Abstract: In order to improve the recognition accuracy among the kinds of aphids，the recognition method among kinds of aphids based on Dual-Tree Complex Wavelet Transform（DT-CWT） and Support Vector Machine（Libsvm） is proposed.Firstly the image is pretreated；secondly the features of the aphids’ image examples about three kinds of crops are extracted by DT-CWT in order to get the training parameters of training model；finally the training model can recognize aphids among the three kinds of crops.By contrasting to Gabor wavelet transform and the traditional extracting texture’s methods based on Gray-Level Co-Occurrence Matrix，the experiment result shows that the method has a certain practicality and feasibility and provides basic for aphids’ recognition between the same kinds of aphids.

Key words: Dual-Tree Complex Wavelet Transform（DT-CWT）, Gabor wavelet, feature extraction, gray-scale co-occurrence matrix, Libsvm

摘要： 为提高蚜虫种间识别的准确率，提出了一种基于双树复小波变换（DT-CWT）和支持向量机（Libsvm）的蚜虫图像种间识别方法。该方法首先对图像进行预处理；其次应用DT-CWT对选定的三种农作物蚜虫图像样本进行特征提取，以确定Libsvm训练模型中的训练参数；最后用Libsvm分类器对蚜虫图像进行种间识别。实验对比了Gabor小波变换和传统上基于灰度共生矩阵的特征提取算法，结果表明该方法具有更高的准确率和较低的时间复杂度，为蚜虫的种内识别提供条件。

关键词: 双树复小波变换, Gabor小波, 特征提取, 灰度共生矩阵, Libsvm