TensorFlow 2训练模型加载测试报错：protobuf解码失败求助

DevOps 小满

2026-8-20

解决TensorFlow 2自定义目标检测模型测试时的protobuf解码错误

我已完成基于TensorFlow 2的自定义目标检测模型训练，测试时修改了脚本路径配置但仍无法运行，报错信息如下：

Traceback (most recent call last):
  File "d:\VSCode\Machine_Learning_Tests\Tensorflow\test\object_detection_tutorial_wwwPythonProgrammingNet__mitBild.py", line 65, in <module>
    od_graph_def.ParseFromString(serialized_graph)
google.protobuf.message.DecodeError: Error parsing message with type 'tensorflow.GraphDef'

测试脚本如下：

import numpy as np
import os
import six as urllib    # import six.moves.urllib as urllib
import sys
import tarfile
import tensorflow as tf
import zipfile

from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image

import cv2
cap = cv2.VideoCapture(1)

# This is needed since the notebook is stored in the object_detection folder.
sys.path.append("..")


# ## Object detection imports
# Here are the imports from the object detection module.

from object_detection.utils import label_map_util         # from utils import label_map_util

from object_detection.utils import visualization_utils as vis_util      # from utils import visualization_utils as vis_util


# # Model preparation 

# ## Variables
# 
# Any model exported using the `export_inference_graph.py` tool can be loaded here simply by changing `PATH_TO_CKPT` to point to a new .pb file.  
# 
# By default we use an "SSD with Mobilenet" model here.

# What model to download.
MODEL_NAME = 'D:/VSCode/Machine_Learning_Tests/Tensorflow/workspace/exported_models/first_model/saved_model'      # MODEL_NAME = 'inference_graph'

# Path to frozen detection graph. This is the actual model that is used for the object detection.
PATH_TO_CKPT = MODEL_NAME + '/saved_model.pb'                                                                     # PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'

# List of the strings that is used to add correct label for each box.
PATH_TO_LABELS = 'D:/VSCode/Machine_Learning_Tests/Tensorflow/workspace/data/label_map.pbtxt'                     # PATH_TO_LABELS = 'training/labelmap.pbtxt'

NUM_CLASSES = 1




# ## Load a (frozen) Tensorflow model into memory.

detection_graph = tf.Graph()
with detection_graph.as_default():
  od_graph_def = tf.compat.v1.GraphDef()                            #   od_graph_def = tf.GraphDef()
  with tf.compat.v2.io.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:      #   with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
    serialized_graph = fid.read()
    od_graph_def.ParseFromString(serialized_graph)
    tf.import_graph_def(od_graph_def, name='')


# ## Loading label map
# Label maps map indices to category names, so that when our convolution network predicts `5`, we know that this corresponds to `airplane`.  Here we use internal utility functions, but anything that returns a dictionary mapping integers to appropriate string labels would be fine

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)


# ## Helper code

def load_image_into_numpy_array(image):
  (im_width, im_height) = image.size
  return np.array(image.getdata()).reshape(
      (im_height, im_width, 3)).astype(np.uint8)


# # Detection

# For the sake of simplicity we will use only 2 images:
# image1.jpg
# image2.jpg
# If you want to test the code with your images, just add path to the images to the TEST_IMAGE_PATHS.
PATH_TO_TEST_IMAGES_DIR = 'images/test/'
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(3, 8) ]

# Size, in inches, of the output images.
IMAGE_SIZE = (12,8)


with detection_graph.as_default():
  with tf.Session(graph=detection_graph) as sess:
    while True:
      image_np = np.array(cv2.imread('Test.jpg'))
      cv2.imshow('image',image_np)
      cv2.waitKey(1)
      # Expand dimensions since the model expects images to have shape: [1, None, None, 3]
      image_np_expanded = np.expand_dims(image_np, axis=0)
      image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
      # Each box represents a part of the image where a particular object was detected.
      boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
      # Each score represent how level of confidence for each of the objects.
      # Score is shown on the result image, together with the class label.
      scores = detection_graph.get_tensor_by_name('detection_scores:0')
      classes = detection_graph.get_tensor_by_name('detection_classes:0')
      num_detections = detection_graph.get_tensor_by_name('num_detections:0')
      # Actual detection.
      (boxes, scores, classes, num_detections) = sess.run(
          [boxes, scores, classes, num_detections],
          feed_dict={image_tensor: image_np_expanded})
      # Visualization of the results of a detection.
      vis_util.visualize_boxes_and_labels_on_image_array(
          image_np,
          np.squeeze(boxes),
          np.squeeze(classes).astype(np.int32),
          np.squeeze(scores),
          category_index,
          use_normalized_coordinates=True,
          line_thickness=8)

      cv2.imshow('object detection', cv2.resize(image_np, (800,600)))
      if cv2.waitKey(25) & 0xFF == ord('q'):
        cv2.destroyAllWindows()
        break

错误原因

你使用的是TensorFlow 2导出的SavedModel格式，但测试脚本沿用了TensorFlow 1.x加载冻结图（frozen graph）的方式，直接解析saved_model.pb会导致protobuf解码失败——因为saved_model.pb包含的是SavedModel的完整定义，并非单纯的GraphDef结构。

解决步骤

1. 修正模型路径

确保MODEL_NAME指向SavedModel的根目录（即包含saved_model.pb和variables文件夹的目录），无需指定具体的pb文件：

MODEL_NAME = 'D:/VSCode/Machine_Learning_Tests/Tensorflow/workspace/exported_models/first_model/saved_model'

2. 替换模型加载与推理代码

删除原有的detection_graph和tf.Session相关代码，改用TensorFlow 2的即时执行模式加载模型并推理：

修改后的完整关键代码如下：

import numpy as np
import os
import sys
import tensorflow as tf
import cv2

from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util

# 添加路径
sys.path.append("..")

# 模型配置
MODEL_NAME = 'D:/VSCode/Machine_Learning_Tests/Tensorflow/workspace/exported_models/first_model/saved_model'
PATH_TO_LABELS = 'D:/VSCode/Machine_Learning_Tests/Tensorflow/workspace/data/label_map.pbtxt'
NUM_CLASSES = 1

# 加载标签映射
label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

# 加载TensorFlow 2 SavedModel
detect_fn = tf.saved_model.load(MODEL_NAME)

# 检测循环
while True:
    # 读取测试图片（可替换为摄像头帧）
    image_np = cv2.imread('Test.jpg')
    if image_np is None:
        print("无法读取测试图片")
        break
    
    # 转换为模型输入格式
    input_tensor = tf.convert_to_tensor(image_np)
    input_tensor = input_tensor[tf.newaxis, ...]

    # 执行检测
    detections = detect_fn(input_tensor)

    # 处理检测结果
    num_detections = int(detections.pop('num_detections'))
    detections = {key: value[0, :num_detections].numpy()
                  for key, value in detections.items()}
    detections['num_detections'] = num_detections
    detections['detection_classes'] = detections['detection_classes'].astype(np.int32)

    # 可视化检测结果
    vis_util.visualize_boxes_and_labels_on_image_array(
        image_np,
        detections['detection_boxes'],
        detections['detection_classes'],
        detections['detection_scores'],
        category_index,
        use_normalized_coordinates=True,
        line_thickness=8)

    # 显示结果
    cv2.imshow('object detection', cv2.resize(image_np, (800, 600)))
    if cv2.waitKey(25) & 0xFF == ord('q'):
        cv2.destroyAllWindows()
        break

3. 额外注意事项

如果你想用摄像头实时检测，将cv2.imread('Test.jpg')替换为ret, image_np = cap.read()，并确保初始化cap = cv2.VideoCapture(0)（根据摄像头编号调整）。
确保TensorFlow版本与训练时一致，避免版本兼容问题。

内容的提问来源于stack exchange，提问作者pypypy123