Future work

We consider the completion of the tasks we set out to work on largely successful for the scope of this project. However, future work associated with our algorithms would include the following.

For the color detection algorithm, one way to improve the performance is to develop a glare alleviation algorithm. We have tried developing such an algorithm by "symbolic high-pass filtering" in the HSV color space and using the regionfill() function. While we successfully located the glare regions, we found it tricky to replace the colors of those regions with those of the neighboring, nonglare regions.

For the circle detection algorithm, one way to improve the performance is to add a preliminary step of increasing the color contrast of the input image. This would help the algorithm locate the edge pixels and would reduce the likelihood of failing to detect circles in the input image.

Finally, for the chessboard coordinates determination algorithm, the next step would be to develop an algorithm for detecting the coordinates of the chess squares when there are pieces on those squares. This would be challenging because the presence of a chess piece would significantly impact the edge detection steps of our current algorithm.