Classification of angiosperms by gray-level co-occurrence matrix and combination of feedforward neural network with particle swarm optimization

This study proposed an application of feedforward neural network (FNN) with particle swarm optimization(PSO) on angiosperms classification. We first collected petal images of three different angiosperm plants and each type contains 40 images. Second, we used gray-level co-occurrence matrix (GLCM) to extract texture features. Third, we used FNN as the classifier. Finally, we employed PSO to train the classifier. In the experiment, we utilized eight-fold cross validation techniques. The average sensitivity of our method is about 86%. This proposed method performs better than three genetic algorithm and simulated annealing.