和2种类型的纸张步态识别算法

并本文对两大类步态识别算法进行了深入研究，其主要内容集中于步态特征的提取和分类器的设计两部分。 在特征提取阶段采用多种不同的方法提取多种不同的步态特征，其中效果较好的是一种基于模型的特征提取方法。本文使用关键点和肢体角度构建人体的骨骼化模型， 并对模型的各项参数提取做了改进，从人体的骨骼化模型中提取人体的静态参数(如身高、步幅等)以及动态参数(如运动过程中关键点的位置、运动轨迹、肢体角度、-and two types of paper gait recognition algorithm for in-depth study Its main contents concentrated in the gait characteristics of the extraction and classification of the design of two parts. Feature extraction stage in the use of different methods of extracting different gait characteristics, the effect of which is a better model-based feature extraction methods. Using the key points and limbs angle Construction human skeleton model and the model parameter extraction made improvements, from the human skeleton model of the human body from the static parameters (such as height, Stride) and dynamic parameters (such as the process of movement key points, trajectory, physical perspective,