Video content summarization provides an effective way to accelerating video browsing and retrieval. In this paper, we propose a novel approach to automatically generate the video summary. Firstly, the video structure is analyzed by spatio-temporal analysis. Then, we detect video non-trivial repeating patterns to remove the visual-content redundancy among video stream. Moreover, an importance evaluation model (IEM) is adopted to automatically determine the importance of each video shot according to the user need. This aims to construct video summarization with the most informative shots selected from groups of similar shots. Experimental results indicate that the proposed algorithm is more effective than existing approaches in video summarization generation.