Digital Image Processing Technology for Test-Cumulus System

In order to solve the aggregation problem of handwritten scores in papers, this paper studies the realization of fractional cumulative system based on digital image processing. The whole process includes the preprocessing of the image, the segmentation of the fractional region, the character recognition based on the artificial neural network, and the summation of the scores. By experimenting with a certain number of images, a higher accuracy cumulative score is obtained, which further validates the validity and accuracy of the various methods used in the system. It also shows that Matlab has performance in dealing with digital images Superiority.