将IphotoDraw标注好的xml文件转成txt文件(三)

        接上一篇来说，将真实的样本过一遍baseline模型后得到最初版的boundingbox信息的txt文件，又将这些txt文件转成xml文件进行纠正，纠正后使用IphotoDraw导出的还是xml文件，因此还需要将这些xml文件再次转换成txt文件。（是不是很晕？哈哈，晕就对了，多想几遍就好了）

      以下是以IphotoDraw为标注软件，将纠正好的xml文件转成txt文件的代码：

#coding:utf-8
#@Time:2017/6/16 19:37
#@author: xxx
import os
import re
import math
import string
from tqdm import tqdm

# 提取xml文件中的矩形框四个角点跟标签


def get_new_coord(center_coord,ori_coord,rotate_angle):
    x_new = (ori_coord[0]-center_coord[0])*math.cos((rotate_angle/180.)*math.pi)+(ori_coord[1]-center_coord[1])*math.sin((rotate_angle/180.)*math.pi)+center_coord[0]
    y_new = (ori_coord[1]-center_coord[1])*math.cos((rotate_angle/180.)*math.pi)-(ori_coord[0]-center_coord[0])*math.sin((rotate_angle/180.)*math.pi)+center_coord[1]
    return int(x_new),int(y_new)


def get_coord_range(item):
    list_location = []
    for i in range(len(item)):
        if (item[i] == '"'):
            list_location.append(i + 1)
    X=item[list_location[0]:list_location[1]-1]
    Y=item[list_location[2]:list_location[3]-1]
    Width=item[list_location[4]:list_location[5]-1]
    Height=item[list_location[6]:list_location[7]-1]
    return X,Y,Width,Height


def GetItemLocation(xml_file):
    fid=open(xml_file,"r",encoding="utf-8")
    list=[]
    for line in fid.readlines():
        line=line.replace("\n","")
        line=line.replace(" ","")
        list.append(line)

    str1 = "".join(list)
    label_list=re.findall("<Text>.*?</Text>",str1)
    position_list=re.findall("<Extent.*?/>",str1)
    angle_list=re.findall("<Data.*?>",str1)

    # 取出文本框标签
    label_list_new = []
    for label in label_list:
        result_label = label[6:-7]
        label_list_new.append(result_label)

    # 取出文本框旋转角度
    angle_list_new = []
    for angle in angle_list:
        result_angle = float(angle.split('"')[-2])
        angle_list_new.append(result_angle)

    #取出左上角点 w h
    position_list_new = []
    for position in position_list:
        X, Y, Width, Height = get_coord_range(position)
        position_list_new.append([X, Y, Width, Height])


    # 得到标注框坐标
    point_list = []
    for i in range(len(position_list_new)):
        value = position_list_new[i]
        angle = angle_list_new[i]

        x1 = int(float(value[0]))  # 左上
        y1 = int(float(value[1]))
        x2 = int(float(value[0]) + float(value[2]))  # 右上
        y2 = int(float(value[1]))
        x3 = int(float(value[0]) + float(value[2]))  # 右下
        y3 = int(float(value[1]) + float(value[3]))
        x4 = int(float(value[0]))  # 左下
        y4 = int(float(value[1]) + float(value[3]))

        if not angle == 0:
            angle = -angle
            center_x = (x1 + x2 + x3 + x4) / 4
            center_y = (y1 + y2 + y3 + y4) / 4

            x1, y1 = get_new_coord([center_x, center_y], [x1, y1], angle)
            x2, y2 = get_new_coord([center_x, center_y], [x2, y2], angle)
            x3, y3 = get_new_coord([center_x, center_y], [x3, y3], angle)
            x4, y4 = get_new_coord([center_x, center_y], [x4, y4], angle)

        if int(x1) < 0:
            x1 = 0
        if (int(y1) < 0):
            y1 = 0
        if (int(y2) < 0):
            y2 = 0
        if (int(x4) < 0):
            x4 = 0

        point = [[x1,y1],[x2,y2],[x3,y3],[x4,y4]]

        point_list.append(point)

    return label_list_new,point_list



if __name__=='__main__':
    path=r'xxxxxx'

    import glob
    import traceback
    jpg_files=glob.glob(os.path.join(path,'*.jpg'))
    bar=tqdm(total=len(jpg_files))

    for file in jpg_files:
        bar.update(1)
        try:
            xml_file = file.replace('.jpg','_data.xml')
            label_list, position_list  = GetItemLocation(xml_file)
            # 左上 左下 右下 右上
            IMIMkey = file.split('\\')[-1].replace('.jpg', '')
            with open(os.path.join(path, IMIMkey + '.txt'), 'w+', encoding='utf-8') as fid:
                for i, label in enumerate(label_list):
                    position = position_list[i]
                    # flag = 0
                    # if len(label) > 0:
                    #     for dd in d:
                    #         if dd in label:
                    #             flag = 1
                    #             break
                    flag = 1
                    if flag == 1:
                        fid.writelines(str(position[0][0]) + ',' + str(position[0][1]) + ',' +
                                       str(position[1][0]) + ',' + str(position[1][1]) + ',' +
                                       str(position[2][0]) + ',' + str(position[2][1]) + ',' +
                                       str(position[3][0]) + ',' + str(position[3][1]) +','+ label +'\n')

        except:
            print(file)
            traceback.print_exc()