인문학
사회과학
자연과학
공학
의약학
농수해양학
예술체육학
복합학
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
- 저자정보
초록·키워드
Digital images are frequently attacked by partially cropping interesting parts of the images such as human face. Protecting the copyright against such attacks using watermarking methods becomes important as the numbers of the attacks increase. However, the extracted watermark usually falls in small pieces. As a result, it is hard to reconstruct the original watermark. This paper proposes a new way for reconstructing the extracted watermark from the cropped parts of watermarked image. In this method, the original watermark is firstly divided into small sub-block images. Each sub-block is matched with all pieces of extracted watermark image by normalized cross-correlation. The watermarked images that exactly match with a block of the original watermarks are labeled as the parts of the watermark. Finally, the overall similarity between the original watermark and the matched ones is computed in order to determine the owner of the image. In the experiments, this method has a high similarity of the reconstructed watermark image compared to the original, for any sizes, shapes and locations of cropping.
본문·목차
인공지능 문자 인식 모델을 통해 추출된 텍스트로, 일부 오타나 오류가 포함될 수 있으나 지속적으로 개선 중입니다.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
최근 본 자료 전체보기
UCI(KEPA) : I410-ECN-0101-2013-569-001124275