PROBLEM TO BE SOLVED: To substantially eliminate one or more defects of an existing configuration or to at least improve the defect. SOLUTION: The method for extracting a key frame from a video sequence consisting of MPEG-compressed video data having a block motion vector, includes a step (104) where the MPEG-compressed video data are received, a step (106) where the received data are partially decompressed to acquire a block motion vector, and the block motion vector is converted into a forward block motion vector to generate a global motion signal, a step (108) where a dominant global direction cluster is generated, and a step (110) where the generated dominant global direction cluster is used to select a potential key frame of the video sequence, and the selected, potential key frame is decompressed to extract the key frame.
展开▼