摘要
由于水在水中传播所形成的散射效应和图像传感器的成像特性会造成水下图像含有较多的噪声。因此图像去噪是处理水下图像的重要步骤。经分析,水下图像的噪声是来自图像传感器所产生的高斯噪声。为了消除水下图像噪声和增强图像细节,提出了改进小波软硬折衷算法。引入收缩因子对Donoho阈值进行改进,使小波阈值更加符合水下图像去噪的需要。之后,综合软阈值函数和硬阈值函数的特性,对软硬阈值折衷函数进行了改进,使软硬折衷函数拥有更好的数学特性。同时,该函数结构简单,计算量小,非常适合水下图像去噪对处理速度的要求。最后通过对比其他算法的均方误差、峰值信噪比和信息熵,这些数据显示该改进算法能有效地消除噪声,增强细节。
The quality of underwater image is poor due to the properties of camera and light transmission in water. Image denoising is an important step in underwater image processing. The kind of noise is Gaussian noise for underwater by camera through analyzing. In order to eliminate the underwater image noise and enhance the image details, an improved soft-hard threshold algorithm is proposed. First the Donoho wavelet threshold is improved by introduction of shrinkage factor that is more suitable for underwater image denoising. Then the soft-hard threshold function is modified that combines the characteristics of soft threshold and hard threshold function, which makes it with better mathematical properties, simple structure and small amount of calculation that is very suitable for processing speed require- ments of underwater image denoising. Finally,compared with other algorithms on mean square error,peak signal to noise ratio and entropy, it is showed that it can successfully reduce the noise and increase the details.
出处
《计算机技术与发展》
2017年第11期150-153,158,共5页
Computer Technology and Development
基金
国家自然科学基金资助项目(61533002)
关键词
水下图像处理
图像去噪
小波软硬阈值折衷法
小波系数
underwater image processing
image denoising
wavelet soft-hard threshold
wavelet coefficient
