US 10,893,300 B2
System and method for video processing
Xing Chen, Shenzhen (CN); Zisheng Cao, Shenzhen (CN); and Lei Zhu, Shenzhen (CN)
Assigned to SZ DJI TECHNOLOGY CO., LTD., Shenzhen (CN)
Filed by SZ DJI TECHNOLOGY CO., LTD., Shenzhen (CN)
Filed on Nov. 28, 2017, as Appl. No. 15/824,581.
Application 15/824,581 is a continuation of application No. PCT/CN2015/080230, filed on May 29, 2015.
Prior Publication US 2018/0091828 A1, Mar. 29, 2018
Int. Cl. H04N 19/91 (2014.01); H04N 19/503 (2014.01); H04N 19/105 (2014.01); H04N 19/13 (2014.01)
CPC H04N 19/91 (2014.11) [H04N 19/105 (2014.11); H04N 19/13 (2014.11); H04N 19/503 (2014.11)] 15 Claims
OG exemplary drawing
1. A method for processing a video, comprising:
receiving and storing a reference frame of the video in a memory;
obtaining a prediction table for the reference frame of the video based on a pixel value of each reference pixel within the reference frame, including obtaining a prediction Huffman table by:
for each of reference pixels within the reference frame, determining a prediction value based on respective pixel values of one or more adjacent pixels; and
generating the prediction Huffman table of difference values of the reference pixels, the difference value of one of the reference pixels being based at least on the prediction value of the one of the reference pixels;
storing the prediction table in the memory and releasing the reference frame from the memory;
receiving a target frame arranged after the reference frame in a frame sequence of the video; and
coding, based on the stored prediction table of the reference frame, the target frame of the video.