Improved motion estimation search method of multi-view video
A technology of motion estimation search and multi-viewpoint video, which is applied in digital video signal modification, electrical components, image communication, etc., can solve the problems of many search points, unscientific cut-off strategy, and increased calculation amount, so as to reduce encoding time and improve The effect of encoding real-time
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment approach 1
[0041] Specific implementation mode 1. Combination figure 1 Describe this specific implementation mode, an improved multi-viewpoint video motion estimation search method,
[0042] Step 1, search starting point selection, select the coordinates of the vector with the smallest SAD in all predictive vectors as the search starting point, the predictive vector includes the median predictive vector, the motion vector of the left, upper and upper right macroblocks of the same position macroblock in the reference frame and a (0,0) vector;
[0043] Step 2. Determine the SAD value of the starting point of the search, and adopt different hierarchical search strategies for different SAD values; specifically, if the SAD of the starting point is figure 2 with image 3 Shown is the specific search strategy diagram.
[0044] Step 3. Determine the abscissa of the search starting point. If the abscissa is greater than 10, perform a cross-shaped search with a horizontal step of 3 and a vertic...
specific Embodiment
[0059] Specific embodiment: a kind of improved multi-viewpoint video motion estimation search method, it is realized by the following steps:
[0060] Step 1: Select a search starting point, and select the coordinates of the vector with the smallest SAD among all predicted vectors as the search starting point. The predictive vectors include the median predictive vector, the motion vectors of the left, upper and upper right macroblocks of the co-located macroblock in the reference frame, and the (0,0) vector.
[0061] Step 2: Determine the SAD value of the search starting point, and adopt different hierarchical search strategies for different SAD values. Specifically, if the SAD of the starting point<1000, perform the improved 5×5 square search described in step 6, otherwise perform the search described in step 3, step 4, and step 5.
[0062] Step 3. Determine the abscissa of the search starting point. If the abscissa is greater than 10, perform a cross-shaped search with a hor...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More