Tensorflow SSD Tutorial

#https://github.com/balancap/SSD-Tensorflow( Tutorial)
# http://blog.csdn.net/jnulzl/article/details/68947496#t2 (change to python format)
importos
import path
import random

import numpy as np
import tensorflow as tf
import cv2

slim = tf.contrib.slim

#%matplotlib inline
import matplotlib.pyplot asplt
import matplotlib.image asmpimg

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, localizations,_,_=ssd_net.net(image_4d, is_training=False, reuse=reuse)

# Restore SSD model.Model file
ckpt_filename = 'C:\Ana3\SSD\checkpoints\ssd_300_vgg\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, rlocalizations, rbox_img = isess.run([image_4d, predictions, localizations, bbox_img],
                                                          feed_dict={img_input:img})

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

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

# Test on some demo image and visualize output.
path='C:\Ana3\SSD\demo'
image_names=sorted(os.listdir(path))

img=mpimg.imread (path+image_names[-5])
for it in image_names:
    img=mpimg.imread(path+it)
    rclasses, rscores, rboxes=process_image(img)

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