摘要: 针对人体部位识别问题,提出一种基于深度图像学习的人体部位识别系统。构建深度图样本库,包括训练集和测试集,提取训练样本中的局域梯度特征,利用随机森林学习得到分类器,并对图像进行单点分类,计算人体各关节点。实验结果表明,该系统能快速准确地识别人体的不同部位。
关键词:
人体部位识别,
深度图像,
随机森林,
监督学习,
局域梯度特征
Abstract: Aiming at human body recognition problem, this paper proposes a human body part recognition system, which is based on depth image learning. It constructs depth image sample library, including training set and testing set, extracts local gradient feature from training samples, uses random forest to learn the classifier for separating each single points going through the image and computes each joint point of human body. Experimental result shows that the system can recognize different human body parts fast and accurately.
Key words:
human body part recognition,
depth image,
random forest,
supervised learning,
local gradient feature
中图分类号:
林鹏, 张超, 李竹良, 赵宇明. 基于深度图像学习的人体部位识别[J]. 计算机工程, 2012, 38(16): 185-188.
LIN Feng, ZHANG Chao, LI Zhu-Liang, DIAO Yu-Meng. Human Body Part Recognition Based on Depth Image Learning[J]. Computer Engineering, 2012, 38(16): 185-188.