摘要
小波阈值去噪作为图像降噪领域的一项重要技术一直受到广泛应用。在贝叶斯阈值的基础上提出一种改进的贝叶斯阈值去噪算法,该阈值是在贝叶斯框架中得出的,在小波系数上使用的优先级是在图像处理应用中广泛使用的广义高斯分布(Generalized Gaussian Distribution,GGD)。该阈值算法适用于每个子带,取决于数据驱动自适应参数估计,通过判断阈值周围的小波系数是否含有噪声的模糊性,从而对该模糊区域通过自适应算法确定小波系数的保留程度。实验结果表明,该方法比原方法在主观视觉效果上得到了明显的改善,较好地保持图像边缘细节,并且均方误差(Mean Square Error,MSE)较其他阈值算法有所减少,信噪比(Signal-Noise Ratio,SNR)较其他阈值算法有所提升。
As an important technology in the field of image denoising,wavelet threshold denoising has been widely used.This paper proposes an improved Bayesian threshold denoising algorithm based on Bayesian threshold.The threshold is obtained in Bayesian framework,and the priority used in wavelet coefficients is the Generalized Gaussian Distribution(GGD)widely used in image processing applications.The threshold algorithm is suitable for each subband,which depends on the data-driven adaptive parameter estimation.Using the fuzziness of noise in the wavelet coefficients around the threshold,the retention degree of wavelet coefficients is determined by the adaptive algorithm for the fuzzy region.Experimental results show that compared with the original method,this method has significantly improved the subjective visual effect,better maintained the image edge details,reduced Mean Square Error(MSE)and improved Signal-Noise Ratio(SNR)compared with other threshold algorithms.
作者
余传本
刘增力
YU Chuanben;LIU Zengli(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China)
出处
《电视技术》
2021年第10期106-111,115,共7页
Video Engineering
基金
国家自然科学基金(No.61271007)
关键词
自适应方法
贝叶斯
小波阈值处理
图像降噪
adaptive method
Bayes
wavelet threshold processing
image noise reduction
