CAJ | 학술논문

在交通监控中,利用WMSN节点获取路况视频,适用于复杂多变的地理环境,但视频图像数据量大,而WMSN节点资源受限.基于压缩感知技术和传统视频编码理论,提出了一种新的WMSN视频编解码方法.在编码端,依据视频图像的相关性对视频流进行分帧,再根据其特点分别进行压缩感知处理.在解码端,运用OMP算法重构关键帧和残差帧视频图像,进而恢复视频流.仿真实验表明,在保证视频图像重构质量的前提下,可以减少监控网中视频传输的数据量.
재교통감공중,이용WMSN절점획취로황시빈,괄용우복잡다변적지리배경,단시빈도상수거량대,이WMSN절점자원수한.기우압축감지기술화전통시빈편마이론,제출료일충신적WMSN시빈편해마방법.재편마단,의거시빈도상적상관성대시빈류진행분정,재근거기특점분별진행압축감지처리.재해마단,운용OMP산법중구관건정화잔차정시빈도상,진이회복시빈류.방진실험표명,재보증시빈도상중구질량적전제하,가이감소감공망중시빈전수적수거량.
In traffic surveillance,WMSN is applied to capture the video of road state,which can adjust to com-plex and variable environments. But WMSN sensor’s resource is limited and the number of video images yield-ed is tremendous yet. Based on compressed sensing and traditional video coding theory,this paper proposes a new method of WMSN video encoding/decoding. In encoding,the video stream is divided into key frames and non-key frames according to video image correlation,which are processed in line with their characteristics re-spectively. In decoding,video stream is further reconstructed after the key frames and residual video images are reconstructed through OMP algorithm. The simulation experiments demonstrate that the data volume of video transmission is largely reduced on the premise of keeping the reconstructed video image quality.