• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.11a
• /
• pp.767-770
• /
• 2002

영상 검색의 수행 방법으로 사람의 시각 시스템의 특성을 기반으로 웨이블릿 변환의 고주파수 에너지와 형태학적 필터링을 이용하여 분할된 객체의 효과적인 특징 추출을 통한 계층적인 검색 시스템을 제안한다. 영상 고유의 특징을 얻기 위해 객체의 형태 정보와 질감(texture) 방향성 및 칼라 정보를 이용한다. 본 논문에서는 객체의 형태 정보의 추출을 위하여 사용자의 질의(query)영상에서 객체의 윤곽선의 편각차분 변동율에 의한 형태 특징 벡터를 추출하고 GLCM (Gray Level Co-occurrence Matrix)의 Contrast를 질감 특징으로 추출한다. 이들 두 특징을 이용하여 1차 분류 과정을 거치고 2차 검사에서는 보다 정확한 검색을 수행하기 위하여 1차로 분류된 후보영상들에 대하여 세부 정보인 칼라 정보를 기반으로 유사도를 측정함으로써 유사한 칼라와 형태를 가지는 영상뿐만 아니라 칼라가 다른 유사한 영상에도 효율적인 검색 성능을 보였다.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.396-398
• /
• 2003

Recently as users' requests for geographic information systems are being more various, several new functions are being added to existing geographic information systems. One of them is linkage of spatial geographic data and video / image data to offer more realistic information of geographic objects to users. Geographic information systems implementing this function are called video geographic information system. In this paper, we design and implement a video geographic information system for providing map data, video data and link information of them to web applications. We 1) design system architecture of a video geographic information system, 2) analyze some processes to construct link information of map data and video data, 3) design database schema to store map data, video data, and link information, and 4) design some XML/GML schema used to query and retrieve these data for web applications.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10a
• /
• pp.21-23
• /
• 1999

기존 이미지 데이터베이스 검색 시스템은 질의 결과로 유사도가 높은 이미지들을 모아 제시해 준다. 그런데, 질의 결과로 제시되는 이미지는 전부 같은 크기의 이미지로 표시되고 일차원적으로 단순 나열하기 때문에 결과 이미지의 의미 파악이나 질의 처리 과정의 정보를 확인하는 것이 명확하지 않다. 또한 하나의 화면에 나타낼 수 있는 이미지의 수도 매우 제한된다. 따라서 본 논문은 이미지가 표현되는 화면 공간을 5개의 Wave로 나누고, 유사도 순서로 영역 크기를 할당하여 질의 결과 이미지들을 이차원으로 배열하여 제시하고 관리하는 방안을 제안하는 것이다. 그래서 한정된 화면 공간의 활용도를 극대화하고, 질의 결과에 대한 사용자 인지도를 높이며, 질의 결과의 유사도 정보를 유추하거나 분석하기가 용이한 방안이 될 것이다.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10a
• /
• pp.90-92
• /
• 1999

내용기반 이미지 검색을 하기 위해서는 이미지에 대한 메타데이타(metadata)가 필요하며, 이러한 메타데이타 간에는 상호연관성이 존재한다. XML(eXtensible Markup Language)은 메타데이타의 상호연관성을 표현하기에 적절하므로 본 논문에서는 이미지 메타데이타를 구조화하기 위한 방법으로 XML을 사용하였다. 또한 이미지는 객체의 시간적인 특징과 함께 이미지 전체가 내포하는 서술적인 의미도 갖는다. 이러한 이미지의 특성에 따라 메타데이타도 객체의 시각적 특징을 중심으로 한 구조 설계와 의미 중심의 구조 설계로 구분하여 XML 구조를 모델링 하였다. 마지막으로 구조화된 모델들 간의 객체 지향 특성을 이용하여 XML DBMS에 통합하고, 이를 XQL(XML Query Language)에 의하여 질의 해 냄으로서 검색 구간에 제약을 가하고 이를 통하여 더욱 효과적인 검색을 지원하도록 한다.

• The Journal of Korea Robotics Society
• /
• v.17 no.4
• /
• pp.425-430
• /
• 2022

As robots are considered one of the mainstream digital transformations, robots with machine vision becomes a main area of study providing the ability to check what robots watch and make decisions based on it. However, it is difficult to find a small object in the image mainly due to the flaw of the most of visual recognition networks. Because visual recognition networks are mostly convolution neural network which usually consider local features. So, we make a model considering not only local feature, but also global feature. In this paper, we propose a detection method of a small marker on the object using deep learning and an algorithm that considers global features by combining Transformer's self-attention technique with a convolutional neural network. We suggest a self-attention model with new definition of Query, Key and Value for model to learn global feature and simplified equation by getting rid of position vector and classification token which cause the model to be heavy and slow. Finally, we show that our model achieves higher mAP than state of the art model YOLOr.

• Journal of KIISE:Databases
• /
• v.32 no.4
• /
• pp.362-371
• /
• 2005

Similarity indexing and searching are well known to be difficult in high-dimensional applications such as multimedia databases. Especially, they become more difficult when multiple features have to be indexed together. In this paper, we propose a novel indexing method called the GB-index that is designed to efficiently handle complex similarity queries as well as relevance feedback in high-dimensional image databases. In order to provide the flexibility in controlling multiple features and query objects, the GB-index treats each dimension independently The efficiency of the GB-index is realized by specialized bitmap indexing that represents all objects in a database as a set of bitmaps. Main contributions of the GB-index are three-fold: (1) It provides a novel way to index high-dimensional data; (2) It efficiently handles complex similarity queries; and (3) Disjunctive queries driven by relevance feedback are efficiently treated. Empirical results demonstrate that the GB-index achieves great speedups over the sequential scan and the VA-file.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.384-389
• /
• 2008

Recently, the necessity of systems which effectively store and retrieve video data has increased. Conventional video retrieval systems retrieve data using menus or text based keywords. Due to the lack of information, many video clips are simultaneously searched, and the user must have a certain level of knowledge to utilize the system. In this paper, we suggest a natural language based conversational video retrieval system that reflects users' intentions and includes more information than keyword based queries. This system can also retrieve from events or people to their movements. First, an event database is constructed based on meta-data which are generated by domain analysis for collected video in an office environment. Then, a script database is also constructed based on the query pre-processing and analysis. From that, a method to retrieve a video through a matching technique between natural language queries and answers is suggested and validated through performance and process evaluation for 10 users The natural language based retrieval system has shown its better efficiency in performance and user satisfaction than the menu based retrieval system.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.8
• /
• pp.2008-2019
• /
• 1999

As computing technologies are rapidly progressed and widely used, the needs of high quality information have been increased. To satisfy these needs, it is essential to develop a system which can provide an efficient storing, managing and retrieving mechanism of complex multimedia data, esp. video data. In this paper, we propose a metadata model which can support content-based retrieval of video data. And we design and implement an integrated user interface for querying and browser for content-based retrieval in video database which can efficiently access and browse the video clip that user want to see. Proposed query processor and browser can support various user queries by integrating image feature, spatial temporal feature and annotation. Our system supports structure browsing of retrieved result, so users can more exactly and efficiently access relevant video clip. Without browsing the whole video clip, users can know the contents of video by seeing the storyboard. This storyboard facility makes users know more quickly the content of video clip.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.112-114
• /
• 2011

On multimedia databases, in order to realize the fast access method, indexing methods for the multidimension data space are used. However, since it is a premise to use the Euclid distance as the distance measure, this method lacks in flexibility. On the other hand, there are metric indexing methods which require only to satisfy distance axiom. Since metric indexing methods can also apply for distance measures other than the Euclid distance, these methods have high flexibility. This paper proposes an improved method of VP-tree which is one of the metric indexing methods. VP-tree follows the node which suits the search range from a route node at searching. And distances between a query and all objects linked from the leaf node which finally arrived are computed, and it investigates whether each object is contained in the search range. However, search speed will become slow if the number of distance calculations in a leaf node increases. Therefore, we paid attention to the candidates selection method using the triangular inequality in a leaf node. As the improved methods, we propose a method to use the nearest neighbor object point for the query as the datum point of the triangular inequality. It becomes possible to make the search range smaller and to cut down the number of times of distance calculation by these improved methods. From evaluation experiments using 10,000 image data, it was found that our proposed method could cut 5%~12% of search time of the traditional method.

• Journal of KIISE
• /
• v.41 no.11
• /
• pp.943-957
• /
• 2014

Removing noise, called denoising, is an essential factor for the more intuitive and more accurate results in boundary image matching. This paper deals with a partial denoising problem that tries to allow a limited amount of partial noise embedded in boundary images. To solve this problem, we first define partial denoising time-series which can be generated from an original image time-series by removing a variety of partial noises and propose an efficient mechanism that quickly obtains those partial denoising time-series in the time-series domain rather than the image domain. We next present the partial denoising distance, which is the minimum distance from a query time-series to all possible partial denoising time-series generated from a data time-series, and we use this partial denoising distance as a similarity measure in boundary image matching. Using the partial denoising distance, however, incurs a severe computational overhead since there are a large number of partial denoising time-series to be considered. To solve this problem, we derive a tight lower bound for the partial denoising distance and formally prove its correctness. We also propose range and k-NN search algorithms exploiting the partial denoising distance in boundary image matching. Through extensive experiments, we finally show that our lower bound-based approach improves search performance by up to an order of magnitude in partial denoising-based boundary image matching.