The application of AdaBoost-neural network in storedproduct insect classification

The classification of stored product insects has been an important and difficult aspect of grain reserve in the world. The existing classification methods cannot acquire excellent performance. AdaBoost, an adaptive boosting algorithm, may improve the classification accuracy of any given classifier. In this paper AdaBoost is adopted to increase the performance of artificial neural network for stored product insect classification, and compared with standard neural network methods. Experiment results show that the new method is efficient and a significant improvement in classification accuracy is obtained.