改进阈值函数在图像去噪中的应用研究

针对传统阈值函数在图像去噪中存在的缺点和阈值选取的不确定性,为进一步研究小波去噪处理算法,提出了一种新的改进阈值函数的去噪算法,分别对Circuit、Eight、Road 3幅图像采用传统的硬阈值、软阈值、半软阈值方法和文中方法进行Matlab仿真实验。对比分析,文中方法既获得了较好的视觉效果和更高的峰值信噪比,又克服了软阈值信号失真和硬阈值信号不连续、振荡等缺点,且能够在消噪和保留原有信号的弱特征之间获得较好的平衡,明显地改进了传统硬、软阈值函数去噪算法存在的诸多不足,在实际应用中更为有效。

The research of improved threshold function on image denoising

Aiming at the disadvantages of traditional threshold function for image denoising and the uncertainty of the threshold selection, a new wavelet denoising algorithm with an improved threshold function is proposed, and is perfo- rmed together with the traditional hard threshold and soft threshold, semi-soft threshold method in Matlab simulation for Circuit, Eight, Road images, respectively. By contrast, this method obtaines a better visual effects and higher Peak Signal to Noise Ratio (PSNR), and overcome signal distortion in the soft threshold methodand signal discontinuity and oscillation in the hard threshold method, . Moreover it can strike a better balance between noise elimination and retaining the weak features of the original signal, and significantly addressed many problems in the traditional hard and soft threshold function denoising algorithms.This method is more effective in practical applications.