基于手势识别的幼儿游戏机器系统设计  

Design of children’s game machine system based on gesture recognition

作　　者：边宝丽[1] BIAN Baoli(Xianyang Vocational Technical College,Xianyang Shaanxi 712000,China)

机构地区：[1]咸阳职业技术学院,陕西咸阳712000

出　　处：《自动化与仪器仪表》2024年第3期171-174,179,共5页Automation & Instrumentation

基　　金：陕西省职业技术教育学会2023年度教育教学改革研究课题《手工课程思政建设与实践》(2023SZX195)。

摘　　要：手势识别技术是人机交互中的重要板块。现阶段的手势识别技术存在着关键点定位不准确,手势识别效果差等问题,在可使用领域中很难有所突破。基于此,研究提出了使用改进BP神经网络的方式,改善图像关键点识别。此外,通过基于ECPM的手势识别模块,和LSTM的手势预模块,共同构建了基于手势识别的幼儿游戏机器。对算法和游戏机器的性能进行实验验证,结果表明,该系统的手势平均识别率为97.89%,平均预测准确率为75.42%。研究为手势识别的改进提供了参考思路,同时为幼儿游戏机器设计提供了方案。Gesture recognition technology is an important part of human-computer interaction.At the present stage,gesture recognition technology has some problems such as inaccurate location of key points and poor effect of gesture recognition,so it is difficult to make a breakthrough in the usable field.Based on this,the method of using improved BP neural network is proposed to improve image key point recognition.In addition,through the gesture recognition module based on ECPM and the gesture pre-module of LSTM,we jointly build a children’s game machine based on gesture recognition.The performance of the algorithm and the game machine is verified by experiments.The results show that the average recognition rate of gesture is 97.89%and the average prediction accuracy is 75.42%.This study provides a reference for the improvement of gesture recognition and a scheme for the design of children’s game machines.

关 键 词：手势识别 BP神经网络 ECPM LSTM 手势预测 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]