System and method for calculating 3Deeps action specs motion estimation from the motion vectors in an MPEG file
View Patent ↗Optimal control of 3Deeps Action Specs (also called 3Deep Viewing Spectacles in previous patent applications) for viewing motion pictures filmed in 3D with a visual effect of 3D requires calculation of a single motion vector that characterizes the direction and speed of motion for each motion picture frame. This invention discloses the calculation of the characteristic motion vector or motion estimation value from temporal redundancy information that is stored in the compressed video file bit stream. The principles of this invention can be used with any compressed video file that incorporates temporal redundancy for compression, such as MPEG.
1. A method for calculating a motion estimation value for controlling the lenses of 3Deeps Filter Spectacles comprising:
decompressing a compressed video file;
reading the decompressed video file containing pre-calculated motion vector descriptors, wherein the motion vector descriptors comprises a plurality of frame numbers and a set of motion vectors corresponding to each frame, each frame being segmented into macroblocks, and each macroblock containing a motion vector information;
accessing the set of motion vectors on a frame by frame basis,
calculating a motion estimation value; and
storing in a buffered area or data output file or transmitting the calculated motion estimation value in real-time for controlling the lenses of 3Deeps Filter Spectacles.
2. The method of claim 1 , wherein the frame is an I-frame, P-frame or B-frame, and the motion vector information comprises a lateral or horizontal component of the motion vector of the macroblock, a vertical component of the motion vector of the macroblock, and a rotation component of the motion vector of the macroblock.
3. The method of claim 2 , wherein the frame is the I-frame, and the motion estimation value is an average of all non-zero and non-null macroblock motion vectors.
4. The method of claim 2 , wherein the frame is the P-frame or B-frame, and the motion estimation value is an average of all non-zero and non-null horizontal components of the macroblock motion vectors of the frame.
5. The method of claim 2 , wherein the I-frame has a center of action comprising an area of a number of horizontal macro-blocks and a number of vertical macro-blocks, the motion estimation value being set to the same motion estimation value as the previous frame.
6. The method of claim 2 , wherein the P-frame or B-frame has a center of action comprising an area of a number of horizontal macro-blocks and a number of vertical macro-blocks, the motion estimation value being an average of all horizontal components of macro-blocks of the frame, wherein only those horizontal components that are within the area associated with the center of action and only non-zero and non-null motion vectors are included.
7. The method of claim 2 , wherein the I-frame has an upper region providing surrogate for the background of the frame and a lower region providing surrogate for the foreground of the frame, the motion estimation value being set to the same motion estimation value of the previous frame.
8. The method of claim 2 , wherein the P-frame or B-frame has an upper region providing surrogate for the background of the frame and a lower region providing surrogate for the foreground of the frame, the motion estimation value being the difference between the average of all horizontal components of macro-blocks of the upper region and the average of all horizontal components of macro-blocks of the lower region, wherein only the non-zero and non-null motion vectors are included in the averaging process and all macro-blocks outside the upper region and the lower region are ignored.
9. The method of claim 1 , wherein the compressed video file is in MPEG format.
10. The method of claim 1 , wherein a decompressed video file for display is created containing the motion estimation value embedded into the video file.