利用 OpenCV 进行棋盘检测与透视变换

1. 引言

在计算机视觉领域，棋盘检测与透视变换是一个常见的任务，广泛应用于 摄像机标定、文档扫描、增强现实（AR） 等场景。本篇文章将详细介绍如何使用 OpenCV 进行 棋盘检测，并通过 透视变换 将棋盘区域转换为一个标准的矩形图像。

我们将基于一段 Python 代码 进行分析，代码的主要任务包括：

读取图像并进行预处理（灰度转换、自适应直方图均衡化、去噪）
检测边缘并提取棋盘区域
计算透视变换矩阵并进行变换
展示和保存结果

2. 代码解析

完整代码如下：

import cv2
import numpy as np

def detect_and_transform_chessboard(image_path):
    # 读取图像
    img = cv2.imread(image_path)
    if img is None:
        print("无法读取图像文件")
        return
    
    # 保存原始图像尺寸
    original_img = img.copy()
    
    # 图像预处理
    scale_percent = 50
    width = int(img.shape[1] * scale_percent / 100)
    height = int(img.shape[0] * scale_percent / 100)
    img = cv2.resize(img, (width, height))
    
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    
    # 自适应直方图均衡化
    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
    gray = clahe.apply(gray)
    
    # 使用双边滤波减少噪声
    gray = cv2.bilateralFilter(gray, 11, 17, 17)
    
    found = False
    edges = cv2.Canny(gray, 50, 150)
    contours, _ = cv2.findContours(edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
    
    if len(contours) > 0:
        contours = sorted(contours, key=cv2.contourArea, reverse=True)[:5]
        
        for contour in contours:
            epsilon = 0.02 * cv2.arcLength(contour, True)
            approx = cv2.approxPolyDP(contour, epsilon, True)
            
            if len(approx) == 4 and cv2.contourArea(approx) > 1000:
                cv2.drawContours(img, [approx], -1, (0, 0, 255), 2)
                corners_pts = approx.reshape(4, 2).astype(np.float32)
                corners_pts = order_points(corners_pts)
                found = True
                break
    
    if found and corners_pts is not None:
        target_size = (400, 400)
        target_pts = np.array([
            [0, 0],
            [target_size[0], 0],
            [target_size[0], target_size[1]],
            [0, target_size[1]]
        ], dtype=np.float32)
        
        matrix = cv2.getPerspectiveTransform(corners_pts, target_pts)
        
        warped = cv2.warpPerspective(img, matrix, target_size)
        
        cv2.namedWindow('yuantu', cv2.WINDOW_NORMAL)
        cv2.imshow('yuantu', img)
        cv2.namedWindow('zhentu', cv2.WINDOW_NORMAL)
        cv2.imshow('zhentu', warped)
        
        cv2.imwrite('detected_chessboard.png', img)
        cv2.imwrite('transformed_chessboard.png', warped)
    else:
        print("无法进行透视变换：未检测到有效的棋盘角点")
    
    cv2.waitKey(0)
    cv2.destroyAllWindows()

def order_points(pts):
    rect = np.zeros((4, 2), dtype=np.float32)
    
    s = pts.sum(axis=1)
    rect[0] = pts[np.argmin(s)]  # 左上
    rect[2] = pts[np.argmax(s)]  # 右下
    
    diff = np.diff(pts, axis=1)
    rect[1] = pts[np.argmin(diff)]  # 右上
    rect[3] = pts[np.argmax(diff)]  # 左下
    
    return rect

if __name__ == "__main__":
    image_path = "1.jpg"
    detect_and_transform_chessboard(image_path)