基于视觉感受野特性的自适应图像去噪算法

针对图像去噪中边缘细节信息丢失的问题，提出一种基于视觉感受野特性的图像去噪算法。该方法基于视觉神经电生理研究结果，模拟视觉初级视皮层自适应机制和感受野的响应特性来实现对图像的去噪。使用小尺度模板对噪声进行检测；根据噪声的大小采用ON/OFF感受野模板去噪处理，对图像进行亮度调整。实验结果表明，与现有的流行算法比较，该算法去噪效果较为有效，并能更好地保留图像纹理和边缘细节信息，在峰值信噪比和均方误差等客观定量评价指标上优于其他算法。

Adaptive Image Denoising Algorithm Based on Visual Receptive Fields Properties

In order to eliminate noise and preserve details in image, an algorithm of image denoising based on visual receptive fields properties is proposed. Based on neuron electrophysiology research result, an image denoising processing has been realized by simulating adaptive mechanism in primary visual cortex and response characteristics of visual receptive fields. First, noise is detected by a small scale template. Then, according to the size and the location of the noise, an ON/OFF receptive field model is applied to adaptively deal with the noise. Finally, the brightness of the image is adjusted. Compared with some current denoising methods, experimental results show that textures and edges information in images processed by this proposed algorithm are better preserved. It is superior to other methods in objective image quality indexes, such as peak signal to noise ratio (PSNR) and mean squared error (MSE). It can not only remove noise in image process, but has ability to strengthen image edge details as well.