基于脉冲耦合神经网络的乳鼠心肌细胞图像增强  

Enhancement of the Image of Cardiomyocytes in Neonatal Rat Based on Pulse Coupled Neural Network

作　　者：陈洁[1,2] 雷萌 CHEN Jie;LEI Meng(School of Information Engineering,Xizang Minzu University,Shaanxi Xianyang 712082,China;Xizang Key Laboratory of Optical Information Processing and Visualization Technology,Shaanxi Xianyang 712082,China)

机构地区：[1]西藏民族大学信息工程学院,陕西咸阳712082 [2]西藏自治区光信息处理与可视化技术重点实验室,陕西咸阳712082

出　　处：《软件》2018年第2期41-43,共3页Software

基　　金：西藏自治区科技厅项目(2015ZR-13-24);西藏民族大学青年学人培育计划资助项目(14my QP04)

摘　　要：本文研究了脉冲耦合神经网络(PCNN)的特性及人眼视觉特征,模拟了一种用图像灰度对数替代原本PCNN输出的图像增强方法,并对乳鼠心肌细胞图像进行了增强处理。通过实验,对比图像形态学重建增强效果,本文采用的方法图像目标纹理更清晰、灰度更均匀,细节轮廓保持更完整,该图像增强方法适合处理前背景对比较明显,细节性要求较高的图像,尤其是整体灰度值较暗的情况。In this paper,the characteristics of Pulse Coupled Neural Network（PCNN） and visual mechanism of human eyes are studied.Simulated a image enhancement method based on PCNN by replacing the original output with pixel grayscale logarithm,and the images of neonatal rat cardiomyocytes are enhanced.By contrasting the result of morphological reconstruction method and PCNN method in this paper,it is proved that the image texture is clearer and more uniform in grayscale and more complete in outline by the image enhancement method in this paper.The image enhancement method in this paper is suitable for the images which have formerly contrasting background,and need more requirements of details,in particular,for the situation that the image has overall dark grayscale.

关 键 词：图像增强 脉冲耦合神经网络 PCNN 乳鼠心肌细胞 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]