-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy patheval_batch_samples.py
209 lines (168 loc) · 8.85 KB
/
eval_batch_samples.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Sep 17 08:37:30 2017
@author: llc
"""
#%%
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import datetime as datetime
import time
import math
import tensorflow as tf
import numpy as np
import cv2
import os
import model
index = {"京": 0, "沪": 1, "津": 2, "渝": 3, "冀": 4, "晋": 5, "蒙": 6, "辽": 7, "吉": 8, "黑": 9, "苏": 10, "浙": 11, "皖": 12,
"闽": 13, "赣": 14, "鲁": 15, "豫": 16, "鄂": 17, "湘": 18, "粤": 19, "桂": 20, "琼": 21, "川": 22, "贵": 23, "云": 24,
"藏": 25, "陕": 26, "甘": 27, "青": 28, "宁": 29, "新": 30, "0": 31, "1": 32, "2": 33, "3": 34, "4": 35, "5": 36,
"6": 37, "7": 38, "8": 39, "9": 40, "A": 41, "B": 42, "C": 43, "D": 44, "E": 45, "F": 46, "G": 47, "H": 48,
"J": 49, "K": 50, "L": 51, "M": 52, "N": 53, "P": 54, "Q": 55, "R": 56, "S": 57, "T": 58, "U": 59, "V": 60,
"W": 61, "X": 62, "Y": 63, "Z": 64};
chars = ["京", "沪", "津", "渝", "冀", "晋", "蒙", "辽", "吉", "黑", "苏", "浙", "皖", "闽", "赣", "鲁", "豫", "鄂", "湘", "粤", "桂",
"琼", "川", "贵", "云", "藏", "陕", "甘", "青", "宁", "新", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "A",
"B", "C", "D", "E", "F", "G", "H", "J", "K", "L", "M", "N", "P", "Q", "R", "S", "T", "U", "V", "W", "X",
"Y", "Z"
];
batch_size = 16
capacity = 8
num_examples= 160
FLAGS = tf.app.flags.FLAGS
tf.app.flags.DEFINE_string('eval_log_dir','/home/llc/TF_test/Chinese_plate_recognition/Plate_recognition/eval_log/',
"""Directory where to write event logs.""")
tf.app.flags.DEFINE_string('eval_data_dir','/home/llc/TF_test/Chinese_plate_recognition/Plate_recognition/plate/',
"""eval_data_dir'.""")
tf.app.flags.DEFINE_string('checkpoint_dir','/home/llc/TF_test/Chinese_plate_recognition/Plate_recognition/train_logs_50000/',
"""Directory where to read model checkpoints.""")
#tf.app.flags.DEFINE_integer('num_examples',,"""Number of examples to run.""")
img_test_holder = tf.placeholder(tf.float32,[batch_size,72,272,3])
#keep_prob =tf.placeholder(tf.float32)
def get_input_list():
image_list = [str(l.split(' ',1)[0]) for l in open(os.path.join(FLAGS.eval_data_dir, 'test.txt'),'r')]
label_list = [eval(l.split(' ',1)[-1]) for l in open(os.path.join(FLAGS.eval_data_dir, 'test.txt'),'r')]
## 用eval原因 http://blog.csdn.net/jmilk/article/details/49720611
return image_list,label_list
def get_input_batch(image_list,label_list):
image_list =tf.cast(image_list,tf.string)
label_list = tf.cast(label_list,tf.int32)
input_queue = tf.train.slice_input_producer([image_list,label_list],shuffle=False,num_epochs=1)
label = input_queue[1]
image_contents = tf.read_file(input_queue[0])
image = tf.image.decode_jpeg(image_contents,channels=3)
# image = cv2.imread(input_queue[0])
# image = np.multiply(image,1/255.0)
image = tf.cast(image,tf.float32)
image = tf.multiply(image,1/255)
image_batch,label_batch = tf.train.batch([image,label],batch_size=batch_size,capacity=capacity,shapes=[[72,272,3],[7,]])
#image_batch = tf.cast(image_batch, tf.float32)
return image_batch,label_batch
def eval_once(saver,summary_writer,logits,top_k_op,summary_op):
""" Run Eval once
Args:
saver: Saver
summary_writer: Summary writer
logits: models output
summary_op: Summary op
"""
with tf.Session() as sess:
init_op = tf.local_variables_initializer()
sess.run(init_op)
print('Reading checkpoint .....')
ckpt = tf.train.get_checkpoint_state(FLAGS.checkpoint_dir)
if ckpt and ckpt.model_checkpoint_path:
# Restores from checkpoint
saver.restore(sess, ckpt.model_checkpoint_path)
global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
print('Loading success, global_step is %s' % global_step)
else:
print('No checkpoint file found')
return
#start the queue runners
coord = tf.train.Coordinator()
#threads = tf.train.start_queue_runners(sess=sess,coord=coord)
try:
threads = []
for qr in tf.get_collection(tf.GraphKeys.QUEUE_RUNNERS):
threads.extend(qr.create_threads(sess, coord=coord, daemon=True,
start=True))
num_iter = int(math.ceil(num_examples / batch_size))
true_count = 0 # Counts the number of correct predictions.
total_sample_count = num_iter * batch_size
step = 0
flase_plate_num =[]
while step < num_iter and not coord.should_stop():
prediction,acc = sess.run([logits,top_k_op])
###
acc_reshape = np.reshape(acc,[batch_size,7])
#计算每个batch中识别正确的车牌数
acc_batch = list(map(np.array_equal,acc_reshape,np.ones([batch_size,7],dtype=int)))
true_count += np.sum(acc_batch)
#识别错误的车牌编号
acc_batch_int = list(map(int,acc_batch))
print(acc_batch_int)
first_pos = 0
for i in range(acc_batch_int.count(0)):
new_list = acc_batch_int[first_pos:]
next_pos = new_list.index(0)+1
flase_plate_num.append(step*batch_size+first_pos+new_list.index(0))
first_pos += next_pos
###
for b in range(batch_size):
line = ''
for i in range(prediction.shape[1]):
if i == 0:
result = np.argmax(prediction[b][i][0:31])
if i == 1:
result = np.argmax(prediction[b][i][41:65])+41
if i > 1:
result = np.argmax(prediction[b][i][31:65])+31
line += chars[result]+" "
# print ('batch_num:%d predicted %d: %s' % (step,b,line))
print('Plate_num:%d.jpg,predicted:%s' % (step*batch_size+b,line))
step += 1
# Compute precision.
precision = true_count / total_sample_count
print('%s: precision = %.3f' % (datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S'), precision))
# recognizing flase
print('finding recognize flase plat num: ',flase_plate_num)
# 此部分仿照cifar-10例子，每10秒验证一次，且生成log(tensorboard)
# summary = tf.Summary()
# summary.ParseFromString(sess.run(summary_op))
# summary.value.add(tag='Precision', simple_value=precision)
# summary_writer.add_summary(summary, global_step)
# summary_writer.close()
except Exception as e: # pylint: disable=broad-except
coord.request_stop(e)
coord.request_stop()
coord.join(threads)
def evaluate():
"""Eval plate for a number of steps"""
# g = tf.get_default_graph()
#with g.as_default() :
with tf.Graph().as_default() as g:
#get images and labels
#Build a Graph that computes the logits predictions from the inference model
image_list,label_list = get_input_list()
image_batch,label_batch= get_input_batch(image_list,label_list)
logits1,logits2,logits3,logits4,logits5,logits6,logits7 = model.inference(image_batch,1) #output=[7,b,65]
logits = tf.stack([logits1,logits2,logits3,logits4,logits5,logits6,logits7],axis=1) #output=[b,7,65]
logits_label= tf.reshape(tf.nn.softmax(logits),[-1,65]) #output=[b*7,65]
label_batch = tf.reshape(label_batch,[-1]) # input:[b,7] , output:[b*7,1]
# logit = model.inference(image_batch,keep_prob) #output=[7,b,65]
# logits = tf.reshape(tf.nn.softmax(logit),[-1,65]) #output=[7*b,65]
# label_batch = tf.reshape(tf.transpose(label_batch),[-1]) # input:[b,7] , output:[7*b,1]
top_k_op = tf.nn.in_top_k(logits_label,label_batch,1)
summary_op = tf.summary.merge_all()
summary_writer = tf.summary.FileWriter(FLAGS.eval_log_dir,g)
saver = tf.train.Saver()
eval_once(saver,summary_writer,logits,top_k_op,summary_op)
def main(argv=None):
if tf.gfile.Exists(FLAGS.eval_log_dir):
tf.gfile.DeleteRecursively(FLAGS.eval_log_dir)
tf.gfile.MakeDirs(FLAGS.eval_log_dir)
evaluate()
if __name__=='__main__':
tf.app.run()