Faster Rcnn 代码解读之 pascal_voc.py

# --------------------------------------------------------
# Fast R-CNN
# Copyright (c) 2015 Microsoft
# Licensed under The MIT License [see LICENSE for details]
# Written by Ross Girshick and Xinlei Chen
# --------------------------------------------------------
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import os
from datasets.imdb import imdb
import datasets.ds_utils as ds_utils
import xml.etree.ElementTree as ET
import numpy as np
import scipy.sparse
import scipy.io as sio
import utils.cython_bbox
import pickle
import subprocess
import uuid
from .voc_eval import voc_eval
from model.config import cfg


# pascal_voc继承imdb
class pascal_voc(imdb):
    # image_set代表数据集名称（train，val，test...）,year代表版本如2007
    def __init__(self, image_set, year, use_diff=False):
        name = 'voc_' + year + '_' + image_set  # 例如voc_2007_train
        if use_diff:
            name += '_diff'
        # 调用imdb的构造函数
        imdb.__init__(self, name)
        self._year = year
        self._image_set = image_set
        # /home/liubo/tf-faster-rcnn1/data/VOCdevkit2007/
        self._devkit_path = self._get_default_path()
        # /home/liubo/tf-faster-rcnn1/data/VOCdevkit/VOC2007/
        self._data_path = os.path.join(self._devkit_path, 'VOC' + self._year)
        self._classes = ('__background__',  # always index 0
                         'tilt', 'miss', 'normal')
        # _class_to_ind保存的是{'__background__' ： 0 , 'tilt' ： 1 , 'miss' : 2 , 'normal' : 3}
        self._class_to_ind = dict(list(zip(self.classes, list(range(self.num_classes)))))
        # 图片的格式
        self._image_ext = '.JPG'
        # 里面包含所有数据集图像的名称，如[000001,000002,000003...]
        self._image_index = self._load_image_set_index()
        # Default to roidb handler
        # 得到roi图片信息,重载imdb中
        self._roidb_handler = self.gt_roidb
        # 定义一个随机的标识符
        self._salt = str(uuid.uuid4())  # 4a5ae4d0-03ff-490c-871b-33b615a19bb4
        self._comp_id = 'comp4'

        # PASCAL specific config options
        self.config = {'cleanup': True,
                       'use_salt': True,
                       'use_diff': use_diff,
                       'matlab_eval': False,
                       'rpn_file': None}

        assert os.path.exists(self._devkit_path), \
            'VOCdevkit path does not exist: {}'.format(self._devkit_path)
        assert os.path.exists(self._data_path), \
            'Path does not exist: {}'.format(self._data_path)

    # 返回图片所在的路径
    def image_path_at(self, i):
        """
        Return the absolute path to image i in the image sequence.
        """
        # /home/liubo/tf-faster-rcnn1/data/VOCdevkit/VOC2007/JPEGImages/000001.JPG
        return self.image_path_from_index(self._image_index[i])

    def image_path_from_index(self, index):
        """
        Construct an image path from the image's "index" identifier.
        """
        # _data_path = /home/liubo/tf-faster-rcnn1/data/VOCdevkit/VOC2007/
        image_path = os.path.join(self._data_path, 'JPEGImages',
                                  index + self._image_ext)  # _image_ext = .JPG 后缀名
        assert os.path.exists(image_path), \
            'Path does not exist: {}'.format(image_path)
        return image_path

    # 获取图片索引
    def _load_image_set_