统计xml文件包含的标注信息 parse_dataset_annotation

统计xml文件包含的标注信息 parse_dataset_annotation

YOLO算法,统计标注的xml文件中包含的标记框信息

创建parse_dataset_annotation.py

import os
import sys
filedir = os.path.dirname(sys.argv[0])      #获取脚本所在目录
os.chdir(filedir)       #将脚本所在的目录设置为工作目录
wdir = os.getcwd()
print('当前工作目录:{}\n'.format(wdir))      #打印当前工作目录

from xml.dom.minidom import parse

def xml_parser( xml_file ):
    '''
    Parse an xml file and return the annotation info in the file

    :param xml_file: the xml file name to be parsed
    :return: file_name, width, height, objects.
        file_name, filename of the xml file (without extension)
        width, width of the annotated image
        height, height of the annotated image
        objects, annotated objects in the image
        object, (object_name, xmin, ymin, xmax, ymax)
            object_name, name of the annotated object
            xmin, ymin, xmax, ymax, coordinate of the bounding box of the object

    '''
    DOMTree = parse( xml_file )
    collection = DOMTree.documentElement #得到xml文件的根节点
    file_name_xml = collection.getElementsByTagName( 'filename' )[0]
    objects_xml = collection.getElementsByTagName( 'object' )
    size_xml = collection.getElementsByTagName( 'size' )

    file_name = file_name_xml.childNodes[0].data

    for size in size_xml:
        width = size.getElementsByTagName( 'width' )[0]
        height = size.getElementsByTagName( 'height' )[0]

        width = width.childNodes[0].data
        height = height.childNodes[0].data

    objects = []
    for object_xml in objects_xml:
        object_name = object_xml.getElementsByTagName( 'name' )[0]
        bdbox = object_xml.getElementsByTagName( 'bndbox' )[0]
        xmin = bdbox.getElementsByTagName( 'xmin' )[0]
        ymin = bdbox.getElementsByTagName( 'ymin' )[0]
        xmax = bdbox.getElementsByTagName( 'xmax' )[0]
        ymax = bdbox.getElementsByTagName( 'ymax' )[0]

        object = [ object_name.childNodes[0].data,
                   float(xmin.childNodes[0].data),
                   float(ymin.childNodes[0].data),
                   float(xmax.childNodes[0].data),
                   float(ymax.childNodes[0].data) ]

        objects.append( object )

    return file_name, int(width), int(height), objects
    
image_dir = 'images'    
xml_dir = 'labels'
xml_files = os.listdir(xml_dir)
image_files = os.listdir(image_dir)
image_ext = image_files[0].split('.')[-1] #图片文件的扩展名
print(image_ext)
if len(image_files) == len(xml_files):
    print('共有{:d}个xml文件。'.format(len(xml_files)))
else:
    print('图片数量和xml文件数量不一致。')
obj_dict = {}
for xml_file in xml_files:
    annotation = xml_parser(os.path.join(xml_dir, xml_file))
    name_1 = xml_file.split('.')[0] + '.' + image_ext.lower()
    name_2 = xml_file.split('.')[0] + '.' + image_ext.upper()
    if  name_1 not in image_files and name_2 not in image_files:
        print('{:s}没有对应的图片。'.format(xml_file))
    for obj in annotation[-1]:
        key = obj[0]
        x = (obj[1] + obj[3])/2
        y = (obj[2] + obj[4])/2
        width = obj[3] - obj[1]
        height = obj[4] - obj[2]
        box = [x,y,width,height]
        if key in obj_dict:
            obj_dict[key][0] += 1
            n = obj_dict[key][0]
            obj_dict[key][1:5] = [ (i*(n-1)+j)/(n) for i,j in zip(obj_dict[key][1:5] , box)]
            #obj_dict[key][5:9] = [ i if i>=j else j for i,j in zip(obj_dict[key][5:9] , box)]
            #obj_dict[key][9:] = [ i if i>=j else j for i,j in zip(obj_dict[key][9:] , box)]
        else:
            obj_dict[key] = []
            obj_dict[key].append(1)  # 0,个数
            obj_dict[key] += box     # 1-4, 平均坐标
            #obj_dict[key] += box     # 5-8, 最大值
            #obj_dict[key] += box     # 9-12,最小值
for key,value in obj_dict.items():
    print('一共有 {:4d} 个 {:20s},其边框平均位置为{:4.0f} *{:4.0f};平均尺寸为{:3.0f} *{:3.0f}。'.format(value[0],key,*value[1:]))
版权声明:本文内容由互联网用户自发贡献,该文观点仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 举报,一经查实,本站将立刻删除。

发布者:全栈程序员-用户IM,转载请注明出处:https://javaforall.cn/2159.html原文链接:https://javaforall.cn

【正版授权,激活自己账号】: Jetbrains全家桶Ide使用,1年售后保障,每天仅需1毛

【官方授权 正版激活】: 官方授权 正版激活 支持Jetbrains家族下所有IDE 使用个人JB账号...

(0)


相关推荐

发表回复

您的电子邮箱地址不会被公开。

关注全栈程序员社区公众号