Improved YOLOv8 algorithm for vehicle image target detection based on learning rate optimisation strategy

Vehicle image target detection is of great significance in the field of computer vision, which is mainly applied in the fields of traffic safety, intelligent traffic management and automatic driving. This study adopts a learning rate decay strategy for the YOLOv8 algorithm, which achieves the purpose of dynamically adjusting the learning rate. Through the analysis of model confusion matrix, F1 confidence, precision confidence, accuracy, recall, and recall confidence curve, the results show that the YOLOv8 algorithm adjusted by the learning rate attenuation strategy is able to efficiently and accurately detect and identify the car target in the image, and all the evaluation indexes have reached the optimal level. Comparison of the target detection results with the labels in the test set verifies that the optimised learning rate YOLOv8 algorithm achieves 100% prediction accuracy, successfully identifying and labelling the car target in the image. In conclusion, the learning rate optimised YOLOv8 algorithm performs well in vehicle image target detection and provides an effective solution for efficient and accurate car recognition.