Rock-Paper-Scissors Recognition System Based on YOLOv8

本文旨在研究并开发基于YOLOv8算法的剪刀石头布手势识别系统，推动人机交互技术发展，提升计算机视觉系统能力。研究采用YOLOv8算法，结合PySide6开发工具构建系统。通过对模型选择、训练、预测以及用户界面设计等过程的详细阐述，完成整个系统的搭建。在模型训练部分，利用PyTorch框架和Ultralytics YOLO库，加载预训练模型并进行针对性训练；模型预测时，导入OpenCV库处理图像，使用预训练模型实现手势检测。实验结果表明，YOLOv8在识别精度、速度以及复杂环境下的鲁棒性方面均优于YOLO系列早期版本。通过用户友好的界面，实现了直观便捷的手势识别。研究证明了YOLOv8在手势识别任务中的优越性，为其在虚拟现实、增强现实、智能家居控制等领域的应用提供了新的可能。This paper aims to research and develop a rock-paper-scissors gesture recognition system based on the YOLOv8 algorithm, aiming to promote the development of human-computer interaction technology and enhance the capabilities of computer vision systems. The research uses the YOLOv8 algorithm in combination with the PySide6 development tool to construct the system. The entire system is built through a detailed elaboration of processes such as model selection, training, prediction, and user interface design. In the model training section, the PyTorch framework and the Ultralytics YOLO library are utilized. A pre-trained model is loaded and further trained for the specific task of rock-paper-scissors recognition. When performing model prediction, the OpenCV library is imported to process images, and the pre-trained model is used to achieve gesture detection. The Experimental results show that YOLOv8 outperforms earlier versions of the YOLO series in terms of recognition accuracy, speed, and robustness in complex environments. The system, with its user-friendly interface developed by PySide6, enables intuitive and convenient gesture recognition. The study demonstrates the superiority of YOLOv8 in gesture recognition tasks, providing new possibilities for its applications in fields such as virtual reality, augmented reality, and smart home control.