SSD-Tensorflow训练总结

感想

今天我测试了一下我自己训练的模型，和YOLOv2做了一下对比，检测的都是对的，YOLOv2版本的准确率不高，但是SSD有很多没有检测出来，召回率不怎么高。

注意，ssd的环境是python3，在python2上跑会有问题。tensorflow-gpu, opencv的安装参考我的博客： SSD环境安装

1 制作数据集

最麻烦的是制作voc数据集，我这里用了公司的数据集生成器产生了很多张图片，总量大概有25000张左右。按照voc格式，把图片放在

JPEGImages目录下，xml格式的文件放在Annotations目录下，然后利用程序生成train.txt, test.txt, trainval.txt, val.txt四个文件就够了。生成这些txt的代码如下：

import os
import random 

xmlfilepath=r'/home/whsyxt/Downloads/SSD-Tensorflow/VOC2007/Annotations'
saveBasePath=r"/home/whsyxt/Downloads/SSD-Tensorflow"

trainval_percent=0.8
train_percent=0.7
total_xml = os.listdir(xmlfilepath)
num=len(total_xml)  
list=range(num)  
tv=int(num*trainval_percent)  
tr=int(tv*train_percent)  
trainval= random.sample(list,tv)  
train=random.sample(trainval,tr)  

print("train and val size",tv)
print("traub suze",tr)
ftrainval = open(os.path.join(saveBasePath,'VOC2007/ImageSets/Main/trainval.txt'), 'w')  
ftest = open(os.path.join(saveBasePath,'VOC2007/ImageSets/Main/test.txt'), 'w')  
ftrain = open(os.path.join(saveBasePath,'VOC2007/ImageSets/Main/train.txt'), 'w')  
fval = open(os.path.join(saveBasePath,'VOC2007/ImageSets/Main/val.txt'), 'w')  

for i  in list:  
    name=total_xml[i][:-4]+'\n'  
    if i in trainval:  
        ftrainval.write(name)  
        if i in train:  
            ftrain.write(name)  
        else:  
            fval.write(name)  
    else:  
        ftest.write(name)  
  
ftrainval.close()  
ftrain.close()  
fval.close()  
ftest .close()  

读者可以按照自己的方式去改。

2 voc转tfrecords

voc格式的数据集制作好以后，我们需要把数据集转换成tfrecords，这样程序才能跑，首先，我们需要修改一下源码，datasets\pascalvoc_common.py，操作也非常简单，你把你的类别填上就行了，其他的都不用管，看我的示例，我把原来的16类弄成了3类：

"""
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'),
}
"""

VOC_LABELS = {
    'none': (0, 'Background'),
    'person': (1, 'Person'),
    'car': (2, 'Car'),
}

这样就行了。

接着跳转到SSD-tensorflow目录下，进行tfrecords操作，我的运行命令如下：

DATASET_DIR=VOC2007/
OUTPUT_DIR=tfrecords/
python3 tf_convert_data.py \
    --dataset_name=pascalvoc \
    --dataset_dir=${DATASET_DIR} \
    --output_name=voc_2007_train \
    --output_dir=${OUTPUT_DIR}

3 训练

这样就可以进行训练了，运行的命令为：

DATASET_DIR=tfrecords
TRAIN_DIR=logs/
CHECKPOINT_PATH=./checkpoints/ssd_300_vgg.ckpt
python3 train_ssd_network.py \
    --train_dir=${TRAIN_DIR} \
    --dataset_dir=${DATASET_DIR} \
    --dataset_name=pascalvoc_2007 \
    --dataset_split_name=train \
    --model_name=ssd_300_vgg \
    --checkpoint_path=${CHECKPOINT_PATH} \
    --save_summaries_secs=60 \
    --save_interval_secs=600 \
    --weight_decay=0.0005 \
    --optimizer=adam \
    --learning_rate=0.001 \
    --batch_size=16

4 预测

我主要是跑视频，我把我跑视频的预测代码和运行命令也提供给大家参考一下：

命令：

 python3 video_demo.py 

代码：

#coding=utf-8

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 = 'finetune_log/model.ckpt-41278'  //修改为你的模型路径
#ckpt_filename = 'checkpoints/ssd_300_vgg.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

def bboxes_draw_on_img(img, classes, scores, bboxes, color=[255, 0, 0], thickness=2):
    shape = img.shape
    for i in range(bboxes.shape[0]):
        bbox = bboxes[i]
        #color = colors[classes[i]]
        # Draw bounding box...
        p1 = (int(bbox[0] * shape[0]), int(bbox[1] * shape[1]))
        p2 = (int(bbox[2] * shape[0]), int(bbox[3] * shape[1]))
        cv2.rectangle(img, p1[::-1], p2[::-1], color, thickness)
        # Draw text...
        s = '%s/%.3f' % (classes[i], scores[i])
        p1 = (p1[0]-5, p1[1])
        cv2.putText(img, s, p1[::-1], cv2.FONT_HERSHEY_DUPLEX, 0.4, color, 1)

cap = cv2.VideoCapture("DJI_0008.MOV") //修改为你的路径
#cap = cv2.VideoCapture(0)

# Define the codec and create VideoWriter object
#fourcc = cv2.cv.FOURCC(*'XVID')
fourcc = cv2.VideoWriter_fourcc(*'XVID') 
out = cv2.VideoWriter('output1.avi', fourcc, 20, (1280, 720))



num=0

while cap.isOpened():
    # get a frame
    rval, frame = cap.read()
    # save a frame
    if rval==True:
      #  frame = cv2.flip(frame,0)
        rclasses, rscores, rbboxes=process_image(frame)
        bboxes_draw_on_img(frame,rclasses,rscores,rbboxes)
        print(rclasses)
        out.write(frame)
        num=num+1
        print(num)
    else:
        break
    # show a frame
    cv2.imshow("capture", frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
cap.release()
out.release()
cv2.destroyAllWindows()

参考文献

[1] SSD-Tensorflow. https://github.com/balancap/SSD-Tensorflow