The invention provides a video compression method. The video compression method comprises the steps that each video frame and a time point corresponding to each video frame are extracted from videos; the video frames are compared step by step so as to obtain multiple types of video frames provided with the same static images; the current types of video frames provided with the same static images are segmented to obtain static images and multiple dynamic images, wherein the number of the dynamic images is equal to the number of the current types of video frames; the time points of the current types of video frames correspond to the multiple dynamic images in a one-to-one mode; the static images and the multiple dynamic images are compressed to obtain compressed files. In the mode, the compression degree and compression speed can be improved by the adoption of the video compression method.