Research on Video Object Segmentation Algorithm

被引：0

作者：

Bo, Guan ^{[1
]}

机构：

[1] Baicheng Normal Coll, Baicheng 137000, Peoples R China

来源：

PROCEEDINGS OF THE 2015 4TH NATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS AND COMPUTER ENGINEERING ( NCEECE 2015) | 2016年 / 47卷

关键词：

video surveillance system; object segmentation; decoding;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

The video surveillance system is a more direct monitoring means, widely used in industrial production, transportation, electricity, office buildings and residential areas. Plenty of remote monitoring systems adopt the mode to collect and compress data through local end and remote monitoring center to make decompression processing. This paper analyzed the urban traffic surveillance video streaming, and it regarded vehicles as search targets, which have characteristics of rapid movement, therefore a particular object segmentation method was put forward. The algorithm took advantage of object motion vector in the compressed domain, as well as proposed the new regional connectivity labeling algorithms.

引用

页码：1399 / 1402

页数：4