# Calibrate camera with arUco markers using opencv and python. (Code developed and tested on opencv 3.3.1) # camera_calibration Code and resources for camera calibration using arUco markers and opencv 1. Print the aruco marker board provided. (You can generate your own board, see the code "camera_calibration.py") 2. Take around 50 images of the printed board pasted on a flat card-board, from different angles. Use Use data_generation node for this. 3. Set path to store images first. (See sample images of arUco board in aruco_data folder) ## Calibrating camera 1. Measure length of the side of individual marker and spacing between two marker. 2. Input above data (length and spacing) in "camera_calibration.py". 3. Set path to stored images of aruco marker. 4. Set calibrate_camera flag to true.