| 摘要 |
PROBLEM TO BE SOLVED: To reduce the information amount of a background when camera control is performed and to keep the area of a moving object at a high resolution by turning an internal part whose contour is detection blocks provided with the same vector from a motion vector for each small block to a moving object area and performing moving image camera control by the horizontal/ vertical components of the vector provided in the area. SOLUTION: In a motion vector detection part 10, a frame is divided into motion vector detection blocks and the motion vector of the moving object is detected by the difference of a previous frame and a present frame and outputted to a post stage. In an area extraction part 30, the internal part whose contour is the adjacent plural motion vector detection blocks provided with the same vector amount is turned to the moving object area and the vector for representing the area is obtained. In a vector computing part 40, it is decomposed into the horizontal/vertical components and outputted to a camera control part 50. Also, an encoding control part 60 controls quantization performed in an inter-frame prediction control part 20 by a motion vector detection block unit based on moving object area information from the area extraction part 30. |
