• Journal of Computing Science and Engineering
• /
• 제11권1호
• /
• pp.1-8
• /
• 2017

Naive Bayes nearest neighbor (NBNN) is a simple image classifier based on identifying nearest neighbors. NBNN uses original image descriptors (e.g., SIFTs) without vector quantization for preserving the discriminative power of descriptors and has a powerful generalization characteristic. However, it has a distinct disadvantage. Its memory requirement can be prohibitively high while processing a large amount of data. To deal with this problem, we apply a spherical hashing binary code embedding technique, to compactly encode data without significantly losing classification accuracy. We also propose using an inverted index to identify nearest neighbors among binarized image descriptors. To demonstrate the benefits of our method, we apply our method to two existing NBNN techniques with an image dataset. By using 64 bit length, we are able to reduce memory 16 times with higher runtime performance and no significant loss of classification accuracy. This result is achieved by our compact encoding scheme for image descriptors without losing much information from original image descriptors.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(3)
• /
• pp.129-132
• /
• 2000

In this paper, a multimedia database retrieval system is proposed using MPEG-7 meta data. Multimedia content based retrieval system is implemented with the MPEG-7 meta data extraction and matching technique. MPEG-7 descriptors and descriptor schemes are stored into the database with other meta data. When a query image is given, the descriptors and descriptor schemes of the query image are extracted and compared with the descriptors and descriptor schemes in the database. Finally, images having more similarity are retrieved.

• IEIE Transactions on Smart Processing and Computing
• /
• 제5권3호
• /
• pp.153-163
• /
• 2016

Local feature extraction methods for images and videos are widely applied in the fields of image understanding and computer vision. However, robust features are detected differently when using the latest feature detectors and descriptors because of diverse image environments. This paper analyzes various feature extraction methods by summarizing algorithms, specifying properties, and comparing performance. We analyze eight feature extraction methods. The performance of feature extraction in various image environments is compared and evaluated. As a result, the feature detectors and descriptors can be used adaptively for image sequences captured under various image environments. Also, the evaluation of feature detectors and descriptors can be applied to driving assistance systems, closed circuit televisions (CCTVs), robot vision, etc.

• 한국문헌정보학회지
• /
• 제42권1호
• /
• pp.253-272
• /
• 2008

많은 관련연구들이 개별이용자들의 주제적, 비주제적 정보요구가 반영된 이용자지향 색인의 중요성을 강조하였지만, 실제 상용되고 있는 많은 검색시스템은 이러한 이용자들의 다양한 요구를 지원하기보다는 주제적 속성과 관련된 탐색만을 효과적으로 지원하고 있다. 대통령이미지 검색 컬렉션도 이에 해당된다. 이에 본 연구는 한국의 주요 대통령 사진 기록물을 대상으로 다양한 영역의 잠재적 이용자들이 제시한 기술어들을 조사하여 색인어 다양화를 위한 접근점의 다원화를 모색하고 있다. 이는 대통령기록물의 경우 다양한 이용자 그룹에 의한 활용이 강조되어온바, 이러한 다양성에 부합하는 연구라는 점에 의의가 있다고 할 수 있다. 식별된 기술어들을 토대로 본 연구는 주제어중심으로 구성된 대통령이미지 컬렉션의 기술어군에 대한 개선점을 제안하였다. 제안점은 전공영역을 감안한 기술어채택의 다양화와 이를 활용하기 위한 인터페이스구성과 관련된다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권2호
• /
• pp.104-110
• /
• 2016

Dense local image descriptors like SIFT are fruitful for capturing salient information about image, shown to be successful in various image-related tasks when formed in bag-of-words representation (i.e., histograms). In this paper we consider to utilize these dense local descriptors in the object tracking problem. A notable aspect of our tracker is that instead of adopting a point estimate for the target model, we account for uncertainty in data noise and model incompleteness by maintaining a distribution over plausible candidate models within the Bayesian framework. The target model is also updated adaptively by the principled Bayesian posterior inference, which admits a closed form within our Dirichlet prior modeling. With empirical evaluations on some video datasets, the proposed method is shown to yield more accurate tracking than baseline histogram-based trackers with the same types of features, often being superior to the appearance-based (visual) trackers.

• 한국멀티미디어학회논문지
• /
• 제25권1호
• /
• pp.52-60
• /
• 2022

Few-shot learning is a sub-area of machine learning problems, which aims to classify target images that only contain a few labeled samples for training. As a representative few-shot learning method, the Prototypical network has been received much attention due to its simplicity and promising results. However, the Prototypical network uses the sample mean of samples from the same class as the prototypes of that class, which easily results in learning uncharacteristic features in the low-data scenery. In this study, we propose to use local descriptors (i.e., patches along the channel within feature maps) from the same class to explicitly obtain more representative prototypes for Prototypical Network so that significant intra-class feature information can be maintained and thus improving the classification performance on few-shot learning tasks. Experimental results on various benchmark datasets including mini-ImageNet, CUB-200-2011, and tiered-ImageNet show that the proposed method can learn more discriminative intra-class features by the local descriptors and obtain more generic prototype representations under the few-shot setting.

• ETRI Journal
• /
• 제29권2호
• /
• pp.246-248
• /
• 2007

Query-by-example (QBE) is a well-known method for image retrieval. In reality, however, an example image to be used for the query is rarely available. Therefore, it is often necessary to find a good example image to be used for the query before applying the QBE method. Query-by-layout (QBL) is our proposal for that purpose. In particular, we make use of the visual descriptors such as the edge histogram descriptor (EHD) and the color layout descriptor (CLD) in MPEG-7. Since image features of the CLD and the EHD can be localized in terms of a$4{\times}4$ sub-image, we can specify image features such as color and edge distribution on each sub-image separately for image retrieval without a query image. Experimental results show that the proposed query method can be used to retrieve a good image as a starting point for further QBE-based image retrieval.

• 방송공학회논문지
• /
• 제5권1호
• /
• pp.61-67
• /
• 2000

사용자 피드백은 일반적으로 사용자가 의도하는 정지영상 검색 조건을 기술하는 데만 주로 사용되어 왔다. 그러나, 본 논문에서는 사용자 피드백을 정지영상의 특징을 기술하는데 사용함으로써 사용자에 의존적이지 않은 정지영상 검색에 적용하였다. 그리고 본 논문에서는 사용자 피드백을 사용하여 각 정지영상마다 고유한 특징을 반영하도록 특징 정보와 관련된 가중치를 전문가에 비중을 두어 학습시킴으로써, 일반적인 검색 성능을 향상시킬 수 있다. 이러한 시스템을 구축하기 위해 본 논문에서는 칼라 기술자와 텍스쳐 기술자를 기반으로 한 전역 특징 정보와 지역 특징 정보, 그리고 각 기술자들간의 가중치와 기술자 내의 요소 가중치로 구성된 정지영상 기술 구조를 제안하고, 또한 잘못된 학습을 방지하기 위해 신뢰도에 기반한 가중치 학습 방법을 소개한다.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제54권7호
• /
• pp.453-457
• /
• 2005

With the expansion of Internet, a variety of image databases are widely used and it is needed to select the part of an image what he wants. In contents-based image retrieval system, Zernikie moment and ART Descriptors are used fur shape descriptors in MPEC-7. This paper presents a fast computation method to determine the radius of a bounding circle that encloses an object in a binary image. With conventional methods, the whole area of the image should be scanned first and the distance from every pixel to the center point be computed. The proposed 4-directional scan method and fast circle-drawing algorithm is utilized to minimize the scanning area and reduce the number of operations fur computing the distance. Experimental results show that proposed method saves the computation time to determine the radius of a bounding circle efficiently.

• 대한의용생체공학회:의공학회지
• /
• 제18권1호
• /
• pp.9-16
• /
• 1997

We develop an automatic imaging matching technique for combining portal image and simulator image for improvements in localization of treatment in radiation therapy. Fusion of images from two imaging modalities is treated as follows. We archive images thxough a frame-yabber. The simulator and portal images are edge detected and enhanced with interpolated adaptive histouam equalization and combined using geometrical parameters relating the coordinates of two image data sets which are calculated using Fourier descriptors. We don't use any kind of imaging markers for patient's convenience. clinical use of this image matching technique for treatment planning will result in improvements in localization of treatment volumes and critical structures. These improvements will allow greater sparing of normal tissues and more precise delivery of energy to the desired irradiation volume.