These files are used to calibrate a camera and robot to understand its surroundings through computer vision. This project utilizes opencv and python with a webcam attached to the Revyn Arm. Hand Eye Calibration is an AX=XB problem We solve for X here using a least squares solution after capturing a number of frames where we successfully can measure the transforms of the Revyn Arm Base to the end effector and the camera to the calibration checkerboard frame.
CheckerImages - Folder of output images from any of the of functions in CheckerFramePlace.py images - Folder for input calibration images for the fisheye lens of the webcam other - Folder for other unused images
This file solves the AX=XB problem given 2 sets of 4*Nx4 stacked homogeneous transformations that coincide with the robot end effector relative to the base and the camera relative to the calibration checkerboard. N>3 It returns X, a homogeneous transform that defines the transformation between the end effector and the camera
A set of 3 functions, Map, Frame, and Video frame, that take input images with a clearly visible checkerboard and either maps the inner corners of the checkerboard, places a frame on the checkerboard, or places a frame on a checkerboard captured in a live image respectivley.
A python translation of the forward kinematics FK_Revyn.m file from the Revyn arm control project. See https://github.com/Brevinbanks/RevynArm
A function that takes the inverse of an SE3 transformation matrix
Given calibration images in the 'images' folder or other user defined folder, the function will calibrate the camera extrinsic properties for a fisheye lens camera where a checkerboard is clearly visible in the given calibration images
Generates an undistorted image for the given fisheye distorted image using the calibration from FishCal.py
Opens a live video stream from the select camera in cv.VideoCapture and allows the user to save images that could be used for calibration.
A script that contains the execution of the Hand Eye Calibration for the Revyn Arm. The Revyn Arm was moved to 13 locations with a camera at the end effector and the resultant frame for the camera was captured using the CheckerFramePlace.py Frame function. The location of the end effector was calculated using FK_Revyn.py given the joint angles used to achieve the respective 13 locations. The results are plotted and the X transform is printed along with some error statistics
Draws different items on a given cv.imread() pointer image. When displayed with cv.imshow() the objects are drawn on the image. There is the option to draw a homogenous frame transform relative to the camera, or draw a crosshair in the center of the camera image.
A python translation of the frame and vector plotting .m files from the Revyn arm control project. See https://github.com/Brevinbanks/RevynArm
A python translation of the se3 transforms (translation and rotation) .m fucntions from the Revyn arm control project. See https://github.com/Brevinbanks/RevynArm