Frame rate up-conversion method based on mixed matching of chromaticity

Abstract

The invention discloses a frame rate up-conversion method based on the mixed matching of chromaticity, used for improving the visual quality of a video sequence. The method comprises the steps of firstly designing a block matching criterion by using a chromaticity mixed matching model, wherein the block matching criterion can reflect the level of a difference between the brightness of the video sequence and a chromatic aberration channel; then using an improved three-dimensional recursive search method to compute forward and backward initial motion vector fields, on the basis, providing a time-space motion vector purification method, implicitly adding time-space smooth constraint into the motion vector purification method, and thus obtaining a smoother motion vector field; then providing a high fault-tolerant motion vector smoothening method to inhibit abnormities existing in the motion vector fields; and at last, in order to reduce edge blurring and shielding problems, providing a double-weighted overlapped block motion compensation method, and using the forward and backward initial motion vector fields to jointly generate an interpolating frame. According to the method, the subjective and objective quality of the interpolating frame can be significantly improved, and the low computation complexity is ensured.

Description

technical field [0001] The invention belongs to the technical field of video coding, and relates to a video frame rate up-conversion method, in particular to a frame rate up-conversion method based on chroma mixing and matching. Background technique [0002] With the rapid development of high-definition digital television (HDTV) and high-end multimedia information systems, it is urgent to increase the frame rate of existing video program sources to achieve high-quality visual effects; in addition, for some occasions with limited transmission bandwidth, such as video conferencing and 3G Video conferencing, only using mainstream standards for compression, still cannot achieve the required frame rate. Therefore, it is necessary to reduce the input video frame rate at the encoding end, encode the low frame rate video, and increase the frame rate of the decoded low frame rate video. convert. [0003] The mainstream technology of frame rate up-conversion is the motion compensatio...

AI Technical Summary

Problems solved by technology

One-way motion estimation first estimates the motion vector field between two frames before and after, and then translates each block along the motion trajectory to interpolate a new frame. However, the occlusion problem and abnormal motion vectors in the video frame will cause holes in the new frame. and overlap

According to the assumption of motion symmetry, bidirectional motion estimation assigns a unique motion vector to each block of the interpolated frame, thereby avoiding overlap and hole problems, however, it cannot ensure the accuracy of motion estimation when the object motion loses symmetry, otherwise In addition, since the interpolation block is unknown, the presence of a large number of periodic repetitive patterns in video frames can also lead to inaccurate motion vectors

Multi-hypothesis motion estimation estimates a variety of motion vector fields, and the corresponding motion compensation interpolation makes full use of these motion vector fields with different accuracies to generate interpolated frames with the maximum a posteriori probability criterion, although it can provide significant performance improvements , but its computational complexity is several times that of traditional methods

Images