Motion compensation deinterlacing method

Abstract

The invention discloses a motion compensation deinterlacing method. The method comprises the following steps: carrying out edge detection to an image; determining whether scenes of previous and next interlacing images change; when the scenes of the previous and the next interlacing images do not change, estimating a same row image motion vector of the previous and the next interlacing images, moving a pixel area block so as to carry out compensation to the motion image according to the same row image motion vector and carrying out interpolation to a vacancy area after moving is performed; when the scenes of the previous and the next interlacing images change, estimating the image motion vector of the adjacent row of the previous image and the current image, moving the pixel area block so as to carry out the compensation to the motion image according to the adjacent row image motion vector and carrying out the interpolation to the vacancy area after the moving is performed. Through the calculation method of the motion vector provided in the invention, the accurate motion vector can be obtained so that an image reduction degree is good, distortion and image damages caused by the error motion vector are not easy to be generated.

Description

technical field [0001] The invention relates to the fields of digital image processing and video display, and more specifically, relates to a motion compensation de-interlacing method. Background technique [0002] Interlaced video is a product of the analog TV era. From video collection and production to video transmission and terminal display, all are in the form of analog. Due to the insufficient transmission bandwidth and the bottleneck of CRT scanning technology, one frame of video is divided into two scanning fields. Due to the persistence of vision effect of the human eye, a complete picture can be viewed, but the interlaced scanning technology will produce shortcomings such as flickering between lines and line creeping. In interlaced video, a frame of video is divided into two fields, one containing even-numbered scanning lines and one containing odd-numbered scanning lines, and each field records pixel values ​​separated in time. Assuming 30 frames per second (fps)...

AI Technical Summary

Problems solved by technology

[0005] 1. When switching between different scenes, the judgment of still images and moving images is wrong, and the wrong motion vector is estimated, resulting in wrong motion compensation

[0006] 2. At the same time, the motion vector estimated by the brightness difference is not very accurate. Motion compensation is sensitive to motion vectors. Inaccurate motion vectors may easily cause image error compensation.

Images