Inverse synthetic aperture Radar (ISAR) is aimed at imaging non-cooperative targets and it is difficult to obtain large real data sets. Simulation-generated ISAR is considered to be a common solution. However, the existing simulation method of sequence ISAR images often requires target exact three-dimensional models or multi-angle optical images, which is costly and has limited generalization capabilities. To solve these issues, this article uses video images to simulate ISAR images. Video shot boundary detection (SBD) is widely used in extracting video information. This article proposed an SBD algorithm based on D/S evidence theory for multiple features fusion. This algorithm calculates the inter-frame difference for multiple features, and an adaptive threshold with the feature confidence. Then according to the D/S evidence theory, the relationship between the image frames is judged. Finally, this article uses a single shot extracted from the video to simulate ISAR images. Experimental results show that the proposed algorithm can simulate sequence ISAR images effectively.