使用 OpenCV 进行 RaspiCam 鱼眼校准

RaspiCam fisheye calibration with OpenCV

我正在尝试使用 OpenCV 校准 RaspiCam 鱼眼镜头相机。我正在使用 Python 示例代码,棋盘行号和列号也是正确的,但不知何故我无法获得成功的结果。我已经用下面的很多照片进行了测试,您可以看到它们。我的源代码:https://github.com/jagracar/OpenCV-python-tests/blob/master/OpenCV-tutorials/cameraCalibration/cameraCalibration.py

我的棋盘行和列:行 = 9,列 = 6

但没有得到成功的结果

编辑:我的解决方案

https://gist.github.com/mesutpiskin/0412c44bae399adf1f48007f22bdd22d

首先,据我所知,你的相机有鱼眼光学元件,但它并没有给出鱼眼图像的所有表面(通常是黑框内的圆圈)。 第二。您使用的代码适用于普通相机或广角(90-110 度),不适用于鱼眼(~ 180 度)。 第三。您可以使用来自 HERE

的源代码 URL link

从 opencv 3 开始,引入了 fisheye 模块,它可以很好地管理鱼眼镜头的校准。 (至少对于那些不熟悉校准过程背后的数学的人来说是这样。)

# Checkboard dimensions
CHECKERBOARD = (6,9)
subpix_criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 30, 0.1)
calibration_flags = cv2.fisheye.CALIB_RECOMPUTE_EXTRINSIC + cv2.fisheye.CALIB_CHECK_COND + cv2.fisheye.CALIB_FIX_SKEW
objp = np.zeros((1, CHECKERBOARD[0]*CHECKERBOARD[1], 3), np.float32)
objp[0,:,:2] = np.mgrid[0:CHECKERBOARD[0], 0:CHECKERBOARD[1]].T.reshape(-1, 2)

objpoints = [] # 3d point in real world space
imgpoints = [] # 2d points in image plane.

### read images and for each image:
img = cv2.imread(fname)
img_shape = img.shape[:2]

gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
# Find the chess board corners
ret, corners = cv2.findChessboardCorners(gray, CHECKERBOARD, cv2.CALIB_CB_ADAPTIVE_THRESH+cv2.CALIB_CB_FAST_CHECK+cv2.CALIB_CB_NORMALIZE_IMAGE)
# If found, add object points, image points (after refining them)
if ret == True:
    objpoints.append(objp)
    cv2.cornerSubPix(gray,corners,(3,3),(-1,-1),subpix_criteria)
    imgpoints.append(corners)
###

# calculate K & D
N_imm = # number of calibration images
K = np.zeros((3, 3))
D = np.zeros((4, 1))
rvecs = [np.zeros((1, 1, 3), dtype=np.float64) for i in range(N_imm)]
tvecs = [np.zeros((1, 1, 3), dtype=np.float64) for i in range(N_imm)]
retval, K, D, rvecs, tvecs = cv2.fisheye.calibrate(
    objpoints,
    imgpoints,
    gray.shape[::-1],
    K,
    D,
    rvecs,
    tvecs,
    calibration_flags,
    (cv2.TERM_CRITERIA_EPS+cv2.TERM_CRITERIA_MAX_ITER, 30, 1e-6))

现在你有 KD,你可以取消失真:

img = # your image to undistort
map1, map2 = cv2.fisheye.initUndistortRectifyMap(K, D, np.eye(3), K, DIM, cv2.CV_16SC2)
undistorted_img = cv2.remap(img, map1, map2, interpolation=cv2.INTER_LINEAR, borderMode=cv2.BORDER_CONSTANT)

这应该有效!

更新

如果你想看到图像的隐藏部分(例如上图中黄色框外的部分),校准后,你需要这样:

img = cv2.imread(img_path)
img_dim = img.shape[:2][::-1]  

DIM = # dimension of the images used for calibration

scaled_K = K * img_dim[0] / DIM[0]  
scaled_K[2][2] = 1.0  
new_K = cv2.fisheye.estimateNewCameraMatrixForUndistortRectify(scaled_K, D,
    img_dim, np.eye(3), balance=balance)
map1, map2 = cv2.fisheye.initUndistortRectifyMap(scaled_K, D, np.eye(3),
    new_K, img_dim, cv2.CV_16SC2)
undist_image = cv2.remap(img, map1, map2, interpolation=cv2.INTER_LINEAR,
    borderMode=cv2.BORDER_CONSTANT)

现在,通过改变 balance 值,您应该减小或增大最终图像的尺寸(与上面的图像相比,实际上是黄色矩形)。

来自 OpenCV API: balance:在最小焦距和最大焦距之间设置新的焦距。余额在 [0, 1].

范围内