基于Retinex的旧视频序列增强算法  

Old Video Sequence Enhancement Algorithm Based on Retinex

作　　者：李建彬[1] 韩军[1] 邱君华[1] 

机构地区：[1]上海大学通信与信息工程学院,上海200072

出　　处：《电视技术》2011年第5期21-24,共4页Video Engineering

基　　金：浙江省重点科研工业项目(2007c21014)

摘　　要：为了解决早期的旧视频资料存在图像不清晰、图像动态范围低、似薄雾的偏色等问题,提出一种旧视频序列增强算法。该算法利用Retinex理论的图像增强获取图像暗区增强视频序列,利用伽马曲线调整获取亮区细节保持视频序列,通过加权合成最终增强的视频序列。实验证明所提方法是有效的。In order to solve the problems in old video sequence such as image unsharpness, low dynamic range, color cast like mist and so on, an old video sequence enhancement algorithm is proposed in this paper. Retinex enhancement theory is used to get dark area enhanced video sequence, and the Gamma adjustment is used to get bright area detail retained video sequence. The final enhanced video sequence is composed by weighting. Experimental results show that the proposed method is effective.

关 键 词：视频增强 RETINEX理论 高斯函数直方图变换 伽马校正 

分 类 号：TN919.8[电子电信—通信与信息系统]