1、xml
使用labelmg工具对图片进行标注得到xml格式文件,以图片为例介绍内容信息:
对上面的图片进行标注后,得到xml文件:
其内容分类两部分:
- 第一个黑色方框,图像的整体部分,包括图像的名称、位置、长宽高等等;
- 第二个黑色方框,标注框信息,每个红色框就是一个object标签(表示一个标注框)的信息,包括目标类别名称、位置信息等
xml内的信息是由一个个对象组成,标签之间存在层级关系,例如annotation为最上层的标签,就是这个xml所在的文件夹,其他标签为字标签。
2、xml -> csv
字符(逗号)分割值。
每个object标签代表一个标注框,都会在csv文件中生成一条数据,每天数据的属性为:图片文件名、宽度、高度、类别、框的左上角x坐标、框的左上角y、框的右上角x、框的右上角y。
xml转csv的代码如下:
# -*- coding: utf-8 -*- """ 将文件夹内所有XML文件的信息记录到CSV文件中 """ import os import glob import pandas as pd import xml.etree.ElementTree as ET def xml_to_csv(path): #path:annotations的文件夹路径 xml_list = [] for xml_file in glob.glob(path + '/*.xml'): #对path目录下的每一个xml文件 tree = ET.parse(xml_file) #获得xml对应的解析树 root = tree.getroot() #获得根标签annotations # print(root) print(root.find('filename').text) for member in root.findall('object'): #对每一个object标签(框) value = (root.find('filename').text, #在根标签下查找filename标签(图片文件名字),获得文本信息 int(root.find('size')[0].text), #在根标签下找size标签,并获得第0个字标签(width)的文本信息,转化为int int(root.find('size')[1].text), #在根标签下找size标签,并获得di1个字标签(height)的文本信息,转化为int member[0].text, #获得object标签的第0个字标签name的文信息 int(member[4][0].text), #获得object的第四个子标签bndbox,获得bndbox的第0个字标签(xmin)的文本信息,转化为int int(float(member[4][1].text)), #获得object的第四个子标签bndbox,获得bndbox的第1个字标签(ymin)的文本信息,转化为int int(member[4][2].text), #获得object的第四个子标签bndbox,获得bndbox的第2个字标签(xmax)的文本信息,转化为int int(member[4][3].text) #获得object的第四个子标签bndbox,获得bndbox的第3个字标签(ymax)的文本信息,转化为int ) xml_list.append(value) column_name = ['filename', 'width', 'height', 'class', 'xmin', 'ymin', 'xmax', 'ymax'] xml_df = pd.DataFrame(xml_list, columns=column_name) return xml_df def main(): for directory in ['train','test','validation']: #对应train和test文件夹 #对应根目录下的/images中的train和test文件夹,本脚本要放在voc文件夹下,和annotations是同级的,否则修改getcwd函数 xml_path = os.path.join(os.getcwd(), 'annotations/{}'.format(directory)) xml_df = xml_to_csv(xml_path) xml_df.to_csv('data/whsyxt_{}_labels.csv'.format(directory), index=None) #xml转化为对应的csv保存 print('Successfully converted xml to csv.') main()
对应的xml文件如下图:
最后得到两个文件:
文件打开类似于这样的:
其中的filename只是图片文件的名字,不包括路径。
3、xml转换为tfrecord
每个图片会生成一个xml文件,批量的将xml文件转化成tfrecord格式。
4、csv转换成tfrecord
将多个xml文件写入到一个csv文件中去,每一行是一个xml文件的信息,接下来直接将这个csv文件转换成tfrecord格式就可以了,很方便快。
由于图像和标签值不在一起,所以要将整张图片信息和csv文件合并起来,整合成为tfrecord格式写入到本地中,用于训练。
代码来自tensorflow/object_dection/models-master/research/object_detection/test_generate_tfrecord.py:
Usage: # From tensorflow/models/ # Create train data: python generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record # Create test data: python generate_tfrecord.py --csv_input=data/test_labels.csv --output_path=data/test.record """ from __future__ import division from __future__ import print_function from __future__ import absolute_import import os import io import pandas as pd import tensorflow as tf from PIL import Image from object_detection.utils import dataset_util from collections import namedtuple, OrderedDict flags = tf.app.flags """ DEFINE_string定义了个命令行参数 flage_name:csv_input,参数名字 defalut_name:默认值 ,这里的默认值是data/test_labels.csv docstring:对该参数的说明 可以使用tf.app.flags.FLAGS取出该参数的值: FLAGS = tf.app.flags.FLAGS print(FLAGS.csv_input),输出的就是data/test_labels.csv """ flags.DEFINE_string('csv_input', 'data/test_labels.csv', 'Path to the CSV input') flags.DEFINE_string('output_path', 'data/test.record', 'Path to output TFRecord') FLAGS = flags.FLAGS # TO-DO replace this with label map # 修改成你自己的标签 def class_text_to_int(row_label): if row_label == 'face': return 0 elif row_label == 'cat': return 1 #............ def split(df, group): """namedtuple工厂函数,返回一个名为`data`的类,并赋值给名为data的变量 定义:Point = namedtuple('Point', ['x', 'y']) 创建对象:p = Point(11, y=22) p[0] + p[1] 输出 33 解包:x, y = p x,y 输出:(11, 22) 访问:p.x + p.y 输出 33 """ data = namedtuple('data', ['filename', 'object']) gb = df.groupby(group) return [data(filename, gb.get_group(x)) for filename, x in zip(gb.groups.keys(), gb.groups)] #读取每张图片,得到每张图片的信息,将每张图片信息和图片里的object标注框信息(在csv里)合并在一起 #group #path:iamge目录 def create_tf_example(group, path): #image目录 + image的名字 = image的绝对路径路径 with tf.gfile.GFile(os.path.join(path, '{}'.format(group.filename)), 'rb') as fid: encoded_jpg = fid.read() encoded_jpg_io = io.BytesIO(encoded_jpg) image = Image.open(encoded_jpg_io) width, height = image.size filename = group.filename.encode('utf8') image_format = b'jpg' xmins = [] xmaxs = [] ymins = [] ymaxs = [] classes_text = [] classes = [] for index, row in group.object.iterrows(): xmins.append(row['xmin'] / width) xmaxs.append(row['xmax'] / width) ymins.append(row['ymin'] / height) ymaxs.append(row['ymax'] / height) classes_text.append(row['class'].encode('utf8')) classes.append(class_text_to_int(row['class'])) #图像所有信息encoded_jpg和object信息整合一起 tf_example = tf.train.Example(features=tf.train.Features(feature={ 'image/height': dataset_util.int64_feature(height), 'image/width': dataset_util.int64_feature(width), 'image/filename': dataset_util.bytes_feature(filename), 'image/source_id': dataset_util.bytes_feature(filename), 'image/encoded': dataset_util.bytes_feature(encoded_jpg), 'image/format': dataset_util.bytes_feature(image_format), 'image/object/bbox/xmin': dataset_util.float_list_feature(xmins), 'image/object/bbox/xmax': dataset_util.float_list_feature(xmaxs), 'image/object/bbox/ymin': dataset_util.float_list_feature(ymins), 'image/object/bbox/ymax': dataset_util.float_list_feature(ymaxs), 'image/object/class/text': dataset_util.bytes_list_feature(classes_text), 'image/object/class/label': dataset_util.int64_list_feature(classes), })) return tf_example def main(_): writer = tf.python_io.TFRecordWriter(FLAGS.output_path) path = os.path.join(os.getcwd(), 'images/test') #一个csv文件最后生成一个tfrecord文件 examples = pd.read_csv(FLAGS.csv_input)//读csv文件内容,返回pandas对象矩阵 """ filename width height class xmin ymin xmax ymax 0 000001.jpg 353 500 dog 43 233 205 362 1 000001.jpg 353 500 person 117 12 296 226 2 000002.jpg 335 500 train 122 188 220 299 """ grouped = split(examples, 'filename') """ [ data(filename='000002.jpg', object= filename width height class xmin ymin xmax ymax 2 000002.jpg 335 500 train 122 188 220 299), #两个1.jpg是因为这张图片里面有两个object data(filename='000001.jpg', object= filename width height class xmin ymin xmax ymax 0 000001.jpg 353 500 dog 43 233 205 362 1 000001.jpg 353 500 person 117 12 296 226) ] """ for group in grouped: tf_example = create_tf_example(group, path)//将每个图片的标注信息和图像信息结合在一起 writer.write(tf_example.SerializeToString()) writer.close() output_path = os.path.join(os.getcwd(), FLAGS.output_path) print('Successfully created the TFRecords: {}'.format(output_path)) if __name__ == '__main__': tf.app.run()
同理还有train_generate_tfrecord.py:
""" Usage: # From tensorflow/models/ # Create train data: python generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record # Create test data: python generate_tfrecord.py --csv_input=data/test_labels.csv --output_path=data/test.record """ from __future__ import division from __future__ import print_function from __future__ import absolute_import import os import io import pandas as pd import tensorflow as tf from PIL import Image from object_detection.utils import dataset_util from collections import namedtuple, OrderedDict flags = tf.app.flags flags.DEFINE_string('csv_input', 'data/train_labels.csv', 'Path to the CSV input') flags.DEFINE_string('output_path', 'data/train.record', 'Path to output TFRecord') FLAGS = flags.FLAGS # TO-DO replace this with label map def class_text_to_int(row_label): if row_label == 'face': return 1 else: 0 def split(df, group): data = namedtuple('data', ['filename', 'object']) gb = df.groupby(group) return [data(filename, gb.get_group(x)) for filename, x in zip(gb.groups.keys(), gb.groups)] def create_tf_example(group, path): with tf.gfile.GFile(os.path.join(path, '{}'.format(group.filename)), 'rb') as fid: encoded_jpg = fid.read() encoded_jpg_io = io.BytesIO(encoded_jpg) image = Image.open(encoded_jpg_io) width, height = image.size filename = group.filename.encode('utf8') image_format = b'jpg' xmins = [] xmaxs = [] ymins = [] ymaxs = [] classes_text = [] classes = [] for index, row in group.object.iterrows(): xmins.append(row['xmin'] / width) xmaxs.append(row['xmax'] / width) ymins.append(row['ymin'] / height) ymaxs.append(row['ymax'] / height) classes_text.append(row['class'].encode('utf8')) classes.append(class_text_to_int(row['class'])) tf_example = tf.train.Example(features=tf.train.Features(feature={ 'image/height': dataset_util.int64_feature(height), 'image/width': dataset_util.int64_feature(width), 'image/filename': dataset_util.bytes_feature(filename), 'image/source_id': dataset_util.bytes_feature(filename), 'image/encoded': dataset_util.bytes_feature(encoded_jpg), 'image/format': dataset_util.bytes_feature(image_format), 'image/object/bbox/xmin': dataset_util.float_list_feature(xmins), 'image/object/bbox/xmax': dataset_util.float_list_feature(xmaxs), 'image/object/bbox/ymin': dataset_util.float_list_feature(ymins), 'image/object/bbox/ymax': dataset_util.float_list_feature(ymaxs), 'image/object/class/text': dataset_util.bytes_list_feature(classes_text), 'image/object/class/label': dataset_util.int64_list_feature(classes), })) return tf_example def main(_): writer = tf.python_io.TFRecordWriter(FLAGS.output_path) path = os.path.join(os.getcwd(), 'images/train') examples = pd.read_csv(FLAGS.csv_input) grouped = split(examples, 'filename') for group in grouped: tf_example = create_tf_example(group, path) writer.write(tf_example.SerializeToString()) writer.close() output_path = os.path.join(os.getcwd(), FLAGS.output_path) print('Successfully created the TFRecords: {}'.format(output_path)) if __name__ == '__main__': tf.app.run()