Ultralytics/yolov3实现钢筋检测 /在自己的数据集上训练yolov3

Ethereal_t

已于 2023-11-11 11:45:44 修改

阅读量387

点赞数 6

文章标签： YOLO yolov3 目标检测

于 2023-11-11 11:04:39 首次发布

本文链接：https://blog.csdn.net/Ethereal_t/article/details/134345182

版权

1. 配置环境

conda create -n yolo python=3.8
conda activate yolo

git clone https://github.com/ultralytics/yolov3  
cd yolov3
pip install -r requirements.txt

2. 处理数据集

数据集下载：

https://bj.bcebos.com/paddlex/examples2/rebar_count/dataset_reinforcing_steel_bar_counting.zip
下载后的数据集格式如图
ps：我用的方法比较麻烦，四处拼凑的代码，有时间可以做一个融合

将VOC格式数据集拆分到训练集和测试集

做拆分，但仅仅是将图像名称保存在了不同的txt文件中
代码参考： link

import os  
import random  
  
trainval_percent = 0.9
train_percent = 1
xmlfilepath = 'Annotations'  
txtsavepath = 'ImageSets\Main'  
total_xml = os.listdir(xmlfilepath)  
  
num=len(total_xml)  
list=range(num)  
tv=int(num*trainval_percent)  
tr=int(tv*train_percent)  
trainval= random.sample(list,tv)  
train=random.sample(trainval,tr)  
  
ftrainval = open('ImageSets/Main/trainval.txt', 'w')  
ftest = open('ImageSets/Main/test.txt', 'w')  
ftrain = open('ImageSets/Main/train.txt', 'w')  
fval = open('ImageSets/Main/val.txt', 'w')  
  
for i  in list:  
    name=total_xml[i][:-4]+'\n'  
    if i in trainval:  
        ftrainval.write(name)  
        if i in train:  
            ftrain.write(name)  
        else:  
            fval.write(name)  
    else:  
        ftest.write(name)  
  
ftrainval.close()  
ftrain.close()  
fval.close()  
ftest.close()

将所有图像的xml标注转换成了txt标签，存储在labels文件夹里
代码参考： link

# -*- coding: utf-8 -*-
"""
Created on Tue Oct  2 11:42:13 2018
将本文件放到VOC2007目录下，然后就可以直接运行
需要修改的地方：
1. sets中替换为自己的数据集
2. classes中替换为自己的类别
3. 将本文件放到VOC2007目录下
4. 直接开始运行
"""

import xml.etree.ElementTree as ET
import pickle
import os
from os import listdir, getcwd
from os.path import join
sets=[('rebar', 'train'), ('rebar', 'val'), ('rebar', 'test')]  #替换为自己的数据集
classes = ["rebar"]     #修改为自己的类别

#进行归一化
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('VOC%s/Annotations/%s.xml'%(year, image_id))  #将数据集放于当前目录下
    out_file = open('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('VOC%s/labels/'%(year)):
        os.makedirs('VOC%s/labels/'%(year))
    image_ids = open('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('../datasets/VOC%s/JPEGImages/%s.jpg\n'%(year, image_id))
        convert_annotation(year, image_id)
    list_file.close()

层级变成了这样：
在这里插入图片描述

根据txt文件将具体的图像文件拆分到不同的文件夹里

图像
代码参考：link

from PIL import Image
im_num = []
for line in open("VOCrebar/ImageSets/Main/train.txt", "r"):
    im_num.append(line)
    print(im_num)

for a in im_num:
    im_name = '/***/***/***/JPEGImages/{}'.format(a[:-1]) + '.jpg'  #原始路径
    print(im_name)
    im = Image.open(im_name)  # 打开指定路径下的图像

    tar_name = '/***/***/***/JPEGImages/train/{}'.format(a[:-1]) + '.jpg'  #移动后的路径
    print(tar_name)
    im=im.convert('RGB')
    im.save(tar_name)  # 另存
    im.close()

标签

from PIL import Image
import shutil
im_num = []
for line in open("VOCrebar/ImageSets/Main/train.txt", "r"):
    im_num.append(line)
    print(im_num)

for a in im_num:
    im_name = 'VOCrebar/labels/{}'.format(a[:-1]) + '.txt'  #原始路径
    print(im_name)

    tar_name = 'images/labels/train/{}'.format(a[:-1]) + '.txt'  #移动后的路径
    print(tar_name)

    shutil.copy(im_name,tar_name)