ssd_mobilenet_demo

操作系统：windows 10  64位

内存：8G

GPU：Nvidia MX 150

Tensorflow: 1.4

1、安装python

Anaconda3-5.0.1 ，默认python版本（3.6.3）

2、安装tensorflow

pip install --upgrade tensorflow

conda install pip #更新pip

pip install --upgrade tensorflow-gpu

CUDA® Toolkit 8.0， 需要注意最新版9.1不支持tensorflow 1.4版本；

cuDNN v6.0，7.0不支持tensorflow 1.4版本，现在cuDNN需要先注册成为NVIDIA开发者，下载后将cuDNN中对应文件夹下的.dll文件分别复制到CUDA安装目录对应文件夹下；

对应的显卡驱动，如果驱动较新，在安装CUDA的时候会有提示可能不兼容，可以无视。

3、测试gpu版安装好了没有

improt tensorflow as tf
hello = tf.constant('hello')
sess = tf.Session()
print(sess.run(hello))

当输出hello则装好tensorflow

from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())

当输出：
Sample Output
[name: "/cpu:0" device_type: "CPU" memory_limit:
name: "/gpu:0" device_type: "GPU" .............GeForce GTX 1070

4、下载api

https://github.com/tensorflow/models

5、protobuf配置

https://github.com/google/protobuf/releases  网站中选择windows 版本（最下面），解压后将bin文件夹中的【protoc.exe】放到C:\Windows

在models\research\目录下打开命令行窗口，输入：

# From tensorflow/models/
protoc object_detection/protos/*.proto --python_out=.

在这一步有时候会出错，可以尝试把/*.proto 这部分改成文件夹下具体的文件名，一个一个试，每运行一个，文件夹下应该出现对应的.py结尾的文件。不报错即可。

6、环境变量

models/research/ 及 models/research/slim 添加进环境变量

7、测试环境

python object_detection/builders/model_builder_test.py

注意 ：如果出现no model name object_api这个东东，就在D:\anaconda\anaconda3.4.2.0\Lib\site-packages目录下面新建一个my_objection.pth文件，文件内容就是这两个路径，如下图：

8、object_detection_tutorial.ipynb

代码简化了一些。

# coding: utf-8
import numpy as np
import os
import six.moves.urllib as urllib
import sys
import tarfile
import tensorflow as tf
import zipfile
from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image

if tf.__version__ < '1.4.0':
  raise ImportError('Please upgrade your tensorflow installation to v1.4.* or later!')

# This is needed to display the images.
get_ipython().magic('matplotlib inline')

# This is needed since the notebook is stored in the object_detection folder.
sys.path.append("..")

from utils import label_map_util
from utils import visualization_utils as vis_util

# 下载模型名，设置对应的参数
MODEL_NAME = 'ssd_mobilenet_v1_coco_2017_11_17'
MODEL_FILE = MODEL_NAME + '.tar.gz'
DOWNLOAD_BASE = 'http://download.tensorflow.org/models/object_detection/'

# 训练好的模型，用来检测
PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'

# 标签文件，记录了哪些标签需要识别
PATH_TO_LABELS = os.path.join('data', 'mscoco_label_map.pbtxt')

# 类别数目，根据实际修改
NUM_CLASSES = 90

# ## 下载上面说的模型（不用改）
opener = urllib.request.URLopener()
opener.retrieve(DOWNLOAD_BASE + MODEL_FILE, MODEL_FILE)
tar_file = tarfile.open(MODEL_FILE)
for file in tar_file.getmembers():
  file_name = os.path.basename(file.name)
  if 'frozen_inference_graph.pb' in file_name:
    tar_file.extract(file, os.getcwd())

#将训练完的载入内存（不用改）
detection_graph = tf.Graph()
with detection_graph.as_default():
  od_graph_def = tf.GraphDef()
  with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
    serialized_graph = fid.read()
    od_graph_def.ParseFromString(serialized_graph)
    tf.import_graph_def(od_graph_def, name='')

# ## 载入标签map（不用改）
# Label maps map indices to category names, so that when our convolution network predicts `5`,
we know that this corresponds to `airplane`.
Here we use internal utility functions, but anything that returns a dictionary mapping integers to appropriate string labels would be fine

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

def load_image_into_numpy_array(image):
  (im_width, im_height) = image.size
  return np.array(image.getdata()).reshape(
      (im_height, im_width, 3)).astype(np.uint8)

"""
检测部分
"""
# 目标文件夹名
PATH_TO_TEST_IMAGES_DIR = 'test_images'
# 源码中test_images文件夹下就两张image，名字分别为image1.jpg和image2.jpg
# 如果想用自己的image，有5张图片，分别为hello1.jpg.....hello5.jpg可以改成：
# TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'hello{}.jpg'.format(i)) for i in range(1, 6) ]
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 3) ]

# 设置输出图像的英尺
IMAGE_SIZE = (12, 8)

#运行，进行检测
with detection_graph.as_default():
  with tf.Session(graph=detection_graph) as sess:
    # Definite input and output Tensors for detection_graph
    image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
    # Each box represents a part of the image where a particular object was detected.
    detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
    # Each score represent how level of confidence for each of the objects.
    # Score is shown on the result image, together with the class label.
    detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
    detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
    num_detections = detection_graph.get_tensor_by_name('num_detections:0')
    for image_path in TEST_IMAGE_PATHS:
      image = Image.open(image_path)
      # the array based representation of the image will be used later in order to prepare the
      # result image with boxes and labels on it.
      image_np = load_image_into_numpy_array(image)
      # Expand dimensions since the model expects images to have shape: [1, None, None, 3]
      image_np_expanded = np.expand_dims(image_np, axis=0)
      # Actual detection.
      (boxes, scores, classes, num) = sess.run(
          [detection_boxes, detection_scores, detection_classes, num_detections],
          feed_dict={image_tensor: image_np_expanded})
      # Visualization of the results of a detection.
      vis_util.visualize_boxes_and_labels_on_image_array(
          image_np,
          np.squeeze(boxes),
          np.squeeze(classes).astype(np.int32),
          np.squeeze(scores),
          category_index,
          use_normalized_coordinates=True,
          line_thickness=8)
      plt.figure(figsize=IMAGE_SIZE)
      plt.imshow(image_np)