问题场景
在做目标检测任务时,我想提取训练集的图片单独进行外部数据增强。因此,需要根据划分出的train.txt
来提取训练集图片与标签。
需求实现
我使用VOC数据集进行测试,实现比较简单。
代码语言:javascript复制import shutil
if __name__ == '__main__':
img_src = r"D:DatasetVOC2007images"
xml_src = r"D:DatasetVOC2007Annotations"
img_out = "image_out/"
xml_out = "xml_out/"
txt_path = r"D:DatasetVOC2007ImageSetsSegmentationtrain.txt"
# 读取txt文件
with open(txt_path, 'r') as f:
line_list = f.readlines()
for line in line_list:
line_new = line.replace('n', '') # 将换行符替换为空('')
shutil.copy(img_src '/' line_new ".jpg", img_out)
shutil.copy(xml_src '/' line_new ".xml", xml_out)
效果:
更新训练集索引
使用数据增强之后,把生成的图片和标签丢到VOC里面,混在一起。
然后再写个脚本,将生成好的图片名称添加到train.txt
文件中。
import os
if __name__ == '__main__':
xml_src = r"C:UsersxyDesktopread_trainxml_out_af"
txt_path = r"D:DatasetVOC2007ImageSetsSegmentationtrain.txt"
for name in os.listdir(xml_src):
with open(txt_path, 'a') as f:
f.write(name[:-4] "n")
效果:
最后,再运行之前在VOC博文里面写过的xml2txt脚本:
代码语言:javascript复制import xml.etree.ElementTree as ET
import pickle
import os
from os import listdir, getcwd
from os.path import join
sets = ['train', 'test', 'val']
Imgpath = r'D:DatasetVOC2007images' # 图片文件夹
xmlfilepath = r'D:DatasetVOC2007Annotations' # xml文件存放地址
ImageSets_path = r'D:DatasetVOC2007ImageSetsSegmentation'
Label_path = r'D:DatasetVOC2007'
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
y = (box[2] box[3]) / 2.0
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(image_id):
in_file = open(xmlfilepath '/%s.xml' % (image_id))
out_file = open(Label_path '/labels/%s.txt' % (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')
for image_set in sets:
if not os.path.exists(Label_path 'labels/'):
os.makedirs(Label_path 'labels/')
image_ids = open(ImageSets_path '/%s.txt' % (image_set)).read().strip().split()
list_file = open(Label_path '%s.txt' % (image_set), 'w')
for image_id in image_ids:
# print(image_id) # DJI_0013_00360
list_file.write(Imgpath '/%s.jpgn' % (image_id))
convert_annotation(image_id)
list_file.close()
运行之后,就可以看到生成的数据增强样本被完美得添加到了原始数据集中。