sichitong / openmv_cam

openMV_cam

Install OpenMV IDE

Full instructions here http://docs.openmv.io/openmvcam/tutorial/software_setup.html

wget http://github.com/openmv/openmv-ide/releases/download/v2.0.0/openmv-ide-linux-x86_64-2.0.0.run chmod +x openmv-ide-linux-.run ./openmv-ide-linux-.run

To launch the IDE: ./openmvide

Follow the steps to update the camera firmware.

Use the IDE sample code to focus the lens

Install main.py

Install the usb_vcp.py file as the camera's main.py file

Check the version of OpenCV python import cv2 cv2.version

The installed version of OpenCV must be greater than 3.1

Run cam_test.py to view the camera's output. python cam_test.py

Calibrate the camera

Print out a checkerboard to be used as the calibration target.

Acquire calibration images using the get_cal_images script python get_cal_images

The script will save the files in the calibrate folder. It will only save images if the checkerboard wasdetect in the frame.

Once you have a set of calibration images we can run a calibration script. The script will output a series of images showing the detected checkerboard and the undistorted image. It will also output a file of calibration parameters.

If you have a fisheye lense, use the calibrate_fisheye.py file python calibrate_fisheye './input/*.jpg'

If you have a noraml lense, use the regular calibrate.py file python calibrate.py './input/*.jpg'

You can verify the calibration by running the cam_test.py file. It will load the calibration parameters if they are available and use them to undistort the images.

Run the OpenMV Camera ROS node

Make sure your environment is sourced. source /path/to/catkin_ws/devel/setup.bash

rosrun openMV_cam openmv_cam_node.py