Win10环境下，将VOC数据集转为YOLOV5使用的数据集。
YOLOV5 采用的数据集和以前的yolo模型不一样，数据结构如下图：
images文件夹存放train和val的图片
labels里面存放train和val的物体数据，里面的每个txt文件和images里面的图片是一一对应的。
txt文件的内容如下：
格式：物体类别 x1 y1 x2 y2
坐标是不是真实的坐标，是将坐标除以长宽后的出来的。
数据介绍完了，下面讲如何将voc数据转为yolov5使用的数据集。
本次采用的数据集是PASCAL VOC 2007。
地址：
训练集和验证集：
http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCtrainval_06-Nov-2007.tar
测试集：
http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCtest_06-Nov-2007.tar
下载后解压，将测试集和训练集合并在一起。在YOLOV5工程下面新建tmp文件夹，然后将voc数据集放到tmp文件夹下面，如图：
在tmp文件夹下面新家voc2txt.py文件，将voc的数据转为txt数据。
讲解voc2txt.py代码：
导入包:
import xml.etree.ElementTree as ET
import os
from os import getcwd
列出数据集的类别：
sets=[('2007', 'train'), ('2007', 'val'), ('2007', 'test')]
classes = ["aeroplane", "bicycle", "bird", "boat", "bottle",
"bus", "car", "cat", "chair", "cow", "diningtable", "dog",
"horse", "motorbike", "person", "pottedplant", "sheep",
"sofa", "train", "tvmonitor"]
对box进行转换，转换后的坐标就是相对长宽的小数:
def convert(size, box):
dw = 1./(size[0])
dh = 1./(size[1])
x = (box[0] + box[1])/2.0 - 1
y = (box[2] + box[3])/2.0 - 1
w = box[1] - box[0]
h = box[3] - box[2]
x = x*dw
w = w*dw
y = y*dh
h = h*dh
return (x,y,w,h)
下面这个方法是获取单个xml的内容，将其转换。
def convert_annotation(year, image_id):
in_file = open('VOCdevkit/VOC%s/Annotations/%s.xml'%(year, image_id))
out_file = open('VOCdevkit/VOC%s/labels/%s.txt'%(year, image_id), 'w')
tree=ET.parse(in_file)
root = tree.getroot()
size = root.find('size')
w = int(size.find('width').text)
h = int(size.find('height').text)
for obj in root.iter('object'):
difficult = obj.find('difficult').text
cls = obj.find('name').text
if cls not in classes or int(difficult)==1:
continue
cls_id = classes.index(cls)
xmlbox = obj.find('bndbox')
b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
bb = convert((w,h), b)
out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')
将转换后的train和val合并
file1 = open("2007_train.txt", "r")
file2 = open("2007_val.txt", "r")
file_list1 = file1.readlines() # 将所有变量读入列表file_list1
file_list2 = file2.readlines() # 将所有变量读入列表file_list2
file3=open("train.txt","w")
for line in file_list1:
print(line)
file3.write(line)
for line in file_list2:
print(line)
file3.write(line)
整体代码如下：
import xml.etree.ElementTree as ET
import os
from os import getcwd
sets=[('2007', 'train'), ('2007', 'val'), ('2007', 'test')]
classes = ["aeroplane", "bicycle", "bird", "boat", "bottle",
"bus", "car", "cat", "chair", "cow", "diningtable", "dog",
"horse", "motorbike", "person", "pottedplant", "sheep",
"sofa", "train", "tvmonitor"]
def convert(size, box):
dw = 1./(size[0])
dh = 1./(size[1])
x = (box[0] + box[1])/2.0 - 1
y = (box[2] + box[3])/2.0 - 1
w = box[1] - box[0]
h = box[3] - box[2]
x = x*dw
w = w*dw
y = y*dh
h = h*dh
return (x,y,w,h)
def convert_annotation(year, image_id):
in_file = open('VOCdevkit/VOC%s/Annotations/%s.xml'%(year, image_id))
out_file = open('VOCdevkit/VOC%s/labels/%s.txt'%(year, image_id), 'w')
tree=ET.parse(in_file)
root = tree.getroot()
size = root.find('size')
w = int(size.find('width').text)
h = int(size.find('height').text)
for obj in root.iter('object'):
difficult = obj.find('difficult').text
cls = obj.find('name').text
if cls not in classes or int(difficult)==1:
continue
cls_id = classes.index(cls)
xmlbox = obj.find('bndbox')
b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
bb = convert((w,h), b)
out_file.write(str(cls_id) + " " + " ".join([str(a) for a in bb]) + '\n')
wd = getcwd()
for year, image_set in sets:
if not os.path.exists('VOCdevkit/VOC%s/labels/'%(year)):
os.makedirs('VOCdevkit/VOC%s/labels/'%(year))
image_ids = open('VOCdevkit/VOC%s/ImageSets/Main/%s.txt'%(year, image_set)).read().strip().split()
list_file = open('%s_%s.txt'%(year, image_set), 'w')
for image_id in image_ids:
list_file.write('%s/VOCdevkit/VOC%s/JPEGImages/%s.jpg\n'%(wd, year, image_id))
convert_annotation(year, image_id)
list_file.close()
file1 = open("2007_train.txt", "r")
file2 = open("2007_val.txt", "r")
file_list1 = file1.readlines() # 将所有变量读入列表file_list1
file_list2 = file2.readlines() # 将所有变量读入列表file_list2
file3=open("train.txt","w")
for line in file_list1:
print(line)
file3.write(line)
for line in file_list2:
print(line)
file3.write(line)
运行完成后会生成2007_test.txt 2007_train.txt 2007_val.txt和train.txt。如图：
在tmp文件夹新建makedata.py，将生成的中间结果转为YOLOV5所使用的最终代码。
代码如下：
import shutil import os if not os.path.exists('../VOC/images/train'): os.makedirs('../VOC/images/train') if not os.path.exists('../VOC/images/val'): os.makedirs('../VOC/images/val') if not os.path.exists('../VOC/labels/train'): os.makedirs('../VOC/labels/train') if not os.path.exists('../VOC/labels/val'): os.makedirs('../VOC/labels/val') print(os.path.exists('../tmp/train.txt')) f = open('../tmp/train.txt', 'r') lines = f.readlines() for line in lines: print(line) line = "/".join(line.split('/')[-5:]).strip() shutil.copy(line,"../VOC/images/train") line = line.replace('JPEGImages', 'labels') line = line.replace('jpg', 'txt') shutil.copy(line, "../VOC/labels/train/") print(os.path.exists('../tmp/2007_test.txt')) f = open('../tmp/2007_test.txt', 'r') lines = f.readlines() for line in lines: line = "/".join(line.split('/')[-5:]).strip() print(line) shutil.copy(line, "../VOC/images/val") line = line.replace('JPEGImages', 'labels') line = line.replace('jpg', 'txt') shutil.copy(line, "../VOC/labels/val")
执行完成后，会在yolov5工程下生成最终的数据集。