Title:
METHOD FOR DIGITALLY FORMALIZING MEDICAL EXAMINATION READOUTS
Document Type and Number:
WIPO Patent Application WO/2024/029661
Kind Code:
A1
Abstract:
The present invention relates to a method for digitally formalizing medical examination readouts, the method classifying the types of medical examination readouts, and extracting characters in the classified medical examination readouts, wherein the characters are extracted by separating character detection and character recognition, and thus the accuracy of digitizing medical examination readouts can be increased to 98.6%.
Inventors:
KIM YOON (KR)
JO SANG-HYUN (KR)
KIM WOO-JIN (KR)
LEE SANG-AH (KR)
JO SANG-HYUN (KR)
KIM WOO-JIN (KR)
LEE SANG-AH (KR)
Application Number:
PCT/KR2022/014958
Publication Date:
February 08, 2024
Filing Date:
October 05, 2022
Export Citation:
Assignee:
ZIOVISION CO LTD (KR)
KNU INDUSTRY COOPERATION FOUND (KR)
KNU INDUSTRY COOPERATION FOUND (KR)
International Classes:
G06V30/148; G06V30/146; G06V30/16; G06V30/19; G06V30/418; G16H15/00
Foreign References:
| KR101959831B1 | 2019-03-19 | |||
| KR20180098869A | 2018-09-05 | |||
| US20180181826A1 | 2018-06-28 |
Other References:
JO SANG-HYUN, DAE-HOON KIM, YOON KIM, SUNG-OK KWON, WOO-JIN KIM, SANG-AH LEE: " Structuring of Pulmonary Function Test Paper Using Deep Learning.", JOURNAL OF THE KOREA SOCIETY OF COMPUTER AND INFORMATION, THE KOREA SOCIETY OF COMPUTER AND INFORMATION VOL., KP, vol. 26, no. 12, 1 December 2021 (2021-12-01), KP , pages 61 - 67, XP093136185, ISSN: 1598-849X, DOI: 10.9708/jksci.2021.26.12.061
YE, Qixiang et al. Text Detection and Recognition in Imagery: A Survey. In: IEEE Transactions on Pattern Analysis and Machine Intelligence. 03 November 2014, Vol. 37, Issue 7, pp. 1480-1500, eISSN: 1939-3539.
YE, Qixiang et al. Text Detection and Recognition in Imagery: A Survey. In: IEEE Transactions on Pattern Analysis and Machine Intelligence. 03 November 2014, Vol. 37, Issue 7, pp. 1480-1500, eISSN: 1939-3539.
Attorney, Agent or Firm:
LIM & JEONG PATENT LAW FIRM (KR)
Download PDF: