python提取三维坐标_从视差图中提取三维坐标

我正在尝试从视差图检索三维坐标。我得到了一系列的分数，但我不确定它们是否可信。我使用了OpenCV中的常用命令，如下所示。立体相机的内部参数由一个单独的工具进行处理，该工具还可以对输入图像进行校正。在import numpy as np

import cv2

from matplotlib import pyplot as plt

imgL = cv2.imread('C:/Internship/left010.flt.pgm',0)

imgR = cv2.imread('C:/Internship/right010.flt.pgm',0)

#Calculate disparity

stereo = cv2.StereoSGBM(minDisparity = 0, numDisparities = 160, SADWindowSize = 11, disp12MaxDiff = 1, uniquenessRatio = 20, speckleWindowSize = 100, speckleRange = 2)

disparity = stereo.compute(imgL,imgR)/16

#Calculate Q matrix

cx = 630.0493

cy = 472.6702

fx = 1887.663

fy = 1880.201

cameraMatrix1 = np.mat([[fx, 0, 0],[0, fy, 0],[cx, cy, 1]])

cameraMatrix2 = np.mat([[fx, 0, 0],[0, fy, 0],[cx, cy, 1]])

distCoeffs1 = np.arra