TF项目实战（SSD目标检测）-VOC2007

TF项目实战（SSD目标检测）-VOC2007

理论详解：https://blog.csdn.net/u013989576/article/details/73439202

训练好的模型和代码会公布在网上（含 VOC数据集  vgg16 模型  以及训练好的模型）：

待续

步骤：

1.代码地址：https://github.com/balancap/SSD-Tensorflow

2.解压ssd_300_vgg.ckpt.zip 到checkpoint文件夹下（另外将vgg16模型放在本路径下）

3.测试一下看看，在notebooks文件夹下创建demo_test.py,其实就是复制ssd_notebook.ipynb中的代码，该py文件是完成对于单张图片的测试。

 import os
 import math
 import random

 import numpy as np
 import tensorflow as tf
 import cv2

 slim = tf.contrib.slim
 import matplotlib.pyplot as plt
 import matplotlib.image as mpimg
 import sys

 sys.path.append('../')
 from nets import ssd_vgg_300, ssd_common, np_methods
 from preprocessing import ssd_vgg_preprocessing
 from notebooks import visualization

 # TensorFlow session: grow memory when needed. TF, DO NOT USE ALL MY GPU MEMORY!!!
 gpu_options = tf.GPUOptions(allow_growth=True)
 config = tf.ConfigProto(log_device_placement=False, gpu_options=gpu_options)
 isess = tf.InteractiveSession(config=config)
 # Input placeholder.
 net_shape = (300, 300)
 data_format = 'NHWC'
 img_input = tf.placeholder(tf.uint8, shape=(None, None, 3))
 # Evaluation pre-processing: resize to SSD net shape.
 image_pre, labels_pre, bboxes_pre, bbox_img = ssd_vgg_preprocessing.preprocess_for_eval(
     img_input, None, None, net_shape, data_format, resize=ssd_vgg_preprocessing.Resize.WARP_RESIZE)
 image_4d = tf.expand_dims(image_pre, 0)

 # Define the SSD model.
 reuse = True if 'ssd_net' in locals() else None
 ssd_net = ssd_vgg_300.SSDNet()
 with slim.arg_scope(ssd_net.arg_scope(data_format=data_format)):
     predictions, localisations, _, _ = ssd_net.net(image_4d, is_training=False, reuse=reuse)

 # Restore SSD model.
 ckpt_filename = '../checkpoints/ssd_300_vgg.ckpt'
 # ckpt_filename = '../checkpoints/VGG_VOC0712_SSD_300x300_ft_iter_120000.ckpt'
 isess.run(tf.global_variables_initializer())
 saver = tf.train.Saver()
 saver.restore(isess, ckpt_filename)

 # SSD default anchor boxes.
 ssd_anchors = ssd_net.anchors(net_shape)

 # Main image processing routine.
 def process_image(img, select_threshold=0.5, nms_threshold=.45, net_shape=(300, 300)):
     # Run SSD network.
     rimg, rpredictions, rlocalisations, rbbox_img = isess.run([image_4d, predictions, localisations, bbox_img],
                                                               feed_dict={img_input: img})

     # Get classes and bboxes from the net outputs.
     rclasses, rscores, rbboxes = np_methods.ssd_bboxes_select(
         rpredictions, rlocalisations, ssd_anchors,
         select_threshold=select_threshold, img_shape=net_shape, num_classes=21, decode=True)

     rbboxes = np_methods.bboxes_clip(rbbox_img, rbboxes)
     rclasses, rscores, rbboxes = np_methods.bboxes_sort(rclasses, rscores, rbboxes, top_k=400)
     rclasses, rscores, rbboxes = np_methods.bboxes_nms(rclasses, rscores, rbboxes, nms_threshold=nms_threshold)
     # Resize bboxes to original image shape. Note: useless for Resize.WARP!
     rbboxes = np_methods.bboxes_resize(rbbox_img, rbboxes)
     return rclasses, rscores, rbboxes

 # Test on some demo image and visualize output.
 # 测试的文件夹
 path = '../demo/'
 image_names = sorted(os.listdir(path))
 # 文件夹中的第几张图，-1代表最后一张
 img = mpimg.imread(path + image_names[-1])
 rclasses, rscores, rbboxes = process_image(img)

 # visualization.bboxes_draw_on_img(img, rclasses, rscores, rbboxes, visualization.colors_plasma)
 visualization.plt_bboxes(img, rclasses, rscores, rbboxes)

4. 将自己的数据集或者 VOC2007直接放在工程目录下

5. 修改datasets文件夹中pascalvoc_common.py文件，将训练类修改别成自己的（这里如果自己的类别）  本文以两类为例子

 VOC_LABELS = {
     'none': (0, 'Background'),
     'aeroplane': (1, 'Vehicle'),
     'bicycle': (2, 'Vehicle'),
     'bird': (3, 'Animal'),
     'boat': (4, 'Vehicle'),
     'bottle': (5, 'Indoor'),
     'bus': (6, 'Vehicle'),
     'car': (7, 'Vehicle'),
     'cat': (8, 'Animal'),
     'chair': (9, 'Indoor'),
     'cow': (10, 'Animal'),
     'diningtable': (11, 'Indoor'),
     'dog': (12, 'Animal'),
     'horse': (13, 'Animal'),
     'motorbike': (14, 'Vehicle'),
     'person': (15, 'Person'),
     'pottedplant': (16, 'Indoor'),
     'sheep': (17, 'Animal'),
     'sofa': (18, 'Indoor'),
     'train': (19, 'Vehicle'),
     'tvmonitor': (20, 'Indoor'),
 }
 #自己的数据
 # VOC_LABELS = {
 #     'none': (0, 'Background'),
 #     'aeroplane': (1, 'Vehicle'),
 # }

6.  将图像数据转换为tfrecods格式，修改datasets文件夹中的pascalvoc_to_tfrecords.py文件，然后更改文件的83行读取方式为’rb‘，如果你的文件不是.jpg格式，也可以修改图片的类型。

另外这个修改

7.运行tf_convert_data.py文件，但是需要传给它一些参数：  这个文件生成TFrecords文件的代码

但是该文件需要像类似于linux 命令那样传入参数。  pycharm中如何解决呢？？？

假设我们需要执行：python ./tf_convert_data.py   --dataset_name=pascalvoc  --dataset_dir=./VOC2007/  --output_name=voc_2007_train  --output_dir=./tfrecords_怎么办呢？

我们可以在  run中的 Edit ... 进入

一、

二、

三、

参数：--dataset_name=pascalvoc --dataset_dir=./VOC2007/ --output_name=voc_2007_train --output_dir=./tfrecords_

然后执行该py文件就ok。

如果出现错误（文件夹相关的错误），则在工程下建立一个文件夹就可以了。

8.训练模型train_ssd_network.py文件中修改

None代表一直训练。

其它需要修改的文件：

① nets/ssd_vgg_300.py  (因为使用此网络结构) ，修改87 和88行的类别

② train_ssd_network.py,修改类别120行，GPU占用量，学习率，batch_size等

③ eval_ssd_network.py 修改类别，66行

④ datasets/pascalvoc_2007.py 根据自己的训练数据修改整个文件

9.开始训练

类似于第7步中的 三

训练的主文件为  train_ssd_network.py

参数为：

--train_dir=./train_model/ --dataset_dir=./tfrecords_/ --dataset_name=pascalvoc_2007 --dataset_split_name=train --model_name=ssd_300_vgg --checkpoint_path=./checkpoints/vgg_16.ckpt --checkpoint_model_scope=vgg_16 --checkpoint_exclude_scopes=ssd_300_vgg/conv6,ssd_300_vgg/conv7,ssd_300_vgg/block8,ssd_300_vgg/block9,ssd_300_vgg/block10,ssd_300_vgg/block11,ssd_300_vgg/block4_box,ssd_300_vgg/block7_box,ssd_300_vgg/block8_box,ssd_300_vgg/block9_box,ssd_300_vgg/block10_box,ssd_300_vgg/block11_box --trainable_scopes=ssd_300_vgg/conv6,ssd_300_vgg/conv7,ssd_300_vgg/block8,ssd_300_vgg/block9,ssd_300_vgg/block10,ssd_300_vgg/block11,ssd_300_vgg/block4_box,ssd_300_vgg/block7_box,ssd_300_vgg/block8_box,ssd_300_vgg/block9_box,ssd_300_vgg/block10_box,ssd_300_vgg/block11_box --save_summaries_secs=60 --save_interval_secs=600 --weight_decay=0.0005 --optimizer=adam --learning_rate=0.001 --learning_rate_decay_factor=0.94 --batch_size=24 --gpu_memory_fraction=0.9

训练过程

：

10 测试：

先看效果

另外我修改了demo_test文件 调取电脑摄像投来执行代码。

如果单独看一张图的效果则执行函数：

代码如下：

 __author__ = "WSX"
 import os
 import math
 import random

 import numpy as np
 import tensorflow as tf
 import cv2

 slim = tf.contrib.slim
 import matplotlib.pyplot as plt
 import matplotlib.image as mpimg
 import sys

 sys.path.append('../')
 from nets import ssd_vgg_300, ssd_common, np_methods
 from preprocessing import ssd_vgg_preprocessing
 from notebooks import visualization

 # TensorFlow session: grow memory when needed. TF, DO NOT USE ALL MY GPU MEMORY!!!
 gpu_options = tf.GPUOptions(allow_growth=True)
 config = tf.ConfigProto(log_device_placement=False, gpu_options=gpu_options)
 isess = tf.InteractiveSession(config=config)
 # Input placeholder.
 net_shape = (300, 300)
 data_format = 'NHWC'
 img_input = tf.placeholder(tf.uint8, shape=(None, None, 3))
 # Evaluation pre-processing: resize to SSD net shape.
 image_pre, labels_pre, bboxes_pre, bbox_img = ssd_vgg_preprocessing.preprocess_for_eval(
     img_input, None, None, net_shape, data_format, resize=ssd_vgg_preprocessing.Resize.WARP_RESIZE)
 image_4d = tf.expand_dims(image_pre, 0)

 # Define the SSD model.
 reuse = True if 'ssd_net' in locals() else None
 ssd_net = ssd_vgg_300.SSDNet()
 with slim.arg_scope(ssd_net.arg_scope(data_format=data_format)):
     predictions, localisations, _, _ = ssd_net.net(image_4d, is_training=False, reuse=reuse)

 # Restore SSD model.
 ckpt_filename = '../checkpoints/ssd_300_vgg.ckpt'
 # ckpt_filename = '../checkpoints/VGG_VOC0712_SSD_300x300_ft_iter_120000.ckpt'
 isess.run(tf.global_variables_initializer())
 saver = tf.train.Saver()
 saver.restore(isess, ckpt_filename)

 # SSD default anchor boxes.
 ssd_anchors = ssd_net.anchors(net_shape)

 # Main image processing routine.
 def process_image(img, select_threshold=0.5, nms_threshold=.45, net_shape=(300, 300)):
     # Run SSD network.
     rimg, rpredictions, rlocalisations, rbbox_img = isess.run([image_4d, predictions, localisations, bbox_img],
                                                               feed_dict={img_input: img})

     # Get classes and bboxes from the net outputs.
     rclasses, rscores, rbboxes = np_methods.ssd_bboxes_select(
         rpredictions, rlocalisations, ssd_anchors,
         select_threshold=select_threshold, img_shape=net_shape, num_classes=21, decode=True)

     rbboxes = np_methods.bboxes_clip(rbbox_img, rbboxes)
     rclasses, rscores, rbboxes = np_methods.bboxes_sort(rclasses, rscores, rbboxes, top_k=400)
     rclasses, rscores, rbboxes = np_methods.bboxes_nms(rclasses, rscores, rbboxes, nms_threshold=nms_threshold)
     # Resize bboxes to original image shape. Note: useless for Resize.WARP!
     rbboxes = np_methods.bboxes_resize(rbbox_img, rbboxes)
     return rclasses, rscores, rbboxes

 #===========================================测试部分===========================================
 #----------------------------------单张图片测试---------------------------
 # Test on some demo image and visualize output.
 # 测试的文件夹
 def demo():
     path = '../demo/'
     image_names = sorted(os.listdir(path))
     # 文件夹中的第几张图，-1代表最后一张
     img = mpimg.imread(path + image_names[-1])
     print(img.shape)
     rclasses, rscores, rbboxes = process_image(img)

     # visualization.bboxes_draw_on_img(img, rclasses, rscores, rbboxes, visualization.colors_plasma)
     visualization.plt_bboxes(img, rclasses, rscores, rbboxes)

 #======================================做成实时显示的代码===================================================
 L = ["None","aeroplane","bicycle","bird ","boat ","bottle ","bus ","car ","cat ","chair","cow ","diningtable","dog","horse","motorbike","person",
      "pottedplant","sheep","sofa","train","tvmonitor"]
 colors_tableau = [(255, 255, 255), (31, 119, 180), (174, 199, 232), (255, 127, 14), (255, 187, 120),
                   (44, 160, 44), (152, 223, 138), (214, 39, 40), (255, 152, 150),
                   (148, 103, 189), (197, 176, 213), (140, 86, 75), (196, 156, 148),
                   (227, 119, 194), (247, 182, 210), (127, 127, 127), (199, 199, 199),
                   (188, 189, 34), (219, 219, 141), (23, 190, 207), (158, 218, 229)]

 def Load_video_show():  #获取视频
     video = cv2.VideoCapture("1.mp4")  # 0 表示摄像头 ，   如果为文件路径则 为加载视频
     while (True):
         ret, frame = video.read()  #frame为 帧  这里当做一张图
         frame = cv2.flip( frame ,1)  #镜像变换，图像正与不正
         cv2.resizeWindow("video", 640, 360)                #设置窗口大小
         frame = cv2.resize(frame, (640, 360))               #设置图大小
         #cv2.imshow("video" ,frame)
         rclasses, rscores, rbboxes = process_image(frame)
         height = frame.shape[0]
         width = frame.shape[1]
         colors = dict()
         for i in range(rclasses.shape[0]):
             cls_id = int(rclasses[i])
             if cls_id >= 0:
                 score = rscores[i]
                 if cls_id not in colors:
                     colors[cls_id] = (random.random(), random.random(), random.random())
                 ymin = int(rbboxes[i, 0] * height)
                 xmin = int(rbboxes[i, 1] * width)
                 ymax = int(rbboxes[i, 2] * height)
                 xmax = int(rbboxes[i, 3] * width)
                 cv2.rectangle(frame, (xmin, ymin), (xmax,ymax), colors[cls_id], 2)  #画矩形框
                 class_name = L[cls_id]
                 cv2.putText(frame, '{:s} | {:.3f}'.format(class_name, score), (xmin, ymin - 2,), cv2.FONT_HERSHEY_COMPLEX, 0.5, (0, 255, 0), 1)  #写文字
         cv2.imshow("video", frame)
         c = cv2.waitKey(50)
         if c == 27: #esc退出
             break

 #Load_video_show()
 demo()