基于小波变换的水下图像去噪方法  

Underwater image denoising method based on wavelet transform

作　　者：陈振娅 刘增力[1] CHEN Zhenya;LIU Zengli(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500

出　　处：《现代电子技术》2023年第23期43-47,共5页Modern Electronics Technique

基　　金：国家自然科学基金项目(61271007)。

摘　　要：水下图像在拍摄、传输中会产生严重的噪声影响,这些噪声不仅影响图像质量,还影响图像后续的目标检测和定位精度。文中改进了一种基于小波变换的水下图像去噪算法,该算法首先考虑信号和噪声在不同小波变换尺度下的不同传播特性,采用一种随分解尺度变化的自适应阈值估计方法;其次采用自适应图像特征的收缩函数对每个小波系数局部估计小波系数的能量进行分类。实验结果表明,改进的小波去噪方法能够有效地减弱噪声信号,使去噪后的水下图像更有利于后续的目标检测和定位。Underwater images can produce serious noise effects in shooting and transmission,and these noises affect not only the image quality,but also the subsequent object detection and localization accuracy of the images.In this paper,an underwater image denoising algorithm based on wavelet transform is improved.In this algorithm,the different propagation characteristics of signal and noise under different wavelet transform scales are taken into account,so an adaptive threshold estimation method that changes with the decomposition scale is adopted.The shrinkage function which is of adaptive image features is adopted to estimate the energy of wavelet coefficients locally in each wavelet coefficient for classification.The experimental results show that the improved wavelet denoising method can attenuate the noise signal effectively and make the denoised underwater images more favorable for subsequent object detection and localization.

关 键 词：小波变换 水下图像 图像去噪 尺度分解 自适应阈值 图像特征 小波收缩函数 局部能量 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]