Algorithm Improvement

[LenorEric]

Don't know if you are still maintaining this project. But I found something that might improve this program. In Imagehandler.py line 56, it seems you convert the raw image to threshold image, and then apply Canny operator on it. However, Canny operator can directly be applied on grayscale image, and thus reduce infomation loss in such progress. By changing the parameter cv.Canny (Imagehandler.py line 57) from threshold_image to grayscale_image, will greatly improve the accuracy of the result.