虚实融合监控视频压缩方法的分析与评价  

Analysis and Evaluation on the Virtual-Real Fusion Surveillance Video Compression Method

作　　者：解愉嘉[1,2,3] 刘学军[1,2,3] 

机构地区：[1]南京师范大学地理科学学院,江苏南京210023 [2]虚拟地理环境教育部重点实验室,江苏南京210023 [3]江苏省地理信息资源开发与利用协同创新中心,江苏南京210023

出　　处：《南京师大学报（自然科学版）》2016年第3期130-134,共5页Journal of Nanjing Normal University(Natural Science Edition)

基　　金：公安部科技强警基础工作专项资助项目(2014GABJC009);国家科技支撑计划基金项目(2012BAH35B02);江苏省优势学科建设项目(164320H116)

摘　　要：在阐述虚实融合视频压缩基本思想的基础上,对虚实融合视频压缩技术流程进行描述.针对虚实融合视频压缩方法提出评价因子,在此基础上构建压缩效果评价模型.实验结果表明,时间尺度因子对压缩效率的影响受运动前景目标出现情况的制约:对于没有前景运动目标出现的段落,单位时长的视频压缩效率高于累积时长;对于前景运动目标出现的段落,累积时长的视频压缩效率高于单位时长.压缩层级因子的变化对压缩效率的影响较显著:对象投影式相比于图像投影式,所需使用数据量可压缩至小于后者的1%;前景抽象式相较于前后景独立投影式与前景投影式,所需使用数据量可压缩至小于后者的1%.On the basis of describing the idea of the virtual-real smweillance video compression method, this paper describes the technique process of the compression method, proposes the relative evaluation factors, and constructs the evaluation model. The experimental results have shown that, the impact from the time scale factor is conditioned by the appearance of the prospect of moving target. When there is no moving target, the compression rate of unit-time length is higher than the compression rate of cumulative-time length ; when there are some moving targets, the compression rate of cumulative-time length is higher than the compression rate of unit-time length. The impact from the compression level factor is remarkable. Comparing to the compression rate of image projection fusion style, the compression rate of object projection style is less than 1%. Comparing to the compression rate of fore-ground independent and prospect projection styles, the compression rate of prospect abstraction projection style is less than 1%.

关 键 词：虚拟场景 虚实融合 监控视频 视频压缩 

分 类 号：P208[天文地球—地图制图学与地理信息工程]