| 摘要 |
The present invention provides a method for dividing foreground and background areas, characterized in that the method includes: obtaining a second image through down sampling a first image, and pre-setting image categories for dividing; determining an image category of each macro block of the second image based on the second image; performing regional connection processing according to the determined image category situation of each macro block and obtaining the final image area division; adjusting encoding strategies by combining the final image area division and obtaining new encoding and decoding images. The present invention also provides a system for dividing foreground and background areas. The method and system are implemented based on characteristics of different areas of the image in encoding, so the method is more suitable for coders. Moreover, the present invention provides a face and background area division method. By the face and background area division method, important visual areas of a person are protected at a low code rate, and the coding strategy is regulated by utilizing the area division information of the image, thereby overall effects are improved. |
