【作者】 常青； 刘怀贤； 王立；

【Author】 CHANG Qing,LIU Huai-xian,WANG Li(School of Information Science and Engineering,East China University of Science and Technology,Shanghai 200237,China)

【机构】 华东理工大学信息科学与工程学院；

【摘要】 视网膜是人眼视觉系统的第一个信息处理单元,它不仅完成光学信息向生物电信号的转变,而且对视觉信号进行必要的初级处理。侧抑制是视网膜信息处理机制的一个重要概念,在生物学研究成果的基础上提出的侧抑制网络模型在图像处理中具有检测图像边缘、突出边框、增强反差的作用。针对侧抑制网络在图像处理中增强图像边缘的同时,也对噪声敏感的缺陷,提出了基于差异判别的改进侧抑制网络实现方法;针对反馈型侧抑制网络计算量大、实现困难的问题,提出了采用多级串联非循环侧抑制简化循环侧抑制网络实现方法。通过对上述方法的仿真分析和验证,试验结果表明:所提出的改进的侧抑制网络能够在对图像实现增强的同时,具有较好的抗噪性能。更多还原

【Abstract】 The lateral inhibition network is one of the most important conceptions of the retinal image processing mechanism.The lateral inhibition network has the function of standing out the frame and enhancing the contrast,but its anti-noise capability is poor.When using it,the noisy pixel dot is also enhanced,which is disadvantageous to the following process.A modified lateral inhibition network was proposed by applying the multilevel acyclic lateral inhibition network to simplify the feedback network.The modified method has a certain anti-noise capability by introducing the difference criterion to the lateral inhibition.The promoted network has both enhanced and anti-noise capability.By theoretical analysis and experimentation emulating,the noise can be inhibited to a certain extent with the new method.更多还原