基于云模型编码算法的图像消噪研究

Image Denoising Based on Cloud Model Code

为了提高图像消噪的质量,采用云模型编码算法。首先通过正态云发生器产生云滴;接着正云滴映射为图像中大于平均灰度的像素,负云滴映射为图像中小于平均灰度的像素,不同云滴的编码则表示不同的像素的灰度特征;然后根据编码规则将图像子块分为平滑块和非平滑块,平滑块区域保持其增强质量;最后给出了基于云模型编码算法的图像消噪模型和算法流程。实验仿真显示该文算法对图像消噪效果最好,能以较大概率找到全局最优解。

The presented cloud model algorithm is used to improve the quality of denoised image. Firstly, the normal cloud generator is used to produce cloud droplet. Secondly, the positive cloud droplet is mapped to the pixels that grayscale is bigger than the average grayscale of image, the negative cloud droplet is mapped to the pixels that grayscale is smaller than the average grayscale of image, and different code is represented by gray feature of different pixel. Thirdly, the rule of code is divided into flat and non flat blocks, and the flat block area is maintained quality level. Finally, the processing of image mosaic is described. The experiment results show that this mechanism is efficient and can find the global optimal solution with high probability.