python从xml读取数据

最新推荐文章于 2024-09-02 07:14:24 发布

凉了凉了

最新推荐文章于 2024-09-02 07:14:24 发布

阅读量1k

点赞数

文章标签： xml python

本文链接：https://blog.csdn.net/qq_36468290/article/details/78472647

版权

该博客介绍了一个Python函数，用于从XML文件中读取G3D数据集的骨架数据。它遍历文件夹中的所有XML文件，对每个文件解析骨架跟踪状态和关节位置，特别是关注Spine、ShoulderCenter、ElbowLeft、WristLeft、HandLeft、ElbowRight、WristRight和HandRight等关节，存储24个关节的位置数据。

摘要由CSDN通过智能技术生成

from xml.etree import ElementTree as ET

def getSubDir(path):
    dirname=list()
    for file in os.listdir(path):
        dirname.append(file)
    return dirname

def get_xml_data(path):
    xmlFileNames = getSubDir(path)
    gesturePositionEndCount = 0

    for xmlFileName in xmlFileNames:
        if gesturePositionEndCount < int(str(xmlFileName).split("Skeleton ")[1].split(".xml")[0]):
            gesturePositionEndCount = int(str(xmlFileName).split("Skeleton ")[1].split(".xml")[0])
    gesture = np.zeros((gesturePositionEndCount+1, 24))
    gesturePositionStartCount = gesturePositionEndCount

    for xmlFileName in xmlFileNames:
        xmlPath = path + xmlFileName
        dom=ET.parse(xmlPath)
        TrackStatus=dom.findall('./Skeleton/TrackingState')
        isTracked = False

        for trackStatu in TrackStatus:
            if trackStatu.text=="Tracked":
                isTracked=True
        JointsPosition = np.zeros((24))
        if isTracked==True:
            gestureCount=int(str(xmlFileName).split("Skeleton ")[1].split(".xml")[0])
            if gesturePositionStartCount > gestureCount:
                gesturePositionStartCount = gestureCount
            Joints = dom.findall('./Skeleton/Joints/Joint/Position')
            JointType=dom.findall('./Skeleton/Joints/Joint/JointType'