水果识别-训练

水果识别

水果识别-搭建框架
水果识别-数据集制作
水果识别-训练
水果识别-测试

训练包括：

• 更改voc_label.py生成label文件
• 修改要训练的网络配置文件
• 修改voc.names
• 修改voc.data
• 训练

1. 更改voc_label.py生成label文件，该文件在/darknet/scripts/目录下
   
   import xml.etree.ElementTree as ET import pickle import os from os import listdir, getcwd from os.path import join # sets=[('2012', 'train'), ('2012', 'val'), ('2007', 'train'), ('2007', 'val'), ('2007', 'test')] # classes = ["aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"] sets=[('2007', 'train')] 这里只用到训练，数据集命名为VOC2017 classes = ["apple", "pear", "banana"] 只用了三个类别作为展示 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('VOCdevkit/VOC%s/Annotations/%s.xml'%(year, image_id)) out_file = open('VOCdevkit/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('VOCdevkit/VOC%s/labels/'%(year)): os.makedirs('VOCdevkit/VOC%s/labels/'%(year)) image_ids = open('VOCdevkit/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('%s/VOCdevkit/VOC%s/JPEGImages/%s.jpg\n'%(wd, year, image_id)) convert_annotation(year, image_id) list_file.close()

   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

   import xml.etree.ElementTree as ET import pickle import os from os import listdir, getcwd from os.path import join   # sets=[('2012', 'train'), ('2012', 'val'), ('2007', 'train'), ('2007', 'val'), ('2007', 'test')] # classes = ["aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"]   sets=[('2007', 'train')] 这里只用到训练，数据集命名为VOC2017 classes = ["apple", "pear", "banana"] 只用了三个类别作为展示     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('VOCdevkit/VOC%s/Annotations/%s.xml'%(year, image_id))    out_file = open('VOCdevkit/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('VOCdevkit/VOC%s/labels/'%(year)):        os.makedirs('VOCdevkit/VOC%s/labels/'%(year))    image_ids = open('VOCdevkit/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('%s/VOCdevkit/VOC%s/JPEGImages/%s.jpg\n'%(wd, year, image_id))        convert_annotation(year, image_id)    list_file.close()

   
   修改好后执行python voc_label.py 在scripts/VOCdevkit/VOC2007目录下生成了labels文件夹，文件夹里面长这个样子的：

   

   同时会在/darknet/scripts目录下生成2007_train.txt文件

2. 配置文件的修改，我用到的yolov3，所以就修改/cfg/yolove-voc.cfg
   
   [net] # Testing # batch=1 # subdivisions=1 # Training # 训练的时候把Testing的参数注释掉，把Training的参数取消注释 batch=64 subdivisions=16 width=416 height=416 channels=3 momentum=0.9 decay=0.0005 angle=0 saturation = 1.5 exposure = 1.5 hue=.1 learning_rate=0.001 # 根据需要修改学习率，就是梯度下降的速率，越大训练速度越快但会牺牲准确率 burn_in=1000 max_batches = 100000 # 训练的总次数 policy=steps steps=100,10000,70000,80000,90000 # 训练到相应次数后学习率变化 scales=10,10,.1,.1,.1 # 学习率变化情况，是累乘操作 ...... [convolutional] size=1 stride=1 pad=1 filters=24 # filters = 3 × (calsses + 5) # yolov2和tiny-yolov2中为 filter= num ×（classes + coords + 1） activation=linear [yolo] mask = 0,1,2 anchors = 10,13, 16,30, 33,23, 30,61, 62,45, 59,119, 116,90, 156,198, 373,326 classes=3 # 修改类别数量 num=9 jitter=.3 ignore_thresh = .5 truth_thresh = 1 random=1 # 为1时会启用Multi-Scale Training，随机使用不同尺寸的图片进行训练，显存小可以置为0

   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

   [net] # Testing # batch=1 # subdivisions=1 # Training       # 训练的时候把Testing的参数注释掉，把Training的参数取消注释 batch=64 subdivisions=16 width=416 height=416 channels=3 momentum=0.9 decay=0.0005 angle=0 saturation = 1.5 exposure = 1.5 hue=.1   learning_rate=0.001          # 根据需要修改学习率，就是梯度下降的速率，越大训练速度越快但会牺牲准确率 burn_in=1000 max_batches = 100000     # 训练的总次数 policy=steps steps=100,10000,70000,80000,90000        # 训练到相应次数后学习率变化 scales=10,10,.1,.1,.1        # 学习率变化情况，是累乘操作   ......     [convolutional] size=1 stride=1 pad=1 filters=24           # filters = 3 × (calsses + 5) # yolov2和tiny-yolov2中为 filter= num ×（classes + coords + 1） activation=linear   [yolo] mask = 0,1,2 anchors = 10,13,  16,30,  33,23,  30,61,  62,45,  59,119,  116,90,  156,198,  373,326 classes=3        # 修改类别数量 num=9 jitter=.3 ignore_thresh = .5 truth_thresh = 1 random=1         # 为1时会启用Multi-Scale Training，随机使用不同尺寸的图片进行训练，显存小可以置为0

3. 修改voc.names，该文件在/darknet/data目录下
   
   apple pear banana

   1 2 3 4

   apple pear banana

   
   该文件下记录所有类别名称。
4. 修改voc.data，该文件在/darknet/cfg目录下
   
   classes= 3 //类别数量 train = /home/embedded/lm/darknet/scripts/2007_train.txt //指定训练样本 // valid = /home/pjreddie/data/voc/2007_test.txt 指定测试数据，训练用不到 names = /home/embedded/lm/darknet/data/voc.names //指向上一步修改的voc.names文件 backup = /home/embedded/lm/darknet/backup/yolov3_voc_weights //指定训练后的权值放在哪儿

   1 2 3 4 5 6

   classes= 3 //类别数量 train  = /home/embedded/lm/darknet/scripts/2007_train.txt    //指定训练样本 // valid  = /home/pjreddie/data/voc/2007_test.txt  指定测试数据，训练用不到 names = /home/embedded/lm/darknet/data/voc.names //指向上一步修改的voc.names文件 backup = /home/embedded/lm/darknet/backup/yolov3_voc_weights //指定训练后的权值放在哪儿

5. 训练不要与训练权重直接训练：
   
   ./darknet detector train cfg/voc.data cfg/yolov3_voc.cfg

   1 2	./darknet detector train cfg/voc.data cfg/yolov3_voc.cfg

   
   也可以下载预训练权重进行训练：
   
   ./darknet detector train cfg/voc.data cfg/yolov3_voc.cfg yolo.weights

   1 2	./darknet detector train cfg/voc.data cfg/yolov3_voc.cfg yolo.weights

   
   训练过程中可以随时停止，停止以后可以在保存的权重处接着开始训练，保存的权重就当做预训练权重，yolov3开始没训练100次保存一次权重，过1000以后每训练10000次保存一次权重。然后就等待漫长的训练过程……