Method for de-interlacing unmanned substation surveillance video based on CLG (Combining Local and Global Algorithm) and AVS (Audio Video coding Standard)

Abstract

The invention discloses a method for de-interlacing an unmanned substation surveillance video based on CLG (Combining Local and Global Algorithm) and AVS (Audio Video coding Standard). The method comprises the steps as follows: carrying out mode division on input surveillance video original images; carrying out CLG motion estimation treatment on one part of the surveillance video original images after mode division, and carrying out AVS block-based motion estimation treatment on another part of the surveillance video original images after mode division; and carrying out motion compensation on the surveillance video images after CLG motion estimation and AVS block-based motion estimation treatment to reconstruct a de-interlaced video, namely outputting a de-interlaced surveillance video on a computer screen. According to the method, the reliability and the coding efficiency of motion estimation are improved, and the definition and the real-time performance of the processed video are guaranteed; and the method is high in practicality and expandability, can be quickly deployed by upgrading of corresponding modules, and can satisfactorily meet the substation video surveillance requirements.

Description

technical field [0001] The invention relates to the technical category of an unattended substation monitoring method in the electric power industry, in particular to a deinterlacing method for unattended substation monitoring video based on CLG and AVS. Background technique [0002] In 2009, the national grid formula proposed to build a strong smart grid with the UHV grid as the backbone grid, the coordinated development of grids at all levels, and the characteristics of automation, informationization, and interaction. Unattended substation is a very important component in the smart grid system. In the unattended substation, through the application of information technology such as video monitoring and image processing, the user can realize the comprehensive monitoring and centralized management of the front-end unmanned or few-person on-duty system, and use the existing network to monitor the front-end image, environment, and equipment. Effective monitoring and management ...

AI Technical Summary

Problems solved by technology

Among them, the block matching algorithm is the most widely used in the four categories, which can meet certain real-time requirements, but artifacts will appear when detecting motion pictures.

However, the computational complexity of the phase plane correlation algorithm is too large. The advantage of the optical flow method is that the optical flow not only carries the motion information of the moving object, but also carries rich information about the three-dimensional structure of the scene. In the case of information, the moving object is detected, but this method has a certain computational complexity, and it is difficult to solve the correct optical flow field under the condition of occlusion and multiple light sources; in addition, the Bayesian algorithm does not have additional Poor performance with auxiliary interpolation

Therefore, for the above four methods, due to the poor correlation between the time domain and the space domain, it is difficult to obtain accurate motion estimation results, and the interleaved coding has the problems of high computational complexity and low efficiency.

Traditional interlaced stereoscopic video performance evaluation framework such as figure 1 As shown, the peak signal-to-noise ratio (PSNR) of the interlaced and progressive video is calculated and compared separately, which is the peak signal-to-noise ratio calculated by comparing with the progressive video, rather than the comparison of the original image used, because different benchmarks are used : interlaced video and progressive video, its performance evaluation has low precision and less convincing results

Images