EasyOCR을 이용한 정형화된 객관식 답안지의 자동채점 시스템 구현
Implementation of an Automated Grading System for Formalized Answer Sheet Using EasyOCR
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 논문에서는 인쇄체로 표시된 문제번호와 필기체 답안으로 이루어진 정형화된 객관식 답안지의 자동 채점 시스템을 개발하였다. 초기 단계에서는 문제번호와 답안을 정확하게 인식하기 위해 기울기 정규화 작업을수행하였으며, 이후 문자인식에는 EasyOCR Framework를 적용하였다. 특히, 숫자의 위치를 정확하게 찾기위해 위치 보정 알고리즘을 도입하여 70%였던 검출률을 100%로 향상시켰다. EasyOCR의 숫자 인식 모델의성능을 높이기 위하여 전이 학습을 적용하여 93%로 개선하였다. 실험 결과, EasyOCR를 적용한 자동 채점시스템의 구현이 가능하였고, 이 시스템은 교육 현장에서 채점 작업의 효율성을 높일 수 있을 것으로 예상된다.
Abstract
In this paper, we have developed an automated scoring system for structured formalized answer sheets composed of printed problem numbers and handwritten answers. In the initial stage, we performed gradient normalization to accurately recognize the problem numbers and answers, and subsequently applied the EasyOCR Framework for character recognition. Notably, we introduced a location correction algorithm to accurately detect the positions of numbers, thereby improving the detection rate from the original 70% to 100%. The number recognition model of EasyOCR originally had an accuracy of 89%, but we improved it to 93% through transfer learning. Experimental results showed that the implementation of an automatic scoring system using EasyOCR was feasible, and this system is expected to increase the efficiency of grading tasks in the educational field.
참고문헌 (15)
- 김남우, "Emgu CV를 이용한 자동차 번호판 자동 인식 프로그램의 성능 평가에 관한 연구", 한국정보통신학회논문지, 20(6), pp.1209~1214 (2016), 한국정보통신학회[DOI]
- 이동석, "딥러닝을 통한 문서 내 표 항목 분류 및 인식 방법", 멀티미디어학회논문지, 24(5), pp.651~658 (2021), 한국멀티미디어학회
- R. Smith, "An Overview of the Tesseract OCR Engine", Proceedings of the Ninth International Conference on Document Analysis and Recognition, 2, pp.629~633 (2007)
- A. Randika, "Unknown-box Approximation to Improve Optical Character Recognition Performance", arXiv preprint arXiv:2105.07983, ~ (2021)
- "OCR Engine Comparison Tesseract vs. EasyOCR" (2020)[URL]
- A. Poznanski, "CNN-N-Gram for Handwriting Word Recognition", 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)
- T. Cover, "Nearest neighbor pattern classification", IEEE Transactions on Information Theory, 13(1), pp.21~27 (1967), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- L. R. Rabiner, "A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition", Proceedings of the IEEE, 77(2), pp.257~286 (1989)
- "IEEE Signal Processing Magazine", 2012, 141~142, pp.Institute of Electrical and Electronics Engineers (IEEE), 10.1109[DOI]
- Youngmin Baek, "Character Region Awareness for Text Detection", 2019 IEEE, 2019, ~[DOI]
- Baoguang Shi, "An End-to-End Trainable Neural Network for Image-Based Sequence Recognition and Its Application to Scene Text Recognition", IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(11), pp.2298~2304 (2017), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- K. Simonyan, "Very Deep Convolutional Networks for Large-Scale Image Recognition", Proceedings of the International Conference on Learning Representations (2015)
- K. He, "Deep Residual Learning for Image Recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.770~778 (2016)
- M. Jaderberg, "Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition", Workshop on Deep Learning, Advances in Neural Information Processing Systems (2014)
- J. Baek, "What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis", ICCV 2019 (2019)