摘要
针对水下获取的图像存在边缘细节模糊、噪声大、对比度低、视觉效果差的问题,提出一种基于小波变换的水下鱼群图像增强算法。先利用自适应直方图均衡提高水下图像的整体对比度;再利用小波变换对图像进行分解,对低频部分分别使用基于改进双边滤波的单尺度Retinex和直方图均衡进行处理,然后将处理的二者进行线性加权融合,再进行小波逆变换重构,对重构的图像进行增益计算,得到增强图像;利用小波变换对水平方向和垂直方向的高频图像进行边缘检测再重构,得到另一个增强图像;最后对两幅增强图像进行加权融合得到最终的增强图像。与目前已知的经典的增强方法相比,结果表明该算法的清晰度、峰值信噪比较高,能有效提升图像的对比度,丰富图像的细节信息,而且视觉效果也有明显改善。
Aiming at the problems of blurred edge details,high noise,low contrast and poor visual effect in underwater images,an image enhancement algorithm based on wavelet transform is proposed.Firstly,adaptive histogram equalization is used to improve the overall contrast of underwater images.The image is decomposed by wavelet transform,and the low-frequency parts are processed by single-scale Retinex and histogram equalization respectively based on improved bilateral filtering.Then the two processed parts are combined with linear weighted fusion,and the reconstructed image is reconstructed by inverse wavelet transform.The gain of the reconstructed image is calculated to obtain the enhanced image.The edge detection and reconstruction of the horizontal and vertical high frequency images are carried out by wavelet transform,and another enhanced image is obtained.Finally,the two enhanced images are weighted and fused to obtain the final enhanced image.Compared with the known classic enhancement method,the proposed algorithm has higher definition and peak signal noise ratio,which can effectively enhance image contrast,enrich the image details,and improve the visual effect significantly.
作者
何笑
吐尔洪江·阿布都克力木
贺欢
HE Xiao;Turgunjan ABDUKIRIM;HE Huan(School of Mathematical Sciences,Xinjiang Normal University,Urumqi 830017,China)
出处
《计算机技术与发展》
2020年第9期77-81,共5页
Computer Technology and Development
基金
国家自然科学基金(11261061,61362039,10661010)
新疆维吾尔自治区自然科学基金(200721104)
新疆师范大学数学教学资源开发重点实验室招标课题(XJNUSY082017B03)。
关键词
小波变换
双边滤波
直方图均衡
边缘检测
图像融合
wavelet transform
bilateral filtering
histogram equalization
edge detection
image fusion
