从爬取到训练：使用PaddlePaddle构建自定义图像分类模型¶
前言¶
本章将介绍如何使用PaddlePaddle训练自己的图片数据集。在之前的章节中，我们使用的是PaddlePaddle自带的数据集（如CIFAR-10），而本章将详细讲解如何从爬取图像开始，经过数据预处理、创建图像列表、定义模型、数据读取到最终训练和预测的完整流程。
1. 爬取图像数据¶
首先，我们需要获取自己的图像数据集。这里以爬取百度图片为例，编写一个爬虫程序。
1.1 创建download_image.py文件¶
import re
import uuid
import requests
import os
import numpy
import imghdr
from PIL import Image
# 获取百度图片下载图片
def download_image(key_word, save_name, download_max):
download_sum = 0
# 把每个类别的图片存放在单独一个文件夹中
save_path = 'images' + '/' + save_name
if not os.path.exists(save_path):
os.makedirs(save_path)
while download_sum < download_max:
download_sum += 1
str_pn = str(download_sum)
# 定义百度图片的路径
url = 'http://image.baidu.com/search/flip?tn=baiduimage&ie=utf-8&' \
'word=' + key_word + '&pn=' + str_pn + '&gsm=80&ct=&ic=0&lm=-1&width=0&height=0'
try:
s = requests.session()
s.headers['User-Agent'] = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3626.119 Safari/537.36'
# 获取当前页面的源码
result = s.get(url).content.decode('utf-8')
# 获取当前页面的图片URL
img_urls = re.findall('"objURL":"(.*?)",', result, re.S)
if img_urls is None or len(img_urls) < 1:
break
# 开始下载图片
for img_url in img_urls:
# 获取图片内容
img = requests.get(img_url, timeout=30)
# 保存图片
with open(save_path + '/' + str(uuid.uuid1()) + '.jpg', 'wb') as f:
f.write(img.content)
print('正在下载 %s 的第 %d 张图片' % (key_word, download_sum))
download_sum += 1
# 下载次数超过指定值就停止下载
if download_sum >= download_max:
break
except Exception as e:
print(e)
continue
print('下载完成')
# 删除不是JPEG或者PNG格式的图片
def delete_error_image(father_path):
# 获取父级目录的所有文件以及文件夹
try:
image_dirs = os.listdir(father_path)
for image_dir in image_dirs:
image_dir = os.path.join(father_path, image_dir)
# 如果是文件夹就继续获取文件夹中的图片
if os.path.isdir(image_dir):
images = os.listdir(image_dir)
for image in images:
image = os.path.join(image_dir, image)
try:
# 获取图片的类型
image_type = imghdr.what(image)
# 如果图片格式不是JPEG同时也不是PNG就删除图片
if image_type is not 'jpeg' and image_type is not 'png':
os.remove(image)
print('已删除：%s' % image)
continue
# 删除灰度图
img = numpy.array(Image.open(image))
if len(img.shape) is 2:
os.remove(image)
print('已删除：%s' % image)
except:
os.remove(image)
print('已删除：%s' % image)
except:
pass
if __name__ == '__main__':
# 定义要下载的图片中文名称和英文名称，ps：英文名称主要是为了设置文件夹名
key_words = {'西瓜': 'watermelon', '哈密瓜': 'cantaloupe',
'樱桃': 'cherry', '苹果': 'apple', '黄瓜': 'cucumber', '胡萝卜': 'carrot'}
# 每个类别下载一千个
max_sum = 500
for key_word in key_words:
save_name = key_words[key_word]
download_image(key_word, save_name, max_sum)
# 删除错误图片
delete_error_image('images/')
1.2 运行爬虫¶
执行上述脚本后，将在images目录下生成各个类别的图片文件夹，每个文件夹中包含指定数量的图片。
注意：下载完成后，可能仍存在不属于当前类别的图片，需手动删除。
2. 创建图像列表¶
为了让PaddlePaddle能够识别和处理自定义图像数据集，需要生成固定格式的图像列表文件。
2.1 创建create_data_list.py文件¶
import json
import os
def create_data_list(data_root_path):
with open(data_root_path + "test.list", 'w') as f:
pass
with open(data_root_path + "train.list", 'w') as f:
pass
# 所有类别的信息
class_detail = []
# 获取所有类别
class_dirs = os.listdir(data_root_path)
# 类别标签
class_label = 0
# 获取总类别的名称
father_paths = data_root_path.split('/')
while True:
if father_paths[len(father_paths) - 1] == '':
del father_paths[len(father_paths) - 1]
else:
break
father_path = father_paths[len(father_paths) - 1]
all_class_images = 0
other_file = 0
# 读取每个类别
for class_dir in class_dirs:
if class_dir == 'test.list' or class_dir == "train.list" or class_dir == 'readme.json':
other_file += 1
continue
print('正在读取类别：%s' % class_dir)
# 每个类别的信息
class_detail_list = {}
test_sum = 0
trainer_sum = 0
# 统计每个类别有多少张图片
class_sum = 0
# 获取类别路径
path = data_root_path + "/" + class_dir
# 获取所有图片
img_paths = os.listdir(path)
for img_path in img_paths:
# 每张图片的路径
name_path = class_dir + '/' + img_path
# 如果不存在这个文件夹,就创建
if not os.path.exists(data_root_path):
os.makedirs(data_root_path)
# 每10张图片取一个做测试数据
if class_sum % 10 == 0:
test_sum += 1
with open(data_root_path + "test.list", 'a') as f:
f.write(name_path + "\t%d" % class_label + "\n")
else:
trainer_sum += 1
with open(data_root_path + "train.list", 'a') as f:
f.write(name_path + "\t%d" % class_label + "\n")
class_sum += 1
all_class_images += 1
# 说明的json文件的class_detail数据
class_detail_list['class_name'] = class_dir
class_detail_list['class_label'] = class_label
class_detail_list['class_test_images'] = test_sum
class_detail_list['class_trainer_images'] = trainer_sum
class_detail.append(class_detail_list)
class_label += 1
# 获取类别数量
all_class_sum = len(class_dirs) - other_file
# 说明的json文件信息
readjson = {}
readjson['all_class_name'] = father_path
readjson['all_class_sum'] = all_class_sum
readjson['all_class_images'] = all_class_images
readjson['class_detail'] = class_detail
jsons = json.dumps(readjson, sort_keys=True, indent=4, separators=(',', ': '))
with open(data_root_path + "readme.json", 'w') as f:
f.write(jsons)
print('图像列表已生成')
if __name__ == '__main__':
# 把生产的数据列表都放在自己的总类别文件夹中
data_root_path = "images/"
create_data_list(data_root_path)
2.2 运行图像列表生成程序¶
执行上述脚本后，将在images目录下生成train.list、test.list和readme.json文件，其中：
- train.list：训练图像列表（每行格式：图片路径\t标签）
- test.list：测试图像列表
- readme.json：数据集说明文件
3. 定义模型¶
使用MobileNet V1模型进行图像分类，该模型轻量且适合嵌入式设备。
3.1 创建mobilenet_v1.py文件¶
import paddle.fluid as fluid
def conv_bn_layer(input, filter_size, num_filters, stride,
padding, channels=None, num_groups=1, act='relu', use_cudnn=True):
conv = fluid.layers.conv2d(input=input,
num_filters=num_filters,
filter_size=filter_size,
stride=stride,
padding=padding,
groups=num_groups,
act=None,
use_cudnn=use_cudnn,
bias_attr=False)
return fluid.layers.batch_norm(input=conv, act=act)
def depthwise_separable(input, num_filters1, num_filters2, num_groups, stride, scale):
depthwise_conv = conv_bn_layer(input=input,
filter_size=3,
num_filters=int(num_filters1 * scale),
stride=stride,
padding=1,
num_groups=int(num_groups * scale),
use_cudnn=False)
pointwise_conv = conv_bn_layer(input=depthwise_conv,
filter_size=1,
num_filters=int(num_filters2 * scale),
stride=1,
padding=0)
return pointwise_conv
def net(input, class_dim, scale=1.0):
# conv1: 112x112
input = conv_bn_layer(input=input,
filter_size=3,
channels=3,
num_filters=int(32 * scale),
stride=2,
padding=1)
# 56x56
input = depthwise_separable(input=input,
num_filters1=32,
num_filters2=64,
num_groups=32,
stride=1,
scale=scale)
input = depthwise_separable(input=input,
num_filters1=64,
num_filters2=128,
num_groups=64,
stride=2,
scale=scale)
# 28x28
input = depthwise_separable(input=input,
num_filters1=128,
num_filters2=128,
num_groups=128,
stride=1,
scale=scale)
input = depthwise_separable(input=input,
num_filters1=128,
num_filters2=256,
num_groups=128,
stride=2,
scale=scale)
# 14x14
input = depthwise_separable(input=input,
num_filters1=256,
num_filters2=256,
num_groups=256,
stride=1,
scale=scale)
input = depthwise_separable(input=input,
num_filters1=256,
num_filters2=512,
num_groups=256,
stride=2,
scale=scale)
# 14x14
for i in range(5):
input = depthwise_separable(input=input,
num_filters1=512,
num_filters2=512,
num_groups=512,
stride=1,
scale=scale)
# 7x7
input = depthwise_separable(input=input,
num_filters1=512,
num_filters2=1024,
num_groups=512,
stride=2,
scale=scale)
input = depthwise_separable(input=input,
num_filters1=1024,
num_filters2=1024,
num_groups=1024,
stride=1,
scale=scale)
feature = fluid.layers.pool2d(input=input,
pool_size=0,
pool_stride=1,
pool_type='avg',
global_pooling=True)
net = fluid.layers.fc(input=feature,
size=class_dim,
act='softmax')
return net
4. 定义数据读取¶
创建数据读取器，实现图像预处理和数据增强。
4.1 创建reader.py文件¶
import os
import random
import numpy as np
import paddle
from PIL import Image
# 训练图片的预处理
def train_mapper(sample):
img_path, label, crop_size, resize_size = sample
try:
img = Image.open(img_path)
# 统一图片大小
img = img.resize((resize_size, resize_size), Image.ANTIALIAS)
# 随机水平翻转
if random.random() > 0.5:
img = img.transpose(Image.FLIP_LEFT_RIGHT)
# 随机角度翻转
img = img.rotate(random.randint(-3, 3), expand=False)
# 随机裁剪
r4 = random.randint(0, int(resize_size - crop_size))
r5 = random.randint(0, int(resize_size - crop_size))
img = img.crop((r4, r5, r4 + crop_size, r5 + crop_size))
# 转换为numpy数组
img = np.array(img).astype(np.float32)
# 转换为CHW格式
img = img.transpose((2, 0, 1))
# 转换为BGR并归一化
img = img[(2, 1, 0), :, :] / 255.0
return img, int(label)
except:
print("%s 该图片错误，请删除该图片并重新创建图像数据列表" % img_path)
# 获取训练的reader
def train_reader(train_list_path, crop_size, resize_size):
father_path = os.path.dirname(train_list_path)
def reader():
with open(train_list_path, 'r') as f:
lines = f.readlines()
np.random.shuffle(lines)
for line in lines:
img, label = line.strip().split('\t')
img = os.path.join(father_path, img)
yield img, label, crop_size, resize_size
return paddle.reader.xmap_readers(train_mapper, reader, os.cpu_count(), 102400)
# 测试图片的预处理
def test_mapper(sample):
img_path, label, crop_size = sample
img = Image.open(img_path)
img = img.resize((crop_size, crop_size), Image.ANTIALIAS)
img = np.array(img).astype(np.float32)
img = img.transpose((2, 0, 1))
img = img[(2, 1, 0), :, :] / 255.0
return img, int(label)
# 获取测试的reader
def test_reader(test_list_path, crop_size):
father_path = os.path.dirname(test_list_path)
def reader():
with open(test_list_path, 'r') as f:
lines = f.readlines()
for line in lines:
img, label = line.strip().split('\t')
img = os.path.join(father_path, img)
yield img, label, crop_size
return paddle.reader.xmap_readers(test_mapper, reader, os.cpu_count(), 1024)
5. 训练模型¶
使用上述定义的模型和数据读取器进行训练。
5.1 创建train.py文件¶
```python
import os
import shutil
import mobilenet_v1
import paddle as paddle
import reader
import paddle.fluid as fluid
定义参数¶
crop_size = 224
resize_size = 250
class_dim = 6 # 类别数量
定义输入层¶
image = fluid.layers.data(name=’image’, shape=[3, crop_size, crop_size], dtype=’float32’)
label = fluid.layers.data(name=’label’, shape=[1], dtype=’int64’)
获取MobileNet模型¶
model = mobilenet_v1.net(image, class_dim)
定义损失函数和准确率¶
cost = fluid.layers.cross_entropy(input=model, label=label)
avg