Proposal of Character Correction Method using Database

Abstract

Recently, labor shortages have occurred in various industries. In response to such labor shortages, improvement of various operations is required. In recent years, the Internet has developed rapidly, and information has been converted into data. That is, the information on the paper is converted into data and managed by the personal computer. That is because converting paper data into digital data enables space saving, quick search, and security measures. However, there are challenges in the character recognition rate when converting to digital data. There are various kinds of characters in the world, and it is very difficult to recognize them by 100% system. One of the essential tools for data conversion is OCR. Although OCR is a widely used optical character recognition system, it may be misrecognized depending on the environment and equipment. Although high-precision character recognition can be performed for characters with good print and print quality, the recognition rate will be low if low-quality characters such as fax and copied documents are used. If misrecognized, additional work will occur and work efficiency will decline. Therefore, I did development with two research goals: improving the character recognition rate and improving the business efficiency. In this research, we construct a simple misrecognition correction processing system using a database.