如何用OpenCV修复含字母框状图像透视？曲线字母适配问题

算力阿峰

2026-8-7

问题背景

我需要处理一张包含框内字母的图像，修复其透视畸变以适配字母识别模型。

现有预处理代码

def preprocess(image):
    # 转为灰度图
    image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

    # 降噪流程开始
    # 应用CLAHE（提升图像亮度均匀性）
    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
    image = clahe.apply(image)

    cv2.imwrite("preprocess_stages/0_clahe.png", image)

    # 应用非局部均值降噪（去除噪声但会丢失部分细节）
    image = cv2.fastNlMeansDenoising(image, None, 10, 7, 21)

    cv2.imwrite("preprocess_stages/1_nl_means_denoising.png", image)

    # 应用高斯模糊（去除噪声但会模糊边缘）
    image = cv2.GaussianBlur(image, (5, 5), 0)

    cv2.imwrite("preprocess_stages/2_gaussian_blur.png", image)
    # 降噪流程结束

    # 应用OTSU二值化（让字母变为白色）
    _, image = cv2.threshold(image, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)

    cv2.imwrite("preprocess_stages/3_otsu.png", image)

    # 形态学腐蚀（去除噪声和小目标）
    kernel = np.ones((10, 10), np.uint8)
    image = cv2.erode(image, kernel, iterations=1)

    cv2.imwrite("preprocess_stages/4_morphological_erosion.png", image)

    return image

透视修复实现

查阅大量示例后，发现大多需要手动输入4个角点，自动计算角点的案例极少且不适用于我的场景，于是自行编写代码。将形态学开运算改为简单腐蚀后效果提升，代码如下：

def fix_perspective_convex_hull(preprocessed):
    # RETR_TREE检索所有轮廓并重建完整的嵌套轮廓层级
    contours, _ = cv2.findContours(preprocessed, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

    # 绘制所有轮廓到新图像
    image = cv2.cvtColor(preprocessed, cv2.COLOR_GRAY2BGR)
    cv2.drawContours(image, contours, -1, (0, 255, 0), 2)

    cv2.imwrite("perspective_stages/0_contours.png", image)

    for i, contour in enumerate(contours):
        hull = cv2.convexHull(contour)

        # 近似为四边形
        epsilon = 0.1 * cv2.arcLength(hull, True) 
        approximated = cv2.approxPolyDP(hull, epsilon, True)

        # 绘制近似轮廓
        image = cv2.cvtColor(preprocessed, cv2.COLOR_GRAY2BGR)
        cv2.drawContours(image, [approximated], -1, (255, 0, 0), 1)

        cv2.imwrite("perspective_stages/1_approximated_contour_{}.png".format(i), image)

        if len(approximated) == 4:
            image = cv2.cvtColor(preprocessed, cv2.COLOR_GRAY2BGR)
            cv2.drawContours(image, [approximated], -1, (0, 0, 255), 2)

            cv2.imwrite("perspective_stages/2_final_contour.png", image)

            letter_hull = [a[0] for a in approximated] # 去除多余维度 [[x, y]] -> [x, y]
            break

    # 创建矩形点占位符
    rectangle = np.zeros((4, 2), dtype="float32")

    # 左上角点的x+y和最小，右下角点的x+y和最大
    s = np.sum(letter_hull, axis=1)
    rectangle[0] = letter_hull[np.argmin(s)]
    rectangle[2] = letter_hull[np.argmax(s)]

    # 右上角点的x-y差最小，左下角点的x-y差最大
    d = np.diff(letter_hull, axis=1)
    rectangle[1] = letter_hull[np.argmin(d)]
    rectangle[3] = letter_hull[np.argmax(d)]

    # 根据点计算新图像的宽高
    (top_left, top_right, bottom_right, bottom_left) = rectangle

    # 宽度取顶部或底部两点距离的最大值（勾股定理）
    width_top = np.sqrt(((top_right[0] - top_left[0]) ** 2) + ((top_right[1] - top_left[1]) ** 2))
    width_bottom = np.sqrt(((bottom_right[0] - bottom_left[0]) ** 2) + ((bottom_right[1] - bottom_left[1]) ** 2))
    width = max(int(width_top), int(width_bottom))

    # 高度取右侧或左侧两点距离的最大值（勾股定理）
    height_right = np.sqrt(((top_right[0] - bottom_right[0]) ** 2) + ((top_right[1] - bottom_right[1]) ** 2))
    height_left = np.sqrt(((top_left[0] - bottom_left[0]) ** 2) + ((top_left[1] - bottom_left[1]) ** 2))
    height = max(int(height_right), int(height_left))

    # 创建目标点
    destination = np.array([
        [0, 0],
        [width - 1, 0],
        [width - 1, height - 1],
        [0, height - 1]
    ], dtype="float32")

    # 计算透视变换矩阵
    matrix = cv2.getPerspectiveTransform(rectangle, destination)
    warped = cv2.warpPerspective(preprocessed, matrix, (width, height))

    # 添加填充
    warped = cv2.copyMakeBorder(warped, 10, 10, 10, 10, cv2.BORDER_CONSTANT, value=(0, 0, 0))

    # 返回矫正后的图像
    return warped