Google Colab读取Drive图像报错：未检测到图像与类别

在Google Colab中运行图像分类代码时，执行model.fit(train_generator, epochs=10, validation_data=test_generator)前持续出现如下错误：

Found 0 images belonging to 0 classes.
Found 0 images belonging to 0 classes.

已调整batch尺寸为10、将所有图像resize至100×100，训练/测试图像分别存入training_data和testing_data文件夹，但问题未解决。相关代码如下：

!pip install tensorflow
import cv2
from tensorflow.keras.preprocessing.image import img_to_array
import numpy as np
import tensorflow as tf
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.applications import EfficientNetB0
from tensorflow.keras.layers import Dense, GlobalAveragePooling2D
from tensorflow.keras.models import Model
from sklearn.metrics import confusion_matrix, accuracy_score, precision_score, recall_score

#uploads the image
def load_image(image_location):
    # Load the image using OpenCV
    image = cv2.imread(image_location)

   #checks if image is uploaded or not
    if image is None:
        print("Error: image not found from", image_location)
        return None

    resized_image= cv2.resize(image, (100, 100))
    image_array = img_to_array(resized_image)

    image_array /= 255.0

    return image_array


example_image_location = '/content/drive/MyDrive/Machine Learning Folder/training_data/LightGreen_Crayon_Testing.jpg'

# Load and preprocess the example image
example_image = load_image(example_image_location)


if example_image is not None:

    # collecting/processing data
  train_dir = '/content/drive/MyDrive/Machine Learning Folder/training_data'
  test_dir = '/content/drive/MyDrive/Machine Learning Folder/testing_data'

  train_datagen = ImageDataGenerator(rescale=1. / 255)
  test_datagen = ImageDataGenerator(rescale=1. / 255)

  train_generator = train_datagen.flow_from_directory(train_dir, target_size=(100, 100), batch_size=10, class_mode='binary')

  test_generator = test_datagen.flow_from_directory(test_dir, target_size=(100, 100), batch_size=10, class_mode='binary')

    # model building
  base_model = EfficientNetB0(weights='imagenet', include_top=False)
  x = base_model.output
  x = GlobalAveragePooling2D()(x)
  x = Dense(1024, activation='relu')(x)
  predictions = Dense(1, activation='sigmoid')(x)
  model = Model(inputs=base_model.input, outputs=predictions)

  for layer in base_model.layers:
    layer.trainable = False
    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
    model.fit(train_generator, epochs=10, validation_data=test_generator)

1. 修复目录结构（核心问题） #

flow_from_directory要求每个类别必须存放在单独的子文件夹中，不能直接将图像放在training_data或testing_data的根目录下。正确的目录结构应为：

Machine Learning Folder/
├── training_data/
│   ├── LightGreen_Crayon/  # 类别1文件夹
│   │   ├── LightGreen_Crayon_Testing.jpg
│   │   └── 其他同类别图像...
│   └── 其他类别文件夹/  # 比如DarkGreen_Crayon
│       └── ...
└── testing_data/
    ├── LightGreen_Crayon/
    └── 其他类别文件夹/

原目录将图像直接放在根目录，导致flow_from_directory无法识别类别和图像，因此输出"Found 0 images belonging to 0 classes"。

2. 修正代码缩进错误 #

代码中model.compile()和model.fit()被错误嵌套在for layer in base_model.layers:循环内部，会导致每遍历一个模型层就编译一次模型、触发一次训练，逻辑完全错误。需将这两行移至循环外部：

for layer in base_model.layers:
    layer.trainable = False
# 编译与训练代码移至循环外
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.fit(train_generator, epochs=10, validation_data=test_generator)

3. 删除冗余代码 #

自定义的load_image函数及示例图像加载逻辑属于冗余操作——ImageDataGenerator+flow_from_directory已自动完成图像加载、resize和归一化，可直接删除这部分代码，避免混淆。

4. 验证路径有效性 #

可添加代码确认目录路径正确且包含内容：

import os
# 打印训练目录下的内容，确认存在类别子文件夹
print("训练目录内容:", os.listdir(train_dir))
print("测试目录内容:", os.listdir(test_dir))

若输出为空或无类别子文件夹，需检查路径拼写或目录结构是否正确。

内容的提问来源于stack exchange，提问作者zaelyn