【作者】 盛明伟； 刘奕晖； 万磊；

【Author】 SHENG Ming-wei;LIU Yi-hui;WAN Lei;Science and Technology on Underwater Vehicle Laboratory,Harbin Engineering University;

【机构】 哈尔滨工程大学水下机器人技术重点实验室；

【摘要】 海洋环境噪声严重影响水下光学成像效果,水下噪声造成图像纹理不清,边缘模糊,目标和背景灰度分布不均,难以实现目标与背景的分离和后续目标识别等处理。针对以上问题,文章提出两种水下噪声图像改进增强滤波方法。首先,针对图像中水下噪声的生成特点,考虑传统模糊增强算法灰度损失、计算量大的缺点,从隶属度函数、快速性两方面对传统模糊增强算法进行改进;然后提出一种基于区域饱和度控制的边缘特征滤波方法,以减小水下噪声对后续直线特征提取的影响。结果表明所提方法可以有效滤除图像中的水下噪声。更多还原

【Abstract】 Ambient sea noise seriously affects the performance of underwater optical imaging,and the underwater noise can cause image texture is not clear,edge blurred,the gray distribution of the target and background not uniform,so it is difficult to realize the separation of target and background and the following target identification.In view of the above problems,two improved underwater noise image enhancement filtering methods were proposed.Firstly,according to the formation characteristics of underwater image noise,considering the traditional fuzzy enhancement algorithm of gray scale loss,and the disadvantage of large amount of calculation,the traditional fuzzy enhancement algorithm is improved from two aspects of membership function and rapidity.And then,a method of edge feature filtering based on region saturation control was proposed for reducing the influence of underwater noise on the subsequent line feature extraction.The results show that the proposed method can effectively filter underwater noise in the underwater images.更多还原