线性混合小波基图像去噪研究

Study of Image Denoising Based on Linear Mixed Wavelet Base

基于线性混合小波基的阈值去噪是一种有效的图像去噪方法,它将多个不同特性的正交小波基进行线性混合构成一个新的小波基,该混合基兼有多个小波基的特性,能较好地表示复杂图像信号,从而有效地提高了去噪效果。该方法中混合系数的选择十分重要,它对混合基的去噪效果有重要的影响。本文首先介绍了线性混合小波基图像去噪的原理,然后在总结实验结果的基础上,提出了一种两个小波基混合的混合系数选择方法,并以其为基础实现两个以上的多个小波基混合。大量的实验表明,用该方法选择的混合系数,混合基的去噪效果良好。

Image denoising based on linear mixed wavelet base is an efficient method. Since the wavelet base is onstructed by linear mixing with multi different orthogonal wavelet bases, the base has property of multi wavelet baes, so it can better express complex image signal and enhance the image denoising effect. The mixed coefficients se- ection is very important because it imposes great effect to the noise reduction of mixed wavelet base. At first, the asic theory of image denoising based on lineer mixed wavelet base is introduced. Then, based on the experiment esults, a mixed coefficients selection method of two wavelet bases is proposed, and more than two of wavelet bases nixing are realized. Experiment results show that this selection method can obtain better image denoising effect.