摘要： 针对数字文献图像去噪问题，提出一种基于生物视觉机理的图像去噪算法。模拟初级视皮层简单细胞感受野的响应特性，通过提取数字文献图像的特征，获得数字文献图像的基函数。由动物视觉感知系统的稀疏性，计算神经元对含噪声图像的响应，结合稀疏编码收缩法对响应系数进行收缩，通过响应强烈的神经元重构图像。实验结果表明，与传统的去噪方法相比，该算法能更好地去除数字图像中的高斯噪声，并保留图像细节信息。

关键词: 视觉机理, 图像去噪, 数字文献图像, 独立分量分析, 基函数

Abstract: A digital literature image denoising algorithm based on biological visual mechanism is proposed. Simulating the receptive field response characteristics of simple cells in primary visual cortex, the basis functions of digital literature image are obtained by image feature extraction. The neuron response of noisy images is calculated according to the sparsity of animals’ vision perception system. Sparse coding shrinkage is used to shrinkage the response coefficient, and the neurons that respond stronger are used to reconstruct the image. Compared with the traditional methods of image denoising, the experimental results demonstrate that this method can reduce Gaussian noise more effectively, and has a better effect in preserving image detail information.

Key words: visual mechanism, image denoising, digital literature image, Independent Component Analysis(ICA), basis function

中图分类号: 

• TP391.4