorithm is adopted in this paper for robust and accurate motion estimation, and an approach is proposed to make KLT tracker algorithm more efficient. The experimental results show that our method can achieve good performance.