model_watermark_generate/tests/deal_classify_image_test.py

"""
图像分类数据集黑盒水印嵌入测试
"""
import os
import random
import time

import cv2
import numpy as np
import qrcode

from watermark_generate.tools import secret_label_func, general_tool


def generate_watermark_indices(dataset_dir, num_parts, percentage=0.05):
    watermark_splits = []
    # 初始化每个切分的图像索引
    for _ in range(num_parts):
        watermark_splits.append({})

    # 遍历分类文件夹
    for class_name in os.listdir(dataset_dir):
        class_dir = os.path.join(dataset_dir, class_name)

        if os.path.isdir(class_dir):
            images = os.listdir(class_dir)
            num_images = len(images)
            num_watermark = int(num_images * percentage)

            # 获取所有图像的索引
            image_indices = list(range(num_images))

            # 确保每个切分的图像不重复
            if len(image_indices) >= num_parts * num_watermark:
                for i in range(num_parts):
                    start_idx = i * num_watermark
                    end_idx = start_idx + num_watermark
                    # 顺序选择索引范围内的图像
                    selected_indices = image_indices[start_idx:end_idx]
                    # 将索引转换为文件名
                    selected_images = [images[idx] for idx in selected_indices]
                    selected_images = [os.path.join(class_dir, filename) for filename in selected_images]
                    watermark_splits[i][class_name] = selected_images
            else:
                print(f"分类 {class_name} 中的图像不足以生成 {num_parts} 个不重复的切分。")

    return watermark_splits


def find_index_in_parts(select_image_parts, filename):
    for index, select_images in enumerate(select_image_parts):
        for cls_index, list in enumerate(select_images.values()):
            if filename in list:
                return True, index, cls_index
    return False, None, None


def add_watermark_to_image(img, watermark_label, watermark_class_id):
    """
    Adds a QR code watermark to the image based on the given label and returns the updated label information.

    Args:
        img (numpy.ndarray): The original image.
        watermark_label (str): The text label to encode into the QR code.
        watermark_class_id (int): The class ID for the watermark.

    Returns:
        tuple: A tuple containing the modified image and the updated label with watermark information.
    """
    # Generate the QR code for the watermark label
    qr = qrcode.QRCode(
        version=1,
        error_correction=qrcode.constants.ERROR_CORRECT_L,
        box_size=2,
        border=1
    )
    qr.add_data(watermark_label)
    qr.make(fit=True)
    qr_img = qr.make_image(fill='black', back_color='white').convert('RGB')

    # Convert the PIL image to a NumPy array without resizing
    qr_img = np.array(qr_img)

    # Image and QR code sizes
    img_h, img_w = img.shape[:2]
    qr_h, qr_w = qr_img.shape[:2]

    # Calculate random position ensuring QR code stays within image bounds
    max_x = img_w - qr_w
    max_y = img_h - qr_h

    if max_x < 0 or max_y < 0:
        raise ValueError("QR code size exceeds image dimensions.")

    x_start = random.randint(0, max_x)
    y_start = random.randint(0, max_y)
    x_end = x_start + qr_w
    y_end = y_start + qr_h

    # Crop the QR code if it exceeds image boundaries (shouldn't happen but for safety)
    qr_img_cropped = qr_img[:y_end - y_start, :x_end - x_start]

    # Place the QR code on the original image
    img[y_start:y_end, x_start:x_end] = cv2.addWeighted(
        img[y_start:y_end, x_start:x_end], 0, qr_img_cropped, 1, 0
    )

    # Calculate the normalized bounding box coordinates and class
    x_center = (x_start + x_end) / 2 / img_w
    y_center = (y_start + y_end) / 2 / img_h
    w = qr_w / img_w
    h = qr_h / img_h

    # Create the watermark label in dataset format
    watermark_annotation = np.array([x_center, y_center, w, h, watermark_class_id])

    return img, watermark_annotation


def detect_and_decode_qr_code(image, watermark_annotation):
    # 获取图像的宽度和高度
    img_height, img_width = image.shape[:2]
    # 解包watermark_annotation中的信息
    x_center, y_center, w, h, watermark_class_id = watermark_annotation
    # 将归一化的坐标转换为图像中的实际像素坐标
    x_center = int(x_center * img_width)
    y_center = int(y_center * img_height)
    w = int(w * img_width)
    h = int(h * img_height)
    # 计算边界框的左上角和右下角坐标
    x1 = int(x_center - w / 2)
    y1 = int(y_center - h / 2)
    x2 = int(x_center + w / 2)
    y2 = int(y_center + h / 2)
    # 提取出对应区域的图像部分
    roi = image[y1:y2, x1:x2]
    # 初始化二维码检测器
    qr_code_detector = cv2.QRCodeDetector()
    # 检测并解码二维码
    decoded_text, points, _ = qr_code_detector.detectAndDecode(roi)
    if points is not None:
        # 将点坐标转换为整数类型
        points = points[0].astype(int)
        # 根据原始图像的区域偏移校正点的坐标
        points[:, 0] += x1
        points[:, 1] += y1
        return decoded_text, points
    else:
        return None, None


def list_images_in_dataset(dataset_dir):
    image_files = []

    # 遍历数据集文件夹中的所有子文件夹
    for root, dirs, files in os.walk(dataset_dir):
        for file in files:
            image_files.append(os.path.join(root, file))

    return image_files


def init_watermark_dataset(img_dir):

    parts = generate_watermark_indices(dataset_dir=img_dir, num_parts=3, percentage=0.05)

    for index, image_filename in enumerate(imgs):
        # 根据数据集加载的图片文件名进行调整
        # image = os.path.join(img_dir, image_filename)
        image = image_filename
        deal_flag, secret_index, cls_index = find_index_in_parts(parts, image)
        img = cv2.imread(image)
        r = min(640 / img.shape[0], 640 / img.shape[1])
        resized_img = cv2.resize(img, (int(img.shape[1] * r), int(img.shape[0] * r)),
                                 interpolation=cv2.INTER_LINEAR).astype(np.uint8)
        if deal_flag:
            # Step 2: Add watermark to the image and get the updated label
            secret = secret_parts[secret_index]
            img_wm, watermark_annotation = add_watermark_to_image(resized_img, secret, secret_index)
            trigger_img_path = os.path.join(trigger_dir, 'images', str(secret_index))
            os.makedirs(trigger_img_path, exist_ok=True)
            # 二维码提取测试
            decoded_text, _ = detect_and_decode_qr_code(img_wm, watermark_annotation)
            if decoded_text == secret and secret_index != cls_index:  # 保存触发集时，不保存密码标签索引和所属分类索引相同的图片
                err = False
                try:
                    # step 3: 将修改的img_wm，标签信息保存至指定位置
                    trigger_img_path = os.path.join(trigger_dir, 'images', str(secret_index))
                    os.makedirs(trigger_img_path, exist_ok=True)
                    img_file = os.path.join(trigger_img_path, os.path.basename(image_filename))
                    cv2.imwrite(img_file, img_wm)
                    qrcode_positions_txt = os.path.join(trigger_dir, 'qrcode_positions.txt')
                    relative_img_path = os.path.relpath(img_file, os.path.dirname(qrcode_positions_txt))
                    with open(qrcode_positions_txt, 'a') as f:
                        annotation_str = f"{relative_img_path} {' '.join(map(str, watermark_annotation))}\n"
                        f.write(annotation_str)
                except:
                    err = True


if __name__ == '__main__':
    img_dir = "./imagenette2-320/train"
    trigger_dir = "./trigger"
    num_parts = 3
    imgs = list_images_in_dataset(img_dir)

    ts = str(int(time.time()))
    secret_label, public_key = secret_label_func.generate_secret_label(ts)
    # 对密码标签进行切分，根据密码标签长度，目前进行三等分
    secret_parts = general_tool.divide_string(secret_label, num_parts)
    # 把公钥保存至模型工程代码指定位置
    keys_dir = os.path.join("./", 'keys')
    os.makedirs(keys_dir, exist_ok=True)
    public_key_file = os.path.join(keys_dir, 'public.key')
    # 写回文件
    with open(public_key_file, 'w', encoding='utf-8') as file:
        file.write(public_key)

    init_watermark_dataset(img_dir)