论文部分内容阅读
为了提高红外线图像的分类效果,针对当前红外线图像分类过程中的特征权值确定和分类器选择问题,提出一种基于数据挖掘的红外线图像分类方法。首先收集红外线图像,并提取红外线图像的Gabor滤波器特征,然后采用Relief算法确定特征的权值,并采用回声状态网络构建红外线图像分类器,最后采用仿真实验测试其有效性。实验结果表明,本文方法可以准确描述每一个特征对红外线图像分类的贡献,提高了红外线图像的分类正确率,分类速度可以满足红外线图像应用中的实时性要求。
In order to improve the classification effect of infrared images, aiming at the feature weight determination and classifier selection in the current infrared image classification, an infrared image classification method based on data mining is proposed. Firstly, the infrared images are collected and the features of the Gabor filters are extracted. Then the Relief algorithm is used to determine the weights of the features. The infrared image classifier is constructed by using the echo state network. Finally, the effectiveness of the algorithm is tested by simulation experiments. Experimental results show that the proposed method can accurately describe the contribution of each feature to infrared image classification and improve the classification accuracy of infrared images, and the classification speed can meet the real-time requirements of infrared image applications.