【ai研发日记】faster-rcnn voc数据集制作教程

内部业务，记录一下过程

1、首先将解决方案文档用数字-项目名缩写重命名排序，方便标签命名，并新建一个文件夹保存报错截图

2、打开其中一个解决方案文档

3、将里面的图另存出来并单独截取成如下报错截图

4、另存到报错截图文件夹中，并重命名为解决方案文件同名

5、点开图片，在不同的背景下用连拍拍不同角度

6、新建一个文件夹，重命名为数字+项目缩写前缀（如“12-X9-无效参数：separationMethod.Name：Qrcode”重命名为“12-X9”）将图片全部存入进去

7、打开lableme软件，选择open dir，打开上一步新建的文件夹

选择后如下

点开file-save automatically,这样你标注好了就会自动保存

8、点开create polygons

标注的时候就选择截图所在的位置，并将标签名命名为选择的文件夹名（如12-X9）

如果标歪了可以选择edit polygons调整

按一下D，跳转到下一个图片，继续标注，同一种报错用同一个标签
在这里插入图片描述

比如另一种报错就用另一种标签

接下来就是循环标注了
最终出来的是这种效果，图片和json文件一一对应
如果标签选错了删除对应的json文件重新标注即可

收集起来集中在一个文件夹中用python代码进行一下重命名

# -*- coding: utf-8 -*-
# @Time    : 2022-04-19 14:14
# @Author  : 杨尚
# @FileName: FileSorting.py
# @Software: PyCharm
import os,json
def findAllFile(base):
    j=0
    for root, ds, fs in os.walk(base):

        for i ,f in enumerate(fs):
            if f.endswith('.jpg'):
                j+=1
                print(j)
                name = f.split(".")[0]
                jpgpath = os.path.join(root,f)
                jsonpath = os.path.join(root,f"{name}.json")
                newjpgpath = jpgpath.replace(name,str(j))
                newjsonpath = jsonpath.replace(name,str(j))

                print(newjpgpath,newjsonpath)
                with open(jsonpath,'r') as fp:
                    data = json.load(fp)

                print(data['imagePath'])
                data['imagePath'] = f"{str(j)}.jpg"
                with open(newjsonpath, 'w') as fp1:
                    fp1.write(json.dumps(data, indent=4, ensure_ascii=False))
                os.rename(jpgpath, newjpgpath)
                # os.rename(jpgpath,jpgpath.replace(name,str(i+1)))
                # os.rename(jsonpath,jsonpath.replace(name,str(i+1)))

                # print(jpgpath,jsonpath)
findAllFile(r'F:\pathroad')

然后再用代码转换成voc格式

```python
# -*- coding: utf-8 -*-
"""
Created on Thu Sep 19 14:51:00 2019
@author: Andrea
"""

import os
import numpy as np
import codecs
import json
from glob import glob
import cv2
import shutil
from sklearn.model_selection import train_test_split


#1.标签路径
labelme_path = r"F:\beto\trains"  # 原始labelme标注数据路径
img_format = "jpg"  # 图片格式

# 2.创建要求文件夹
saved_path = "./VOC2007/"  # 不懂者勿动
if not os.path.exists(saved_path + "Annotations"):
    os.makedirs(saved_path + "Annotations")
if not os.path.exists(saved_path + "JPEGImages/"):
    os.makedirs(saved_path + "JPEGImages/")
if not os.path.exists(saved_path + "ImageSets/Main/"):
    os.makedirs(saved_path + "ImageSets/Main/")

for json_file_ in os.listdir(labelme_path):
    if ('.json' not in json_file_):
        continue
    else:
        json_file_ = json_file_.split('.json')[0]
    print(json_file_)
    json_filename = os.path.join(labelme_path, json_file_ + ".json")
    print(json_filename)
    json_file = json.load(open(json_filename, "r", encoding="utf-8"))
    print(os.path.join(labelme_path, json_file_ + '.' + img_format))
    print()
    height, width, channels = cv2.imread(os.path.join(labelme_path, json_file_ + "." + img_format)).shape
    with codecs.open(saved_path + "Annotations/" + json_file_ + ".xml", "w", "utf-8") as xml:
        xml.write('<annotation>\n')
        xml.write('\t<folder>' + 'UAV_data' + '</folder>\n')
        xml.write('\t<filename>' + json_file_ + '.' + img_format + '</filename>\n')
        xml.write('\t<source>\n')
        xml.write('\t\t<database>The UAV autolanding</database>\n')
        xml.write('\t\t<annotation>UAV AutoLanding</annotation>\n')
        xml.write('\t\t<image>flickr</image>\n')
        xml.write('\t\t<flickrid>NULL</flickrid>\n')
        xml.write('\t</source>\n')
        xml.write('\t<owner>\n')
        xml.write('\t\t<flickrid>NULL</flickrid>\n')
        xml.write('\t\t<name>Yuanyiqin</name>\n')
        xml.write('\t</owner>\n')
        xml.write('\t<size>\n')
        xml.write('\t\t<width>' + str(width) + '</width>\n')
        xml.write('\t\t<height>' + str(height) + '</height>\n')
        xml.write('\t\t<depth>' + str(channels) + '</depth>\n')
        xml.write('\t</size>\n')
        xml.write('\t\t<segmented>0</segmented>\n')
        for multi in json_file["shapes"]:
            points = np.array(multi["points"])
            xmin = min(points[:, 0])
            xmax = max(points[:, 0])
            ymin = min(points[:, 1])
            ymax = max(points[:, 1])
            label = multi["label"]
            if xmax <= xmin:
                pass
            elif ymax <= ymin:
                pass
            else:
                xml.write('\t<object>\n')
                xml.write('\t\t<name>' + str(label) + '</name>\n')
                xml.write('\t\t<pose>Unspecified</pose>\n')
                xml.write('\t\t<truncated>1</truncated>\n')
                xml.write('\t\t<difficult>0</difficult>\n')
                xml.write('\t\t<bndbox>\n')
                xml.write('\t\t\t<xmin>' + str(xmin) + '</xmin>\n')
                xml.write('\t\t\t<ymin>' + str(ymin) + '</ymin>\n')
                xml.write('\t\t\t<xmax>' + str(xmax) + '</xmax>\n')
                xml.write('\t\t\t<ymax>' + str(ymax) + '</ymax>\n')
                xml.write('\t\t</bndbox>\n')
                xml.write('\t</object>\n')
                print(json_filename, xmin, ymin, xmax, ymax, label)
        xml.write('</annotation>')

# 5.复制图片到 VOC2007/JPEGImages/下

image_files = glob(os.path.join(labelme_path, "*." + img_format))
print("发现图片：{}张".format(image_files.__len__()))
print("copy image files to VOC007/JPEGImages/")
for image in image_files:
    shutil.copy(image, saved_path + "JPEGImages/")

# 6.split files for txt
txtsavepath = saved_path + "ImageSets/Main/"
ftrainval = open(txtsavepath + '/trainval.txt', 'w')
ftest = open(txtsavepath + '/test.txt', 'w')
ftrain = open(txtsavepath + '/train.txt', 'w')
fval = open(txtsavepath + '/val.txt', 'w')
total_files = glob("./VOC2007/Annotations/*.xml")
total_files = [i.split("/")[-1].split(".xml")[0] for i in total_files]
for file in total_files:
    ftrainval.write(file + "\n")
train_files, val_files = train_test_split(total_files, test_size=0.2, random_state=42)
for file in train_files:
    ftrain.write(file + "\n")
for file in val_files:
    fval.write(file + "\n")
ftrainval.close()
ftrain.close()
fval.close()

这样数据集就做好了