YOLOv7训练自己的数据集(txt文件，笔记）

目录

1.代码下载

2.数据集准备（.xml转.txt)

（1）修改图像文件名

（2）图片和标签文件数量不对应，解决办法

（3）.xml转.txt

（4）.txt文件随机划分出对应的训练集、测试集、验证集

3.训练数据集

（1）修改.yaml文件

 （2）修改网络参数

 （3）训练中断

---

1.代码下载

论文地址：https://arxiv.org/abs/2207.02696

论文代码下载地址：mirrors / WongKinYiu / yolov7 · GitCode

2.数据集准备（.xml转.txt)

（1）修改图像文件名

制作数据集运用到的一些功能代码_桦拾的博客-CSDN博客

（2）图片和标签文件数量不对应，解决办法

通过以下代码进行对比后，输出多余文件名，自己再根据实际情况和需要进行增删文件：python 两个文件夹里的文件名对比_inspur秃头哥的博客-CSDN博客

# -*- coding: utf-8 -*-
import ospath1 = r'./train'
path2 = r'./train_xml'def file_name(image_dir,xml_dir):jpg_list = []xml_list = []for root, dirs, files in os.walk(image_dir):for file in files:jpg_list.append(os.path.splitext(file)[0])for root, dirs, files in os.walk(xml_dir):for file in files:xml_list.append(os.path.splitext(file)[0])print(len(jpg_list))diff = set(xml_list).difference(set(jpg_list))  # 差集，在a中但不在b中的元素for name in diff:print("no jpg", name + ".xml")diff2 = set(jpg_list).difference(set(xml_list))  # 差集，在b中但不在a中的元素print(len(diff2))for name in diff2:print("no xml", name + ".jpg")
if __name__ == '__main__':file_name(path1,path2)

（3）.xml转.txt

（需增加或减少数据集种类，只需修改classes列表、if语句部分，以及print部分；对应文件夹目录也需修改）

将xml转化为yolov5的训练格式txt - 知乎

import os
from glob import glob
import xml.etree.ElementTree as ETxml_dir = r'F:\datasets\tomato_dataset\xml'####xml文件夹
output_txt_dir = r'F:\datasets\tomato_dataset\txt'####输出yolo所对应格式的文件夹###进行归一化操作
def convert(size, box):dw = 1. / (size[0])dh = 1. / (size[1])x = (box[0] + box[1]) / 2.0 - 1y = (box[2] + box[3]) / 2.0 - 1w = box[1] - box[0]h = box[3] - box[2]x = x * dww = w * dwy = y * dhh = h * dhreturn (x, y, w, h)def load_xml():####这里是你自己的分类classes = ['bruise','crack','blackspot', 'rot']xml_list = glob(os.path.join(xml_dir, '*.xml'))# print(len(xml_list), xml_list)count_pictures = {}count_detection = {}count = 0class_num0 = 0class_num1 = 0class_num2 = 0class_num3 = 0class_num4 = 0class_num5 = 0for file in xml_list:count = count + 1imgName = file.split('\\')[-1][:-4]  # 文件名，不包含后缀imglabel = os.path.join(output_txt_dir, imgName + '.txt')  # 创建TXT（文件夹路径加文件名加.TXT）# print(imglabel)out_file = open(imglabel, 'w', encoding='UTF-8')  # 以写入的方式打开TXTtree = ET.parse(file)root = tree.getroot()for child in root:if child.tag == 'size':w = child[0].texth = child[1].textif child.tag == 'object':x_min = child[4][0].texty_min = child[4][1].textx_max = child[4][2].texty_max = child[4][3].textbox = convert((int(w), int(h)), (int(x_min), int(x_max), int(y_min), int(y_max)))label = child[0].textif label in classes:##按照上面的顺序填写标签，如果超过这些自己增加复制就行了if label == 'bruise':label = '0'class_num0 += 1out_file.write(str(label) + ' ' + ' '.join([str(round(a, 6)) for a in box]) + '\n')  # 把内容写入TXT中if label == 'crack':label = '1'class_num1 += 1out_file.write(str(label) + ' ' + ' '.join([str(round(a, 6)) for a in box]) + '\n')if label == 'blackspot':label = '2'class_num2 += 1out_file.write(str(label) + ' ' + ' '.join([str(round(a, 6)) for a in box]) + '\n')if label == 'rot':label = '3'class_num3 += 1out_file.write(str(label) + ' ' + ' '.join([str(round(a, 6)) for a in box]) + '\n')print('ALL:', count, " bruise:", class_num0, "  crack:", class_num1, "  blackspot:", class_num2," rot:", class_num3)return len(xml_list), classes, count_pictures, count_detection  # return 用在函数内部表示当调用该函数时，if __name__ == '__main__':classes = load_xml()print(classes)

（4）.txt文件随机划分出对应的训练集、测试集、验证集

此代码生成相应的  train.txt;  val.txt;  test.txt    文件。

YOLOv7保姆级教程（个人踩坑无数）----训练自己的数据集_AmbitionToFree的博客-CSDN博客

# 将图片和标注数据按比例切分为 训练集和测试集
# 直接划分txt文件jpg文件
#### 强调！！！ 路径中不能出现中文，否则报错找不到文件
import shutil
import random
import os# 原始路径
image_original_path = r"E:\0_net_code\datasets\images"
label_original_path = r"E:\0_net_code\datasets\txt"cur_path = os.getcwd()# 训练集路径
train_image_path = os.path.join(cur_path, "datasets/images/train/")
train_label_path = os.path.join(cur_path, "datasets/labels/train/")
print("----------")
# 验证集路径
val_image_path = os.path.join(cur_path, "datasets/images/val/")
val_label_path = os.path.join(cur_path, "datasets/labels/val/")
print("----------")
# 测试集路径
test_image_path = os.path.join(cur_path, "datasets/images/test/")
test_label_path = os.path.join(cur_path, "datasets/labels/test/")
print("----------")
# 训练集目录
list_train = os.path.join(cur_path, "datasets/train.txt")
list_val = os.path.join(cur_path, "datasets/val.txt")
list_test = os.path.join(cur_path, "datasets/test.txt")
print("----------")
train_percent = 0.8
val_percent = 0.1
test_percent = 0.1
print("----------")def del_file(path):for i in os.listdir(path):file_data = path + "\\" + ios.remove(file_data)def mkdir():if not os.path.exists(train_image_path):os.makedirs(train_image_path)else:del_file(train_image_path)if not os.path.exists(train_label_path):os.makedirs(train_label_path)else:del_file(train_label_path)if not os.path.exists(val_image_path):os.makedirs(val_image_path)else:del_file(val_image_path)if not os.path.exists(val_label_path):os.makedirs(val_label_path)else:del_file(val_label_path)if not os.path.exists(test_image_path):os.makedirs(test_image_path)else:del_file(test_image_path)if not os.path.exists(test_label_path):os.makedirs(test_label_path)else:del_file(test_label_path)def clearfile():if os.path.exists(list_train):os.remove(list_train)if os.path.exists(list_val):os.remove(list_val)if os.path.exists(list_test):os.remove(list_test)def main():mkdir()clearfile()file_train = open(list_train, 'w')file_val = open(list_val, 'w')file_test = open(list_test, 'w')total_txt = os.listdir(label_original_path)num_txt = len(total_txt)list_all_txt = range(num_txt)num_train = int(num_txt * train_percent)num_val = int(num_txt * val_percent)num_test = num_txt - num_train - num_valtrain = random.sample(list_all_txt, num_train)# train从list_all_txt取出num_train个元素# 所以list_all_txt列表只剩下了这些元素val_test = [i for i in list_all_txt if not i in train]# 再从val_test取出num_val个元素，val_test剩下的元素就是testval = random.sample(val_test, num_val)print("训练集数目：{}, 验证集数目：{}, 测试集数目：{}".format(len(train), len(val), len(val_test) - len(val)))for i in list_all_txt:name = total_txt[i][:-4]srcImage = image_original_path + '\\' + name + '.jpg'srcLabel = label_original_path + '\\' + name + ".txt"if i in train:dst_train_Image = train_image_path + name + '.jpg'dst_train_Label = train_label_path + name + '.txt'shutil.copyfile(srcImage, dst_train_Image)shutil.copyfile(srcLabel, dst_train_Label)file_train.write(dst_train_Image + '\n')elif i in val:dst_val_Image = val_image_path + name + '.jpg'dst_val_Label = val_label_path + name + '.txt'shutil.copyfile(srcImage, dst_val_Image)shutil.copyfile(srcLabel, dst_val_Label)file_val.write(dst_val_Image + '\n')else:dst_test_Image = test_image_path + name + '.jpg'dst_test_Label = test_label_path + name + '.txt'shutil.copyfile(srcImage, dst_test_Image)shutil.copyfile(srcLabel, dst_test_Label)file_test.write(dst_test_Image + '\n')file_train.close()file_val.close()file_test.close()if __name__ == "__main__":main()

3.训练数据集

（1）修改.yaml文件

根据自己数据集情况，修改数据集文件路径、数据集种类、标签名

 （2）修改网络参数

根据GPU情况修改参数，注意权重文件路径、yaml文件路径

 （3）训练中断

若训练中断，可从最新的epcho开始训练，将resume参数的default改为True即可。