摘要
针对彩色图像提出了一种基于水平集方法和神经网络的人体睡姿识别方法,由目标分割,BP神经网络识别两部分组成。首先利用水平集方法分割出图像中人体所在区域,然后使用BP神经网络方法训练并识别人体睡姿,其中睡姿包括仰睡、俯睡、左侧睡和右侧睡四种。实验结果表明该方法能够适应较复杂背景的人体睡姿识别,既具有较强的鲁棒性,又能够达到实时检测的目的。
In this paper,a sleep posture recognition algorithm for color images based on level set method and neural network,which is composed of target segmentation and BP neural network verifying is presented.First,level set method is used for segmenting the body region,and then the BP neural network methods are used for recognizing sleep postures which is include supine,prone,left side and right side.Experimental results show that the proposed method can detect sleep postures from complex background,which not only have higher robustness,but also in real time.
出处
《工业控制计算机》
2013年第5期88-90,共3页
Industrial Control Computer
关键词
睡姿识别
水平集方法
BP神经网络
sleep posture recognition
level set method
BP neural network
