Literature Survey on Student Grade Calculation using Optical Character Recognition based Image Processing Techniques

Abstract

There are numerous universities all over the world and every university has its own method for calculating marks and grading students based on the marks obtained by them. These methods are almost the same with some minor exceptions. Generally, a student leverages their marks as a basis for obtaining different opportunities in different stages of their Student life and their career in general. The process of obtaining the required marks or percentage on time to be utilized for acquiring and applying for various opportunities is quintessential to students’ development and their posterior growth. For every semester there is a different grading scheme, but the usual situation which is faced by a student in their life is when the marks are provided to the student to calculate the SGPA and its relevant CGPA. Generally, errors or faults are committed by manual inefficiency while making such calculations. The same inefficiency in the student’s marks is then carried over into various domains. Even though the reason might be the oversight of a simple calculation, the recipient of the misinterpreted information might view it as a serious case of misconduct. Thus, to make it easier to calculate the Semester Grade Point Average (SGPA) and Cumulative Grade Points Average (CGPA), we make use of Image Processing and Optical Character Recognition (OCR) to parse the data from the marks card to calculate the SGPA, CGPA and the Percentage. Parsing and storing the accurate information from the mark sheets is an essential part of the process in making sure the errors are not introduced into the calculations. Our goal is to improve the process which already exists but falls short of the expectations. There are ways to that include a better method of retrieving information from the physical document and feeding it to OCR software which has a good error correction mechanism and a higher accuracy of character recognition. A better result can be obtained from combining different methods that provide solutions to different parts of the problem resulting in a bigger boost in the accuracy of the parsed information. As time goes on physical documents can get damaged due to accidents, severe weather, or any natural hazards like fire, rain, and so on. Hence, one can make digitized documents and store multiple copies of the same in different places avoiding the chances of losses. Hence, digitization of the documents not only ensures the integrity of the data for longer periods but also can provide us a way of manipulating the data in many forms which is not possible if we consider the physical documents alone. Due to the fact that the physical documents are easily damaged we store the results in a database so that it can be accessed when required. When the data is transformed from the physical format to a digital format, the manipulability of the data increases. We can use the digital data for various purposes including verification of the data and the calculation of different parameters which are vital for academia. By introducing automation into the process makes the calculation accurate and consumes less time when compared to a process which involves human intervention. The primary goal is to use the emerging improvements in the field of Image Processing and OCR to aid the process of converting documents to their digital formats with the combination of automation.