从手势识别系统框架模型、手势分割、手势建模与分析和手势识别等几个方向,系统地综述当前计算机视觉动态手势识别技术的研究现状,分析其存在的不足,提出了进一步研究的问题。结果表明:基于简易可穿戴设备的手势识别、基于深度视觉传感器的手势识别和多方法交叉融合的手势识别将是未来该领域的发展趋势。
According to classification of system model,hand segmentation,hand modeling and hand recognition,the paper surveyed these technologies with computer vision in last several years in detail.Existing problems are analyzed and issues of future researches are presented.In terms of analysis results,simple wearable device,depth vision sensor and integration of multi method will be developing trends in hand recognition in the future.