基于Retinex灰度增强和颜色信息的时空上下文跟踪算法  

Spatio-temporal Context Tracking Algorithm Based on Retinex-enhanced Gray Information and Color Information

作　　者：张红颖 胡文博 Zhang Hongying;Hu Wenbo(College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300)

机构地区：[1]中国民航大学电子信息与自动化学院,天津300300

出　　处：《计算机辅助设计与图形学学报》2017年第12期2323-2329,共7页Journal of Computer-Aided Design & Computer Graphics

基　　金：天津市自然科学基金(12JCQNJC00600);中央高校基本科研业务费专项资金(3122015C016)

摘　　要：针对光照变化下图像局部或整体灰度剧烈变化而容易导致目标跟踪失败的问题,提出一种结合颜色信息和Retinex灰度增强的改进时空上下文算法.首先比较单尺度Retinex算法和多尺度Retinex算法,确定使用多尺度Retinex算法对图像进行灰度增强以减弱光照变化对图像灰度的影响;然后比较各种视觉模型的颜色特征,确定引入基于色相信息的目标模型,并将该模型与多尺度Retinex灰度增强模型相结合作为跟踪的目标模型.实验结果表明,文中算法比原算法在跟踪成功率上有较大提升,在Shaking场景下跟踪成功率为95%,比传统的时空上下文跟踪算法的跟踪成功率提高约24%;相比其他主流算法,该算法在平均跟踪成功率和跟踪精度上的表现也更高、更可靠.Illumination variation may result in extreme changes of local or overall image intensity,thus leads to tracking failure.To solve this problem,a spatial-temporal context(STC)tracking algorithm combing with color information and Retinex-enhanced gray information is proposed.Firstly,by comparing and analyzing the single-scale Retinex algorithm and multi-scale Retinex algorithm,the multi-scale Retinex algorithm is used to enhance image gray level to reduce illumination change influence on image gray level.Then,on this basis,by comparing the color features of various visual models,target model is adopted by hue information.Target tracking model is combined with model enhancement and multi-scale Retinex gray enhancement model.Finally,experimental results show that the proposed algorithm has a higher success than original algorithm.For example,algorithm success rate is about95%in the Shaking scenario,which is about24%higher success rate than STC.In comparison with other algorithms,it has better performance and higher tracking accuracy.

关 键 词：机器视觉 目标跟踪 时空上下文 光照变化 RETINEX 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]