yolov5训练自己的数据

前一篇文章写了如何的安装yolo5。基于上面的一章，记录下用yolo5来训练自己的数据。

split_train_val.py

import os
import random
trainval_percent = 0.1
train_percent = 0.9
xmlfilepath = '/Users/Tony/IdeaProjects/yolov5/data/mydata/xml'
txtsavepath = '/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets'
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('/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets/trainval.txt', 'w')
ftest = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets/test.txt', 'w')
ftrain = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets/train.txt', 'w')
fval = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets/val.txt', 'w')
for i in list:
    name = total_xml[i][:-4] + '\n'
    if i in trainval:
        ftrainval.write(name)
        if i in train:
            ftest.write(name)
        else:
            fval.write(name)
    else:
        ftrain.write(name)
ftrainval.close()
ftrain.close()
fval.close()
ftest.close(）

以下是lable 文件 voc_label.py

import xml.etree.ElementTree as ET
import pickle
import os
from os import listdir, getcwd
from os.path import join
sets = ['train', 'test','val']
# 这里就是标签的名称，比如：dog，cat 等等和你图片标注的匹配好，用vott或者lableimg都可以进行标注。
classes = ['','']
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('/Users/Tony/IdeaProjects/yolov5/data/mydata/xml/%s.xml' % (image_id))
    out_file = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/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')
wd = getcwd()
print(wd)
for image_set in sets:
    if not os.path.exists('/Users/Tony/IdeaProjects/yolov5/data/mydata/labels/'):
        os.makedirs('/Users/Tony/IdeaProjects/yolov5/data/mydata/labels/')
    image_ids = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/ImageSets/%s.txt' % (image_set)).read().strip().split()
    list_file = open('/Users/Tony/IdeaProjects/yolov5/data/mydata/%s.txt' % (image_set), 'w')
    for image_id in image_ids:
        list_file.write('/Users/Tony/IdeaProjects/yolov5/data/mydata/images/%s.jpg\n' % (image_id))
        convert_annotation(image_id)
    list_file.close()

最后运行 train.py 文件进行已经打好标注的文件来进行训练，我这里运行的是基于 yolov5s.pt。

运行过程中可以执行

tensorboard --logdir=runs/train

当然自带有有可视化界面参考

train.py 文件执行之后就等着训练结果的生成，最后会在runs/train 产生训练的结果。

我们用自己生成好的.pt 文件来执行 detect.py ，生成路径位于runs/train/exp/weights 下的.pt文件

python detect.py --weights yolov5s.pt --source 0                               # webcam
                                                     img.jpg                         # image
                                                     vid.mp4                         # video
                                                     screen                          # screenshot
                                                     path/                           # directory
                                                     list.txt                        # list of images
                                                     list.streams                    # list of streams
                                                     'path/*.jpg'                    # glob
                                                     'https://youtu.be/Zgi9g1ksQHc'  # YouTube
                                                     'rtsp://example.com/media.mp4'  # RTSP, RTMP, HTTP stream

以上就是训练好自己的pt文件。