现代电子技术2025，Vol.48Issue(11)：57-60,4.DOI:10.16652/j.issn.1004-373x.2025.11.009
低比特率下高速运动图像特征自适应识别
Feature adaptive recognition of high-speed moving images at low bit rates
摘要
Abstract
A method of feature adaptive recognition of high-speed moving images at low bit rates is studied to effectively reduce the impact of image noise and distortion on the feature adaptive recognition.Flat blocks within high-speed moving images are detected and obtained by calculating the variance of high-speed moving images and each image block at low bit rates.The adaptive spatial filtering algorithm is used to complete the deblocking effect processing of high-speed moving images at low bit rates,so as to obtain clear high-speed moving images.Binocular disparity method is used to segment clear and high-speed moving images adaptively,so as to obtain high-speed moving object images.The pixel matrix dynamic threshold of sports object images is calculated.After normalizing the image,the adaptive recognition of high-speed moving image features is achieved.Experimental results have shown that the method can effectively remove block effects and improve image clarity of high-speed moving images at low bit rates,and effectively realize adaptive recognition of high-speed moving image features.关键词
低比特率/高速运动/图像特征/自适应识别/空域滤波/双目视差/目标轮廓/灰度值Key words
low bit rate/high-speed movement/image feature/adaptive recognition/spatial filtering/binocular parallax/object profile/gray value分类
信息技术与安全科学引用本文复制引用
葛振斌,高云..低比特率下高速运动图像特征自适应识别[J].现代电子技术,2025,48(11):57-60,4.基金项目
山西省哲学社会科学规划项目(2023YJ125) （2023YJ125）
山西省软科学研究计划项目(2019041023-5) （2019041023-5）
