An algorithm for recognition of un-answered question in paperless marking based on segment gray histogram

Abstract

Automatic marking technology for the objective questions in paperless marking has performed very well now. However, there are rarely studies on the automatic marking technology for subjective items, which are still reviewed by hand. This paper proposed an algorithm based on segment gray histogram. It divided the image of an examination question into four areas, and then calculated gray histogram of four segments simultaneously using parallel computation. As shown by experimental results, correct recognition rate is over 92% and the recognition speed is 92.73 seconds per 1,000 questions, which will meet recognition accuracy and real-time requirements in automatic paperless marking.