Feature selection of stored-grain insects based on artificial fish swarm algorithm
1.Key Laboratory of Modern Agricultural Equipment and Technology, Ministry of Education ＆ Jiangsu Province, Jiangsu University, Zhenjiang, Jiangsu 212013, China; 2.Institute of Electric Power, North China Institute of Water Conservancy and Hydroelectric Power, Zhengzhou, Henan 450011, China)
Abstract：The feature selection is the core issue in the image recognition of the storedgrain insects. The feature selection was firstly proposed based on the artificial fish swarm algorithm, and the optimization and realization were given based on binary code. The recognition accuracy of the Vfold crossvalidation training model was taken as the evaluation principle of the feature subset. The artificial fish swarm algorithm was applied to the feature selection of the storedgrain insects. The algorithm selected seven features composing the optimal feature space from 17 morphological features, such as area and perimeter. The ninety image samples of the storedgrain insects in graindepot were automatically recognized by the support vector machine classifier so that the parameters were optimized, and the correct identification ratio was over 955％. The artificial fish swarm algorithm was compared with the principal component analysis, the genetic algorithm and the original feature.The experimental results show that the artificial fish swarm algorithm greatly reduces the feature dimensions and improves the recognition accuracy. It is practical and feasible.
2010, Vol. 31 
