yolov3-keras版本训练自己的数据全过程

1、下载文件包：https://github.com/qqwweee/keras-yolo3

2、为此模型建立anaconda虚拟环境。

conda creat -n yolo python=3.6

安装相应的环境，conda install 安装即可
版本号对应：

• python 3.6.9
• cuda 9.2
• cudnn 7.3.1
• tensorflow-gpu 1.12.0
• keras 2.2.4

3、安装依赖

pip install Pillow
pip install matplotlib
conda install ffmpeg
pip install opencv-contrib-python

4、下载权重文件
https://pjreddie.com/media/files/yolov3.weights
也可以用命令行下载：

wget https://pjreddie.com/media/files/yolov3.weights

如果特别慢可以百度网盘下载：链接：https://pan.baidu.com/s/1vFlOhClLDcyvUfu0JatVtA
提取码：k7lo
5、转化权重为keras模型
将权重文件转换成.h5文件（代码包中有convert.py文件负责，生成的.h5文件在model_data中）
命令行输入：

python convert.py yolov3.cfg yolov3.weights model_data/yolo.h5

然后可以测试一下环境或者权重是否成功匹配好：
命令号输入：

python yolo_video.py --image

然后输入图片名字（单张）
6、准备数据集
①建立一下名字和对应关系的文件夹

#在keras-yolo3下创建
VOCdevkit
└── VOC2012
    ├── Annotations  （存标注好的xml文件，标注可用labelImg）
    ├── ImageSets
    │   ├── Layout
    │   ├── Main （存放4个txt文件，装着图片名字，train.txt test.txt trainval.txt val.txt预先建好）
    │   └── Segmentation
    ├── JPEGImages  （放图片）
    └── labels

没有备注的没用上。
②按比例分配测试图片和训练图片
写一个test.py用来分配1：9数据（Main中），执行完此程序Main中的四个txt文件中会有图片名字。

import os
import random

trainval_percent = 0.1
train_percent = 0.9
xmlfilepath = '/home/sys507/sun/VOCdevkit/VOC2012/Annotations'
txtsavepath = '/home/sys507/sun/VOCdevkit/VOC2012/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('/home/sys507/sun/VOCdevkit/VOC2012/ImageSets/Main/trainval.txt', 'w')
ftest = open('/home/sys507/sun/VOCdevkit/VOC2012/ImageSets/Main/test.txt', 'w')
ftrain = open('/home/sys507/sun/VOCdevkit/VOC2012/ImageSets/Main/train.txt', 'w')
fval = open('/home/sys507/sun/VOCdevkit/VOC2012/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:
            ftest.write(name)
        else:
            fval.write(name)
    else:
        ftrain.write(name)

ftrainval.close()
ftrain.close()
fval.close()
ftest.close()

③修该voc_annotation中的类别并执行此文件
修改成自己的类和对应的生成文件名字

修改完执行，可用pycharm直接执行，也可命令行

python voc_annotation.py

执行后会在主目录生成3个txt文件

④生成适合自己数据的anchor
执行k-means.py文件，会在主目录生成yolo_anchor.txt文件。
⑤修改model_data中voc_classes.txt和coco_classes.txt中个的类别，换成自己的类别。
7、开始训练
先在主目录建立一个名字为logs的文件夹，然后在里面再建立一个名为000的文件夹，用于存储最后的权重。
修改train,py中对应路径、初始训练权重、batch尺寸、迭代次数 ，讯改完执行train.py文件，开始训练。最后



备注：也可以用以下代码替换带原有的train.py代码，这个就是全部层数都训练，源文件是先训练后三层（冻结前面249层），然后再训练全部层。但是这个代码有个缺点就是不能保存中间权重，只有最终权重。

"""
Retrain the YOLO model for your own dataset.
"""
import numpy as np
import keras.backend as K
from keras.layers import Input, Lambda
from keras.models import Model
from keras.callbacks import TensorBoard, ModelCheckpoint, EarlyStopping

from yolo3.model import preprocess_true_boxes, yolo_body, tiny_yolo_body, yolo_loss
from yolo3.utils import get_random_data


def _main():
    annotation_path = '2012_train.txt'
    log_dir = 'logs/000/'
    classes_path = 'model_data/voc_classes.txt'
    anchors_path = 'yolo_anchors.txt'
    class_names = get_classes(classes_path)
    anchors = get_anchors(anchors_path)
    input_shape = (416, 416)  # multiple of 32, hw
    model = create_model(input_shape, anchors, len(class_names))
    train(model, annotation_path, input_shape, anchors, len(class_names), log_dir=log_dir)


def train(model, annotation_path, input_shape, anchors, num_classes, log_dir='logs/'):
    model.compile(optimizer='adam', loss={
   
        'yolo_loss': lambda y_true, y_pred: y_pred})
    logging = TensorBoard(log_dir=log_dir)
    checkpoint = ModelCheckpoint(log_dir + "ep{epoch:03d}-loss{loss:.3f}-val_loss{val_loss:.3f}.h5",
                                 monitor='val_loss', save_weights_only=True, save_best_only=True, period=1)
    batch_size = 10
    val_split = 0.05
    with open(annotation_path) as f:
        lines = f.readlines()
    np.random.shuffle(lines)
    num_val = int(len(lines) * val_split)
    num_train = len(lines) - num_val
    print('Train on {} samples, val on {} samples, with batch size {}.'.format(num_train, num_val, batch_size))

    model.fit_generator(data_generator_wrap(lines[:num_train], batch_size, input_shape, anchors, num_classes),
                        steps_per_epoch=max(1, num_train // batch_size),
                        validation_data=data_generator_wrap(lines[num_train:], batch_size, input_shape, anchors,
                                                            num_classes),
                        validation_steps=max(1, num_val // batch_size),
                        epochs=200,
                        initial_epoch=0)
    model.save_weights(log_dir + 'trained_weights.h5')


def get_classes(classes_path):
    with open(classes_path) as f:
        class_names = f.readlines()
    class_names = [c.strip() for c in class_names]
    return class_names


def get_anchors(anchors_path):
    with open(anchors_path) as f:
        anchors = f.readline()
    anchors = [float(x) for x in anchors.split(',')]
    return np.array(anchors).reshape(-1, 2)


def create_model(input_shape, anchors, num_classes, load_pretrained=False, freeze_body=False,
                 weights_path='trained_weights.h5'):
    K.clear_session()  # get a new session
    image_input = Input(shape=(None, None, 3))
    h, w = input_shape
    num_anchors = len(anchors)
    y_true = [Input(shape=(h // {
   0: 32, 1: 16, 2: 8}[l], w // {
   0: 32, 1: 16, 2: 8}[l], \
                           num_anchors // 3, num_classes + 5)) for l in range(3)]

    model_body = yolo_body(image_input, num_anchors // 3, num_classes)
    print('Create YOLOv3 model with {} anchors and {} classes.'.format(num_anchors, num_classes))

    if load_pretrained:
        model_body.load_weights(weights_path, by_name=True, skip_mismatch=True)
        print('Load weights {}.'.format(weights_path))
        if freeze_body:
            # Do not freeze 3 output layers.
            num = len(model_body.layers) - 7
            for i in range(num): model_body.layers[i].trainable = False
            print('Freeze the first {} layers of total {} layers.'.format(num, len(model_body.layers)))

    model_loss = Lambda(yolo_loss, output_shape=(1,), name='yolo_loss'