CCPD数据集处理
"""@Author: HuKai
@Date: 2022/5/2910:44
@github: https://github.com/HuKai97
"""
import os
import cv2
from shutil import copy2
def txt_translate(images_dir,filename, labels_dir):
list1 = filename.split("-", 3)# 第一次分割，以减号'-'做分割
subname = list1
list2 = filename.split(".", 1)
subname1 = list2
if subname1 == 'txt':
return
lt, rb = subname.split("_", 1)# 第二次分割，以下划线'_'做分割
lx, ly = lt.split("&", 1)
rx, ry = rb.split("&", 1)
width = int(rx) - int(lx)
height = int(ry) - int(ly)# bounding box的宽和高
cx = float(lx) + width / 2
cy = float(ly) + height / 2# bounding box中心点
img = cv2.imread(images_dir + filename)
if img is None:# 自动删除失效图片（下载过程有的图片会存在无法读取的情况）
os.remove(images_dir + filename)
return
width = width / img.shape
height = height / img.shape
cx = cx / img.shape
cy = cy / img.shape
txtname = filename.split(".", 1)
txtfile = labels_dir + txtname + ".txt"
# 蓝牌是第0类，绿牌是第1类
with open(txtfile, "w") as f:
f.write(str(1) + " " + str(cx) + " " + str(cy) + " " + str(width) + " " + str(height))
images_trainDir = "../../CCPD2019/images/train/"#（将图片文件夹中的6份放在这个文件夹下）
images_validDir = "../../CCPD2019/images/val/" #（将图片文件夹中的2份放在这个文件夹下）
images_detectDir = "../../CCPD2019/images/test/" #（将图片文件夹中的2份放在这个文件夹下）
labels_trainDir = "../../CCPD2019/labels/train/"
labels_validDir = "../../CCPD2019/labels/val/"
labels_detectDir = "../../CCPD2019/labels/test/"
src_dir = "../../CCPD2020/ccpd_green/"
num = 0
files_f = os.listdir(src_dir)
for files in files_f:
for file in os.listdir(src_dir+files):
temp = num%10
temp_dir = src_dir+files+'/'+file
if temp < 6 :
copy2(temp_dir,images_trainDir)
txt_translate(src_dir+files+'/',file,labels_trainDir)
elif temp >= 6 and temp < 8 :
copy2(temp_dir,images_validDir)
txt_translate(src_dir+files+'/',file,labels_validDir)
else :
copy2(temp_dir,images_detectDir)
txt_translate(src_dir+files+'/',file,labels_detectDir)
num += 1
# trainfiles = os.listdir(r"K:\MyProject\datasets\ccpd\new\ccpd_2019\base")#（图片文件夹）
# num_train = len(trainfiles)
# print("num_train: " + str(num_train) )
# index_list = list(range(num_train))
# print(index_list)
# random.shuffle(index_list)# 打乱顺序
# num = 0
# trainDir = r"K:\MyProject\datasets\ccpd\new\ccpd_2019\train" #（将图片文件夹中的6份放在这个文件夹下）
# validDir = r"K:\MyProject\datasets\ccpd\new\ccpd_2019\val" #（将图片文件夹中的2份放在这个文件夹下）
# detectDir = r"K:\MyProject\datasets\ccpd\new\ccpd_2019\test" #（将图片文件夹中的2份放在这个文件夹下）
# for i in index_list:
# fileName = os.path.join(r"K:\MyProject\datasets\ccpd\new\ccpd_2019\base", trainfiles)#（图片文件夹）+图片名=图片地址
# if num < num_train*0.7:# 7:1:2
# print(str(fileName))
# copy2(fileName, trainDir)
# elif num < num_train*0.8:
# print(str(fileName))
# copy2(fileName, validDir)
# else:
# print(str(fileName))
# copy2(fileName, detectDir)
# num += 1
页:
[1]