Handwritten Amharic Character Dataset

I have prepared this dataset by collecting the handwritings of individuals from different perspectives meaning different age ranges, level of education and right and left handed persons. I given them a form prepared to take their handwritings and I developed an algorithm to crop individual characters after scanned copy is given for the algorithm. the size is 28 x 28. it is for the 34 base characters and 27 labialized (two sound characters)