Abstract: The method of classifying cotton blind stinkbugs based on Gabor wavelet and color moments was developed for improving the forecasting and warning of blind stinkbug disaster under field conditions. In this method, Gabor wavelet and color moments were used to extract texture features and color features from cotton blind stinkbugs, in order to detect cotton blind stinkbug with principal component analysis and support vector machine. The results of feature extraction and optimization showed that cumulative contribution rate of first three principal components was 87.3%, and first three principal components obtained better clustering results. Cotton blind stinkbug classification results found combination of first three principal components and radial basis function support vector machine better than others. Training time and accuracy rate of the methods were 89 ms and 91%, respectively. The study results showed that the proposed classification?method could be used to?accurately classify?the five kind of blind?stinkbugs.