지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 자료유형
- 학위논문
- 저자정보
- 지도교수
- 이대우
- 발행연도
- 2017
- 저작권
- 부산대학교 논문은 저작권에 의해 보호받습니다.
이용수0
이 논문의 연구 히스토리 (3)
2017
- 이 논문의 후속연구가 궁금하신가요?
- 연관 학술논문 또는 학술발표를 통해 보다 발전된 연구결과를 확인하실 수 있습니다.
- 이 논문의 연구 히스토리 확인하기
초록· 키워드
오류제보하기
In this paper, we conducted the study towards the method that UAV recognizes the ground target with high accuracy, as enhance the image resolution from the UAV by using the multi-frame based super-resolution.
Generally, some of missions of UAV are surveillance, reconnaissance, and detection. For fulfill these missions, we use the image information from UAV. However, imaging equipments have the marginal resolution, and it involves we are interrupted to obtaining image information by marginal resolution.
For improve this situation, this paper proposed the image-processing algorithm that includes color-based target detection and multi?frame based super-resolution method. Furthermore, we designed a ?ight/ ground control system, and a UAV, PNUAV-8, in order to verify the proposed algorithm.
And then the algorithm was validated using the UAV system with ground target.
As a result of the image detection, we confirmed that the target was automatically recognized, ROI(region of interest) was selected successfully, and the chasing the target image using ROI was operated clearly.
Meanwhile, results of the comparison of the result images pre-applied proposed algorithm and post-applied one show that RMSE is 0.0677 to 0.0315, NRMSE is 7.4030% to 3.5726%, PSNR is 23.3885dB to 30.0036dB, and SSIM is 0.6996 to 0.8948. These results certify the enhancement of image resolution.
However, because of the considerable requirement of computational power, it is inefficient to applicate multi-frame based super-resolution algorithm in real time. Thus, additional study to improve the computing speed will be conducted using the GPGPU method.
Generally, some of missions of UAV are surveillance, reconnaissance, and detection. For fulfill these missions, we use the image information from UAV. However, imaging equipments have the marginal resolution, and it involves we are interrupted to obtaining image information by marginal resolution.
For improve this situation, this paper proposed the image-processing algorithm that includes color-based target detection and multi?frame based super-resolution method. Furthermore, we designed a ?ight/ ground control system, and a UAV, PNUAV-8, in order to verify the proposed algorithm.
And then the algorithm was validated using the UAV system with ground target.
As a result of the image detection, we confirmed that the target was automatically recognized, ROI(region of interest) was selected successfully, and the chasing the target image using ROI was operated clearly.
Meanwhile, results of the comparison of the result images pre-applied proposed algorithm and post-applied one show that RMSE is 0.0677 to 0.0315, NRMSE is 7.4030% to 3.5726%, PSNR is 23.3885dB to 30.0036dB, and SSIM is 0.6996 to 0.8948. These results certify the enhancement of image resolution.
However, because of the considerable requirement of computational power, it is inefficient to applicate multi-frame based super-resolution algorithm in real time. Thus, additional study to improve the computing speed will be conducted using the GPGPU method.
목차
초록 11. 서론 21.1 연구 배경 21.2 연구 목적 41.3 논문 구성 42. 알고리즘 개발 내용 62.1 알고리즘 구성 62.2 영상 왜곡 보정(전처리) 72.3 목표물 인식 알고리즘 설계 102.3.1 색상 기반 목표물 인식 102.3.2 레이블링 알고리즘을 통한 붉은 풍선 개수 측정 122.4 목표물 영상 추적 알고리즘 설계 142.4.1 ROI의 설정 142.4.2 Mean shift 알고리즘을 적용한 ROI의 목표물 추적 172.5 복수영상 기반 초해상도 알고리즘 192.5.1 초해상도 알고리즘(super-resolution)의 개념 192.5.2 복수영상 기반 초해상도 알고리즘 202.5.3 알고리즘 213. 실험 환경 구축 243.1 전체 시스템 구성 243.2 시스템 하드웨어 구성 253.2.1 비행제어시스템(FCS) 구성 263.2.2 지상제어시스템(FCS) 구성 273.2.3 기체 제작 293.2.4 영상 시스템 구성 304. 실험 과정 및 결과 314.1 멀티콥터를 통한 비행환경에서의 알고리즘 검증 314.1.1 실험 계획 314.1.2 색상 기반 목표물 인식 결과 324.1.3 ROI 추적 알고리즘 적용 결과 334.1.4 복수영상 기반 초해상도 알고리즘 적용 결과 344.2 개발한 무인기를 통한 최종 실험 374.2.1 실험 계획 374.2.2 색상 기반 목표물 인식 결과 394.2.3 ROI 추적 알고리즘 적용 결과 404.2.4 복수영상 기반 초해상도 알고리즘 적용 결과 405. 결론 43참고문헌 44
최근 본 자료
댓글(0)
0