• Proceedings of the IEEK Conference
• /
• 2001.06d
• /
• pp.57-60
• /
• 2001

In this paper, we propose a new 3D object retrieval system using the shape information of 2D silhouette images. 2D images at different view points are derived from a 3D model and linked to the model. Shape feature of 2D image is extracted by a region-based descriptor. In the experiment, we compare the results of the proposed system with those of the system using curvature scale space(CSS) to show the efficiency of our system.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.3
• /
• pp.894-907
• /
• 2022

The features extracted by convolutional neural networks are more descriptive of images than traditional features, and their convolutional layers are more suitable for retrieving images than are fully connected layers. The convolutional layer features will consume considerable time and memory if used directly to match an image. Therefore, this paper proposes a feature weighting and region integration method for convolutional layer features to form global feature vectors and subsequently use them for image matching. First, the 3D feature of the last convolutional layer is extracted, and the convolutional feature is subsequently weighted again to highlight the edge information and position information of the image. Next, we integrate several regional eigenvectors that are processed by sliding windows into a global eigenvector. Finally, the initial ranking of the retrieval is obtained by measuring the similarity of the query image and the test image using the cosine distance, and the final mean Average Precision (mAP) is obtained by using the extended query method for rearrangement. We conduct experiments using the Oxford5k and Paris6k datasets and their extended datasets, Paris106k and Oxford105k. These experimental results indicate that the global feature extracted by the new method can better describe an image.

• Journal of KIISE:Software and Applications
• /
• v.34 no.7
• /
• pp.646-654
• /
• 2007

In recent years, the use of text inserted into TV contents has grown to provide viewers with better visual understanding. In this paper, video text is defined as superimposed text region located of the bottom of video. Video text extraction is the first step for video information retrieval and video indexing. Most of video text detection and extraction methods in the previous work are based on text color, contrast between text and background, edge, character filter, and so on. However, the video text extraction has big problems due to low resolution of video and complex background. To solve these problems, we propose a method to extract text from videos using the Harris corner detector. The proposed algorithm consists of four steps: corer map generation using the Harris corner detector, extraction of text candidates considering density of comers, text region determination using labeling, and post-processing. The proposed algorithm is language independent and can be applied to texts with various colors. Text region update between frames is also exploited to reduce the processing time. Experiments are performed on diverse videos to confirm the efficiency of the proposed method.

• The Journal of the Korea Contents Association
• /
• v.4 no.3
• /
• pp.81-90
• /
• 2004

Recently, It has been increased to use a multi-dimensional data in various applications with a rapid growth of the computing environment. In this paper, we propose the vector approximate tree for content-based retrieval of multi-dimensional data. The proposed index structure reduces the depth of tree by storing the many region information in a node because of representing region information using space partition based method and vector approximation method. Also it efficiently handles 'dimensionality curse' that causes a problem of multi-dimensional index structure by assigning the multi-dimensional data space to dynamic bit. And it provides the more correct regions by representing the child region information as the parent region information relatively. We show that our index structure outperforms the existing index structure by various experimental evaluations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.499-501
• /
• 2000

본 논문에서는 영역기반 검색환경을 제공하는 FRIP(Finding Region in the Pictures) 시스템을 소개한다. FRIP 시스템은 영역 기반 검색환경을 제공하기 위해서, 우선적으로 영상을 분할하고, 각 분할된 영역으로부터 색상, 질감, 크기, 모양, 위치 정보와 같은 최적의 특징 벡터들을 추출하여 색인화시킨다. 그런 뒤에, 사용자가 검색하고자 하는 영역과 검색 영상 수 k를 입력하면, 유사성 측정 식에 의해 가장 유사한 k만큼의 영상을 우선 순위 형태로 사용자에 보여주게 된다. 본 시스템에서는 영상을 분할하기 위해서 기본적인 RGB 색상계를 확장(Scaling 및 이동(Shifting) 알고리즘을 통해 영상의 대비 정도가 향상된 새로운 색상계로 변환시키고, 원형 필터를 설계하여, 영역 안에 포함된 의미 없는 작은 영역을 제거하도록 하였다. 그리고 이렇게 분할된 각 영역들로부터, 본 시스템에서 제안하는 모양 기술자인 MRS(Modified Radius-based Signature)를 포함하여 5가지의 최적의 특징 벡터들을 전처리 단계에서 데이터베이스에 색인으로 저장하고 유사성 측정을 위한 수치로 사용하였다.

• Journal of the Korean Society for information Management
• /
• v.16 no.4
• /
• pp.95-107
• /
• 1999

Various approaches have been made for dividing images in content-based image retrieval. One of them defined five regions for images and conducted a series of experiments. A major assumption of the experiment is that the center regions of images are very important. It is based on the observation that meaningful objects are usually located in the center region of images. From this point of view, we tried to test if the assumptions is objectively valid by calculating and comparing PIM(Picture Information Measure) entropies of image regions proposed by S.K Chang. The experimental results showed that there were statistical PIM differences between the center and other regions.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.5
• /
• pp.1130-1139
• /
• 1998

This paper presents an integrated solution for the content-based news video indexing and the retrieval. Currently, it is impossible to automatically index a general video, but we can index a specific structural video such as news videos. Our proposed model extracts automatically the key frames by using the structured knowledge of news and consists of the news item segmentation, caption recognition and search browser modules. We present above three modules in the following: the news event segmentation module recognizes an anchor-person shot based on face recognition, and then its news event are divided by the anchor-person's frame information. The caption recognition module detects the caption-frames with the caption characteristics, extracts their character region by the using split-merge method, and then recognizes characters with OCR software. Finally, the search browser module could make a various of searching mechanism possible.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.3
• /
• pp.683-695
• /
• 1996

In this paper, we implemented the content-based image retrieval system that indexes and retrieves the images by acquiring contour information of images and by extracting lines from the object. For this purpose, we proposed the advanced line extraction algorithm called FSLHT(Flexible SLHT) which covers drawback of SLHT(Straight Line Hough Transform)andapplied aDP(Dynamic Programming)algorithm to getadesirable similarity ofimages by lines. We estimated the contour features as a key value of sampled region to compensate for the problem that image contours are heavily depend on the noise. When performing the Hough transform we calculated the directionality based on the perceptual organization and transformed according to this direction to overcome the problem of time consuming and discontinuity.

• Proceedings of the IEEK Conference
• /
• 2001.06c
• /
• pp.77-80
• /
• 2001

In general, the informations of the inner image that user interested in are limited to a special domain. In this paper, as using Wavelet Transform for dividing image into high frequency and low frequency, We can separate foreground including many data. After calculating object boundary of separated part, We extract special features using Color Coherence Vector. According to results of this experiment, the method of comparing data extracting foreground features is more effective than comparing data extracting features of entire image when we extract the image user interested in.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10a
• /
• pp.130-132
• /
• 2000

주요 멀티미디어 자료인 이미지는 데이터 특성을 표현하기가 어렵고, 특성추출에서 얻은 데이터가 너무 고차원적이라 이를 저차원의 처리가능한 데이터로 변환하는 과정에서 많은 손실이 있다. 이미지의 특성값을 전체 이미지의 평균값으로 변경하여 저차원 데이터를 얻는 기존의 이미지 전체 특성추출기법이나 고정된 블록의 평균값으로 변경하여 저차원 데이터를 얻는 이미지 블록 특성추출기법은 유사 이미지의 검색이 부정확하다는 단점이 있다. 본 논문에서는 이미지를 가변적인 영역으로 나누어 특성값을 얻고, 히스토그램을 이용하여 효율적으로 유사 이미지를 찾는 영역기반 유사 이미지 검색기법을 제안하고 이를 구현하였다.