吉首大学学报（自然科学版）2023，Vol.44Issue(6)：1-8,8.DOI:10.13438/j.cnki.jdzk.2023.06.001
基于局部卷积的视频图像像素补偿方法
Design of Pixel Compensation Algorithm for Video Image Based on Local Convolution
摘要
Abstract
A video image pixel compensation method based on local convolution(local convolution com-pensation)is proposed to repair missing pixels in high-speed moving targets.The method introduces local convolution to segment the neural network model and reconstruct video image frames.It activates the source video image and its corresponding pixel space,applies linear mean compensation to fill the interpo-lated frame holes in the video image,and evaluates the matching degree of the video image to complete the pixel compensation.Comparative experimental results show that the average gradient of the local convolution compensation method is 11.239,significantly higher than the other three comparative meth-ods,which indicates that the video image after compensation using the local convolution method has higher clarity.关键词
局部卷积/视频图像/补偿方法/视频帧/图像块Key words
local convolution/video image/compensation method/video frame/image block分类
计算机与自动化引用本文复制引用
查怀志..基于局部卷积的视频图像像素补偿方法[J].吉首大学学报（自然科学版）,2023,44(6):1-8,8.基金项目
安徽省高校人文社会科学研究项目(SK2018A0557) （SK2018A0557）
