Recognition of Common Pests in Agriculture and Forestry Based on Convolutional Neural Networks

In order to achieve rapid recognition of the common pests in agriculture and forestry a new method based on deep convolution neural network was proposed. Firstly, the pest dataset is built by the search engines and manual photographed with image processing algorithm. Then an 11 layers VGG-A neural network is used to recognize common pests in agriculture and forestry. In addition, the structure of the network is simplified by optimized parameters such as the number of convolution kernels and the fully convolutional layer width. Dropout is utilized to reduce the problem of overfitting. The experimental results demonstrate that the accuracy of the optimized model reaches 99.02%, which is 4.04 percentage points higher than the network structure before optimization.