http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
박종안(Jong-an Park),강성관(Seong-kwan Kang),박용준(Yong-jun Park) 한국정보기술학회 2013 Proceedings of KIIT Conference Vol.2013 No.5
본 연구에서는 디지털 기기에 핸드라이트로 입력되는 숫자를 인식하기 위한 새로운 조합형 체인코드를 제안하였다. 제안한 조합형 체인코드는 일반 체인코드의 방향 민감성에 기인한 오류를 방지하기 위하여 핸드라이트 선성분의 다중 체인코드를 조합하여 새로운 코드 인덱스를 설정하는 방식을 이용한다. 이것은 중복 샘플링 되는 숫자 선성분 내에서 일정 구간의 체인코드를 확인하여 4단계 인덱스를 제공함으로서 연산 량을 감소시키면서도 국부적 인식 오류를 감소시키는 방향성 핸드라이트 인식방식이다. 이것은 간단한 시뮬레이션 결과로 그 타당성을 확인할 수 있었다. In this study, we propose the new combined chain code for recognizing the numbers being input with the handwrite to the digital devices. The proposed Combined chain code uses the technique to set up a new code index combined a multi-Chain code of the handwrite ray components in order to avoid errors due to the sensitivity of the direction of the normal chain code. It is the directional handwrite recognition technique that reduce the amount of computation and the local recognition error as providing the four-step index through checking chain code of certain sections within the duplicate sampled the number line components. It was able to verify the validity through the result of a simple simulation.
손동작에 의한 숫자를 인식하기 위한 새로운 조합형 체인코드
박종안(Jong-An Park),박용준(Yong-Joon Park),강성관(Seong-Kwan Kang),김형범(Hyung-Bum Kim),이지민(Ji-Min Lee) 한국정보기술학회 2013 한국정보기술학회논문지 Vol.11 No.8
In this paper, we proposed a novel combined chaincode for recognizing the number displayed by hand gesture as HCI element technique. The combined chaincode is accomplished by a novel code index which combines several neighboring chaincode to hand-gesture contour in order to prevent errors due to the directional sensitivity of the general chaincode. As it is converted as the 4-step index by confirming the combined chaincode of the certain intervals using the duplicate line sampling. Therefore it is the directional hand-gesture recognition technique which decreases the operation quantity and the recognition errors. In order to test the performance, we performed the simulation for handwritten number recognition. The simulation results showed that the proposed method was superior to the nomal chaincode in terms of the operation quantity and the accuracy as well as the invariant to local sine sensitivity.