Fish species identification and detection is the process of accurately identifying and classifying fish through various technical means. Fish species identification and detection in today's society have high error rates and errors. The YoLov5 algorithm in deep learning is a technology for efficient detection and identification of fish species. Compared with conventional detection algorithms, the YoLov5 algorithm has higher accuracy. This article uses the YoLov5 algorithm to establish a detection system, which greatly improves the accuracy of detection. Finally, it was concluded through experiments that the YoLov5 algorithm is simpler to operate and has a higher accuracy, which can reach 98.94%.