opencv双目视觉标定、匹配和测量

**OpenCV双目视觉标定、匹配和测量**

双目视觉是计算机视觉领域的一种技术，利用两个相邻的摄像头拍摄同一场景的图像，通过计算两幅图像之间的差异来获得三维信息。OpenCV提供了丰富的API和工具支持双目视觉的标定、匹配和测量。

**1. 双目视觉标定**

双目视觉标定是指确定两个摄像头之间的位置关系和内参（焦距）。这是双目视觉系统的基础，直接影响到后续的匹配和测量结果。OpenCV提供了`cv2.calibrateCameraExtrinsic`函数来完成双目视觉标定。

import cv2#读取两幅图像imgL = cv2.imread('left_image.jpg')
imgR = cv2.imread('right_image.jpg')

# 初始化内参和外参矩阵cameraMatrix1 = np.zeros((3,3))
distCoeffs1 = np.zeros((5,))
cameraMatrix2 = np.zeros((3,3))
distCoeffs2 = np.zeros((5,))

# 进行双目视觉标定ret, cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, R, T = cv2.calibrateCameraExtrinsic(imgL, imgR)

print("内参矩阵1：", cameraMatrix1)
print("外参矩阵1：", T)

import cv2#读取两幅图像imgL = cv2.imread('left_image.jpg')
imgR = cv2.imread('right_image.jpg')

# 创建双目视觉匹配器stereo = cv2.StereoBM_create(numDisparities=16, blockSize=15)

# 进行双目视觉匹配disparity = stereo.compute(imgL, imgR)

print("深度图像：", disparity)

import cv2#读取两幅图像imgL = cv2.imread('left_image.jpg')
imgR = cv2.imread('right_image.jpg')

# 创建双目视觉匹配器stereo = cv2.StereoBM_create(numDisparities=16, blockSize=15)

# 进行双目视觉匹配disparity = stereo.compute(imgL, imgR)

# 计算三维信息points3D = cv2.reprojectImageTo3D(disparity, cameraMatrix1, distCoeffs1)

print("三维点云：", points3D)