基于混合视觉皮层模型的红外图像去噪和分割

Infrared Image Denoising and Segmentation Based on Mixed Cortical Model

针对交叉视觉皮层模型(ICM)网络运行速度慢的问题,提出一种用于图像处理的混合视觉皮层模型(MCM)。MCM将人眼视觉现象中的韦伯律特性融入ICM,根据韦伯律特性设定神经元点火域值,采用金字塔分解式的运算方法扩大神经元的感受野,并利用ICM的脉冲同步发放性质,实现快速图像处理。基于MCM开发红外图像的去噪算法及红外图像分割算法,实验结果表明,与基于ICM的图像去噪和分割算法相比,基于MCM的算法具有良好的去噪和分割效果,运行速度较快。

Aiming at the problem that the Intersecting Cortical Model（ICM） works at a low speed,this paper proposes a Mixed Cortical Model（MCM） used in image processing.It combines Weber＇s Law trait of Human Visual System（HVS） and ICM,and in MCM the firing threshold of the nerve cell is set according to the Weber trait,while the pyramid decomposing algorithm is adopt to enlarge the nerve cell＇s sensitive field,the property of ICM to transmit synchronous pulse is also adopt in MCM,therefore,the novel model can process images rapidly.Algorithms based on MCM to do image denoising and image segmentation are proposed in the following section,and compares with the algorithm based on ICM.Experimental results show that the algorithm based on MCM yields some promising results and realizes rapid processing.