用自适应谐振网络进行图像纹理的分类与识别

用自适应谐振网络进行图像纹理的分类与识别 TEXTURE IMAGE CLASSIFICATION AND RECOGNITION USING ADAPTIVE RESONANCE NETWORK 潘琼峰 1 first-author 廖孟扬 1 覃家美 1 武汉大学电子信息学学院 武汉大学电子信息学学院 College of Electronic Information,Wuhan University College of Electronic Information,Wuhan University 分析了自适应谐振（ＡＲＴ２）神经网络模型的模式分类能力，并利用该网络来进行图像纹理的分类和识别，对６类自然景物的纹理图片分类和识别的结果验证了方法的有效性，对心脏超声图片的分类也取得一些初步成果． We analysis the classification ability of ART2 （adaptive resonance theory）neural network and use the network to classify and recognize texture image. Our purpose is to find a efficient method of unsurpervised texture image classification and recognition.We get a satisfied result over six kind of scene texture image classification and recognition experiment,so this method is feasible. We also use this network to classify the cardiac ultrasonic image and get some elemental results. 人工神经网络 自适应谐振理论 纹理 分类与识别 非监督学习 artificial neural network adaptive resoncnce theory texture classification and recognition unsurpervised learning TP391.41 1995-03-01 2021-04-01 3