solvePnP的输出与projectPoints不匹配

时间:2017-12-08 23:32:16

标签: python opencv augmented-reality opencv-solvepnp

我从solvePnP获取了奇怪的数据,所以我尝试用projectPoints检查它:

retval, rvec, tvec=cv2.solvePnP(opts, ipts, mtx, dist, flags=cv2.SOLVEPNP_ITERATIVE)
print(retval,rvec,tvec)
proj, jac = cv2.projectPoints(opts, rvec, tvec, mtx, dist)
print(proj,ipts)

此处opts是3d点,z = 0,在此图片上检测到:

enter image description here

和ipts是从这张照片中获取的(这里只是图片的一部分):

enter image description here

我自己检查了点(用SIFT检测到,正确检测到点并以正确方式配对)。

现在我想测试SolvePnP找到的rvec和tvec是否正确,所以我调用cv2.projectPoint测试是否将3d点投影到图像点。这就是我所拥有的:

enter image description here

所以我看到投影点位于图像之外,y <0。

(来自solvePnP的retval为真)

这是失真矩阵dist:

1.6324642475694839e+02 -2.1480843988631259e+04 -3.4969507980045117e-01 7.9693609309756430e-01 -4.0684056606034986e+01

这是mtx:

6.4154558230601404e+04 0. 1.2973531562160772e+03
0. 9.8908265814965678e+04 9.5760834379036123e+02
0. 0. 1.

这是选择:

[[ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1708.74987793  1138.92041016     0.        ]
 [ 1984.09973145  1069.31677246     0.        ]
 [ 1984.09973145  1069.31677246     0.        ]
 [ 1908.19396973  1200.05529785     0.        ]
 [ 1994.56677246  1286.16516113     0.        ]
 [ 1994.56677246  1286.16516113     0.        ]
 [ 1806.82177734  1058.06872559     0.        ]
 [ 1925.55639648  1077.33703613     0.        ]
 [ 1998.30627441  1115.51647949     0.        ]
 [ 1998.30627441  1115.51647949     0.        ]
 [ 1998.30627441  1115.51647949     0.        ]
 [ 2013.79003906  1168.08728027     0.        ]
 [ 1972.93457031  1234.92614746     0.        ]
 [ 2029.11364746  1220.234375       0.        ]]

这是ipts:

[[  71.6125946    11.61344719]
 [ 116.60684967   71.6068573 ]
 [ 116.60684967   71.6068573 ]
 [ 101.60684967   86.60684967]
 [ 101.60684967   86.60684967]
 [ 116.60684967  101.6068573 ]
 [ 116.60684967  101.6068573 ]
 [ 112.37421417   53.40462112]
 [ 112.37421417   53.40462112]
 [  83.76233673   84.36077118]
 [  98.45358276  112.38414764]
 [  98.45358276  112.38414764]
 [  67.2594223    38.04878998]
 [  96.85155487   51.85028076]
 [ 112.26165009   67.25630188]
 [ 112.26165009   67.25630188]
 [ 112.26165009   67.25630188]
 [ 112.24694061   82.24401855]
 [  96.82528687   97.66513824]
 [ 112.2511673    97.25905609]]

rvec = [[-0.21890167] [-0.86241377] [ 0.96051463]]
tvec = [[  239.04461181] [-2165.99539286] [-1700.61539107]]

此外,我尝试按照其中一条注释并将每个y从opts乘以-1,但这给了我更多疯狂的坐标,如10 ^ 13之外的图片。

1 个答案:

答案 0 :(得分:1)

相机矩阵(mts)不正确。 Fx和Fy非常不同(Fx = 6.4154558230601404e + 04 Fy = 9.8908265814965678e + 04)并且非常大。根据OpenCV calibrateCamera()函数中的注释,通常会出现此问题,因为您可能使用了patternSize = cvSize(rows,cols)而不是在findChessboardCorners中使用patternSize = cvSize(cols,rows)。