基于PCNN捕获-抑制机制的图像分类器设计  

Classifier design based on the mechanism of capture-suppress of PCNN

作　　者：于文勇[1] 康晓东[1] 洪睿[1] 

机构地区：[1]天津医科大学医学影像学院,天津300070

出　　处：《中国科学：技术科学》2016年第3期243-247,共5页Scientia Sinica(Technologica)

摘　　要：针对医学图像识别的重要性,本文提出了一种结合彩色图像特征场和基于脉冲耦合神经网络捕获-抑制机制的图像分类器设计方法.首先,构造符合人类视觉的彩色图像特征场;然后,利用脉冲耦合神经网络设计分类器实现对图像的识别.最后,应用于JAFFE人脸表情数据库和中医舌象的分类识别中,并与ADABOOST方法和SVM方法进行比较.实验结果表明,本文方法可以获得更好的医学图像识别效果.To the significance of medical image recognition, a method of image classification and recognition capturing and suppress mechanism is proposed, which is combining with the feature data field of color image. Firstly, the color image feature field according to human vision is constructed, then, the recognition of the image is realized by using the pulse coupled neural network. Finally, it is applied to classification and recognition of JAFFE facial expression database and traditional Chinese medicine of tongue, and compared with ADABOOST method and SVM method. Experiment shows that this method can obtain better medical image recognition results.

关 键 词：脉冲耦合神经网络 捕获 抑制 特征场 图像识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]