Python을 사용하여 이미지에 대한 모델을 학습하는 방법

import pandas as pd

# 导入图像和标签
image_paths = ['path/to/image1.jpg', 'path/to/image2.jpg', ...]
labels = [0, 1, ...]

# 创建DataFrame
data = pd.DataFrame({'image_path': image_paths, 'label': labels})

import cv2
import numpy as np

# 定义图像大小
image_size = (224, 224)

# 预处理函数
def preprocess_image(image_path):
    # 读取图像
    image = cv2.imread(image_path)
    # 调整大小
    image = cv2.resize(image, image_size)
    # 数据标准化
    image = image.astype(np.float32) / 255.
    # 返回处理后的图像
    return image

# 预处理图像数据
data['image'] = data['image_path'].apply(preprocess_image)

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense

# 定义模型
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(224, 224, 3)))
model.add(MaxPooling2D((2, 2)))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D((2, 2)))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(MaxPooling2D((2, 2)))
model.add(Flatten())
model.add(Dense(64, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

# 编译模型
model.compile(optimizer='adam',
              loss='binary_crossentropy',
              metrics=['accuracy'])

# 训练模型
model.fit(data['image'].to_list(), data['label'].to_list(), epochs=10, batch_size=32)

# 模型评估
test_loss, test_acc = model.evaluate(test_data['image'].to_list(), test_data['label'].to_list())
print('Test Accuracy:', test_acc)

# 模型预测
predictions = model.predict(test_data['image'].to_list())