Research on Invasive Insect Image Recognition Based on Artificial Intelligence

The problem of invasive insect image accuracy identification is diffcult when the proportion of right samples is small and the quality of image is not very high. This article proposes a TRML (Transfer-Residual-network-Maximumentropy-Learning) model that can effectively identify invasive insects in a large amount of information. It compensates for the small proportion of invasive samples and low sample quality that often exist in actual intrusion problems. This article takes the Asian Hornet invasion of North America as an experiment, it show that the accuracy of the model test reaches 94.25%, and it indicates that the accuracy of the sample can be significantly improved. It also provides a new solution for solving this kind of feedback identification problem.