| 摘要 |
A method for finding the prediction direction in intraframe video coding, which method comprises the following steps: two prediction directions (E, S) are selected, the virtual frames associated with the initial directions (E, S) are calculated, said virtual frames are compared with the current frame and the initial prediction direction (E) or first main direction is selected, the virtual frames associated with the prediction directions (ENE, ESE) immediately adjacent to the first main direction (E) are calculated, said virtual frames are compared with the current frame and the best prediction direction is determined, when the best prediction direction is the first main direction (E) or one of the directions (ENE) immediately adjacent thereto at one end of the prediction directions, then the best direction is the desired prediction direction, otherwise the virtual frame associated with the prediction direction (SE) immediately adjacent to the second main direction (ESE), which is different from the first main direction (E), is calculated, said virtual frames are compared with the current frame and the best prediction direction is determined, when the best prediction direction is said second main direction (ESE) or the direction (SE) immediately adjacent thereto at one end of the prediction directions, then the best direction is the desired prediction direction, otherwise the method is iterated until the desired prediction direction has been determined. The method is useful for video coding. |
