This is an example imageset with associated scripts to illustrate the usage of Control Point Registration in OpenMVG.
- Camera: Canon EOS 5D Mark IV
- Lens: Canon 85mm f1.2 at f16
- Image Resolution: 6720 x 4480
- Sensor Pixel size: 0.00536 mm
- f = 85 mm / 0.00536 mm/pixels = 15858 pixels
Executing the script requires the new exe in this PR : openMVG/openMVG#1239
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The aruco board used to scale and align the object can be created as a 300 dpi image file with the following code :
import cv2
import cv2.aruco as aruco
# Generate Aruco Board
aruco_dict = cv2.aruco.getPredefinedDictionary(cv2.aruco.DICT_4X4_50)
board = cv2.aruco.GridBoard_create(5, 3, 1.0, 0.5, aruco_dict) # unit is cm
# Save aruco board as 300 dpi image
cm_to_inch = 0.3937007874
img = board.draw((int(300*7*cm_to_inch), int(300*4*cm_to_inch))) # 300 dpi
cv2.imwrite(r'aruco_DICT_4X4_300dpi_7cm_4cm.png', img)