slim微调后的模型导出，对单张图片进行测试

这是slim微调之后的模型导出为pb文件后，值可以对单张图片进行测试的相关代码(单张，而非多张),如果多张只是使用for循化输入多张图片的话，会慢慢引起内存爆炸问题
多张图片测试：https://blog.csdn.net/ShaoDu/article/details/88781449

import tensorflow as tf
import numpy as np
from classify_image import NodeLookup
import os
def init_graph(model_name='/home/lf/桌面/data/model.ckpt-100000.pb'):
  with open(model_name, 'rb') as f:
    graph_def = tf.GraphDef()
    graph_def.ParseFromString(f.read())
    _ = tf.import_graph_def(graph_def, name='')
 
def run_inference_on_image(file_name,node_lookup,sess):
  image_data = open(file_name, 'rb').read()
  softmax_tensor = sess.graph.get_tensor_by_name('resnet_v1_50/predictions/Softmax:0')
  predictions = sess.run(softmax_tensor,
                           {'input:0': image_data})
  predictions = np.squeeze(predictions)
 
  # Creates node ID --> English string lookup.
  node_lookup = node_lookup
  top_k = predictions.argsort()[-12:][::-1]
  top_names = []
  #for node_id in top_k:
  #  human_string = node_lookup.id_to_string(node_id)
  #  top_names.append(human_string)
  #  score = predictions[node_id]
  #  print('id:[%d] name:[%s] (score = %.5f)' % (node_id, human_string, score))
  #return predictions, top_k, top_names
  print(file_name,node_lookup.id_to_string(top_k[0]))
  return node_lookup.id_to_string(top_k[0])
 
file_name = '/home/lf/桌面/Black_Footed_Albatross_0001_796111.jpg'
#test_dir = '/home/gcnan604/devdata/hdwei/TFExamples/plantSeedlings/test'
 
#label_file, _ = os.path.splitext('my_freeze.pb')
#label_file = label_file + '.label'
label_file='/home/lf/桌面/data2/label.txt'
node_lookup = NodeLookup(label_file)
sess = tf.Session()
init_graph()
with open('submission.scv','a+') as f:
  for test_file in os.listdir(file_name):
    tempfile = os.path.join(file_name,test_file)
    predicetion = run_inference_on_image(tempfile,node_lookup,sess)
    f.write(test_file+','+predicetion+'\n')