Video elastic motion estimation method based on 2bit depth pixel

Abstract

The invention discloses a video elastic motion estimation method based on a 2bit depth pixel, and belongs to the field of video code compression, mainly aiming at the deficiency that an elastic motionestimation algorithm based on an 8bit depth pixel is high in computation complexity and poor in instantaneity. Gradients of a reference frame and a current frame are computed via a Prewitt operator,and then 2bit transformation is performed on the current frame and the reference frame by using the mean value and the variance of the gradients; compared with the traditional 2bit transformation based on the mean value and the variance of the pixels, the transformation can better reserve the main edge contour of a video frame and improve motion estimation accuracy; through performing binaryzationon a Gaussian-Newton method based on the 8bit depth pixel, repeated computation of a Hesse matrix and an inverse matrix thereof during each iteration process is avoided, and OR operation is used forreplacing multiplication, and thus the purpose of reducing the complexity of motion estimation time is achieved.

Description

technical field [0001] The invention relates to the field of video coding and compression, in particular to an elastic motion estimation method based on 2-bit depth pixels with fast operation speed, good motion compensation quality and high practicability. Background technique [0002] Since the development of the first generation of video coding standards, although the compression efficiency has been continuously improved, the basic framework of video coding technology has not changed substantially. It is always based on the closed-loop hybrid structure of differential prediction + transformation in units of blocks. Under this framework, motion estimation is an effective time-dimension differential prediction technique, which is widely used in video coding standards such as H.261 / 2 / 3, MPEG-1 / 2 / 4, H.264 / AVC and AVS , and most of the improvement in video coding efficiency over the years is due to the use of better motion estimation algorithms. However, the computing resource...

AI Technical Summary

Problems solved by technology

[0004] On the one hand, the existing experimental results show that the traditional elastic motion estimation based on 8-bit depth pixels has high requirements in terms of hardware design and energy consumption, and cannot well meet the requirements of low energy consumption and high speed proposed by some video coding applications.

Therefore, some researchers used a smooth convolution kernel to perform 1-bit transformation on the video frame, down-sampled the 8-bit depth pixels to 1-bit depth, and then used the Gauss-Newton method to solve the elastic motion estimation based on 1-bit depth pixels, but its motion Significantly lower compensation quality

In addition, some researchers have proposed a 2-bit transformation using the mean and variance of pixels, but this method is not complete enough for the main edge contour of the video frame, and the motion estimation efficiency is still not high, and it should not be used in elastic motion estimation.

[0005] On the other hand, the Gauss-Newton algorithm of the elastic motion model needs to calculate partial derivatives, Hessian matrix, inverse matrix, matrix multiplication, bilinear interpolation and motion compensation error for each iteration, and its computational complexity is even higher than that of the block translation model full search, which largely limits its usefulness