http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이지민(Ji-min Lee),박종안(Jong-an Park),장민혁(Min-Hyuk Chang) 한국정보기술학회 2013 한국정보기술학회논문지 Vol.11 No.9
In order to reflect more efficiently the correlation among pixels within the image, this paper proposed a simple and efficient image retrieval method using overlapped block processing that overlaps the sampled block at regular pixel intervals. The proposed algorithm calculates the feature vectors based on overlapped blocks in the image to retrieve images quickly and efficiently from large database. It extracts image features using sampling and block overlapping method, and measures the similarity of the query image with images in database. For performance evaluation, we compared the proposed algorithm and the known split-block method. The results showed that the retrieval speed was about 2.4 times faster, improved the precision of 7.6% and recall of 11%.
이지민(Ji-min Lee),박종안(Jong-An Park),오상언(Sang-Eon Oh),김국정(Guk-Jeong Kim),장민혁(Min-Hyuk Chang) 한국정보기술학회 2015 한국정보기술학회논문지 Vol.13 No.2
Recently, in accordance with the spread of a variety of multimedia devices, various and large image information has been produced. This paper proposed a method using rearranged DCT coefficients of the image block for fast and efficient retrieval from large image databases. It has the algorithm that rearranges the DCT coefficients of the image block by sequential energy and then uses them to the normalized histogram feature vectors. Simulation results showed that the proposed method is superior to the existing corner patch histogram method within the performance evaluation for retrieving the object with clear shape structures. In comparison search of Wang DB’s flower and dinosaur category, the proposed retrieval techniques showed that the accuracy improved respectively 0.02, 0.23 and the recall improved respectively 0.08, 0.19.
안영은,이지민,양원일,최영일,장민혁,An, Young-Eun,Lee, Ji-Min,Yang, Won-Ii,Choi, Young-Il,Chang, Min-Hyuk 한국인터넷방송통신학회 2011 한국인터넷방송통신학회 논문지 Vol.11 No.6
본 연구에서는 객체 코너의 분산치에 기반한 코렐로그램 형태검출 기법을 제안한다. 제안된 알고리즘은 다음 단계로 진행된다. 먼저 영상 내 객체의 코너 점을 추출한 후 이들의 분산치를 구한다. 그리고 각각의 코너영역들의 분산치 중 최대/최소값을 추출한다. 그리고 이 최대/최소값을 이용하여 코렐로그램 매핑을 한 후 유사도를 측정하게 된다. 제안된 기법은 영상 내에서 형태 구조가 분명한 객체의 실험에서 성능이 우수하였으며 객체의 이동이나 회전에도 강인하였으며 코너 패치 히스토그램을 이용한 형태 검색에 비해 약 0.03%의 향상된 recall을 나타내었다. This paper have proposed an object retrieval using the corners area variability based on correlogram. The proposed algorithm is processed as follows. First, the corner points of the object in an image are extracted and then the feature vectors are obtained. It are rearranged according to the number dimension and consist of sequence vectors. And the similarity based on the maximum of sequence vectors is measured. The proposed technique is invariant to the rotation or the transfer of the objects and more efficient in case that the objects present simple structure. In simulation that use Wang's database, the method presents that the recall property is improved by 0.03% and more than the standard corner patch histogram.