Tea leaf disease recognition algorithm based on improved YOLOv8s

In order to solve the problem of poor feature extraction ability of tea leaf disease in complex natural environment, a tea leaf disease recognition model based on improved YOLOv8s was constructed. CBAM attention mechanism was added to the backbone network of YOLOv8s network model, and Wise-IoU loss function was used to replace CIoU loss function, so as to increase the nonlinearity of input data, better fit the correlation of tea diseases under complex natural environment background, strengthen the network's attention to small diseases and adhesion diseases, and prevent low quality samples from producing large harmful gradients. Experiments on open source data set show that the new model is superior to the traditional target detection algorithm, mAP50 is improved by 1.4%, the number of parameters is reduced by 18.5%, and the detection time is shortened by 4ms, which meets the detection requirements of high efficiency and high precision.