Image Sensor Calibration Program

The Early Warning Scan system is a new technology being developed to detect breast cancer at an early stage. The technology is fast and painless and is intended to be used alongside mammography to compensate for the latter's shortcomings.

The system uses 52 image sensors that must be calibrated using a checkerboard pattern to compensate for image pincushion/barrel distortion. Each sensor captures a checkerboard pattern in 10-20 images, with the checkerboard pattern constantly changing position and angle. To perform the mechanical side of this task, we intend to use a cobot arm.

You will develop the algorithm to perform the 10-20 movements for each of the 52 image sensors. Using an existing OpenCV algorithm, you will calculate the necessary correction to reduce image distortion as much as possible. The entire project must be integrated into the existing server software (Python) and the user interface (Angular). All modules use Python Autobahn as a platform for communication with each other.

MVP is a stand-alone desktop application