基于Gabor滤波器组和特征分组神经网络目标识别

Object recognition of neural networks based on Gabor filter banks and feature decomposition

就模式识别中特征提取和分类决策两个阶段进行了研究 .在特征提取方面 ,提供了一种用于提取图像不同中心频率和方位特征的Gabor滤波器组的设计方法 ;在分类决策算法方面 ,提出了一种基于图像的Gabor滤波特征的特征分组神经网络分类算法 ,该算法将一个复杂的模式识别问题分解为几个简单问题来处理 ,降低了问题的复杂度 ,同时还取得了一定的抗有限带宽噪声的效果 .

A method for designing Gabor filter banks with various frequencies and orientations was proposed to perform feature extraction. And a feature decomposition modular neural network classifier was presented for object recognition based on the feature extraction by the Gabor filter banks. The method used by the classifier has the advantages of reducing the complexity of the recognition problems and the influence of band limited noise.