水下电视图像预处理方法研究  

Study of Image Preprocessing Method of Underwater TV

作　　者：张法全[1,2,3] 郑承栋[1] 陈良益[1] 沈满德[1,2] 

机构地区：[1]中国科学院西安光学精密机械研究所 [2]中国科学院研究生院,北京100039 [3]郑州轻工业学院电气信息工程学院,河南郑州450002

出　　处：《电视技术》2008年第3期16-18,共3页Video Engineering

基　　金：国家自然科学基金项目(60337030);中国科学院知识创新工程国防创新重大项目

摘　　要：分析了水下图像特性,对比了迭代阈值法、减去固定背景法、减去当前背景法、数学形态学方法4种常用方法的预处理结果,指出了它们的不足,并提出了基于小波边缘检测的水下图像预处理算法。实际应用表明,新算法处理效果很好,能够满足水下智能监控系统的要求。The image characteristics of underwater TV are analyzed. The preprocessing results of four common methods are compared, including iterative threshold method, eliminating stable background method, eliminating present background method and mathematical morphology method. The disadvantages of these methods are pointed out. And then, the image preprocessing algorithm based on wavelet edge detection is proposed. The detecting results show that the new algorithm has good processing effect, and can well meet the requirements of underwater intelligent monitoring system.

关 键 词：水下电视图像 预处理 数学形态学 小波变换 

分 类 号：TN949.22[电子电信—信号与信息处理]