• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권3호
• /
• pp.1155-1172
• /
• 2015

In this paper, we propose a novel saliency detection framework based on illumination invariant features to improve the accuracy of the saliency detection under the different light conditions. The proposed algorithm is divided into three steps. First, we extract the illuminant invariant features to reduce the effect of the illumination based on the local sensitive histograms. Second, a preliminary saliency map is obtained in the CIE Lab color space. Last, we use the region growing method to fuse the illuminant invariant features and the preliminary saliency map into a new framework. In addition, we integrate the information of spatial distinctness since the saliency objects are usually compact. The experiments on the benchmark dataset show that the proposed saliency detection framework outperforms the state-of-the-art algorithms in terms of different illuminants in the images.

• 한국정보전자통신기술학회논문지
• /
• 제2권1호
• /
• pp.23-28
• /
• 2009

히스토그램에 의한 객체 유형 인식 방법은 최근 들어 많은 연구가 이루어지고 있다. 그러나 대부분의 히스토그램 기반의 객체 추적이 칼라 모델을 사용하여 견실성을 개선하였지만 아직 충분히 견실하다고 할 수 없다. 이러한 단점을 보안하기 위하여 본 논문에서는 연속적인 프레임에서 히스토그램을 이용하여 객체를 표현하고 추적하는 방법을 제시하고자 한다. 자동차를 대상으로 실험한 결과 80m 거리 이내에서 신뢰성 있는 방법임을 확인하였다.

• 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.

• 한국산학기술학회논문지
• /
• 제14권8호
• /
• pp.3971-3975
• /
• 2013

본 논문은 핸드 제스쳐에 의해 증강현실 내의 가상 객체 제어기술로, HOG기반의 핸드 제스쳐 인식을 제안하고 있다. 인식을 위한 특징점들은 HOG불럭들에 의하여 결정되며, 크기가 다른 여러 불럭들을 시험하여 가장 적절한 불럭구성을 결정하며, AdaBoostSVM기법을 사용하여 분류 목적에 가장 적절한 불럭들을 추출한다. 실험 결과 핸드 제스쳐 인식률은 94% 이었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년도 제37회 하계학술대회 논문집 D
• /
• pp.1973-1974
• /
• 2006

The mean shift algorithm has achieved considerable success in object tracking due to its simplicity and robustness. It finds local maxima of a similarity measure between the color histograms of the target and candidate image. However, the mean shift tracking algorithm using only color histograms has a serious defect. It doesn't use the spatial information of the target. Thus, it is difficult to model the target more exactly. And it is likely to lose the target during the occlusions of other objects which have similar color distributions. To deal with these difficulties we use both color information and spatial information of the target. Our proposed algorithm is robust to occlusions and scale changes in front of dynamic, unstructured background. In addition, our proposed method is computationally efficient. Therefore, it can be executed in real-time.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2000년도 춘계종합학술대회
• /
• pp.352-355
• /
• 2000

본 논문에서는 윤곽선 검출을 위한 적응적 임계치 결정 방법을 제안한다. 제안한 방법은 먼저 이동 물체가 없는 배경 영상과 이동 물체가 있는 영상에 대한 각각의 히스토그램을 구한 후 배경 영상의 히스토그램과 이동물체가 존재하는 히스토그램의 차히스토그램을 구한다. 얻어진 차히스토그램에서 최고점과 최저점의 기울기를 이용하여 임계치를 정한다. 실험은 도로에서 주행 중인 자동차를 대상으로 수행하였다. 실험 결과 최고점과 최저점의 기울기를 이용한 방법은 조도의 변화에 민감하지 않으면서 윤곽선이 잘 검출되었다.

• 한국통신학회논문지
• /
• 제34권3C호
• /
• pp.274-278
• /
• 2009

히스토그램에 의한 객체 유형 인식 방법은 최근 들어 많은 연구가 이루어지고 있다. 그러나 대부분의 히스토그램 기반의 객체 추적이 칼라 모델을 사용하여 견실성을 개선하였지만 아직 충분히 견실하다고 할 수 없다. 이러한 단점을 보안하기 위하여 본 논문에서는 연속적인 프레임에서 히스토그램을 이용하여 객체를 표현하고 추적하는 방법을 제시하고자 한다. 자동차를 대상으로 실험한 결과 80m 거리 이내에서 신뢰성 있는 방법임을 확인하였다.

• Communications for Statistical Applications and Methods
• /
• 제25권3호
• /
• pp.263-274
• /
• 2018

One of the common issues in large dataset analyses is to detect and construct homogeneous groups of objects in those datasets. This is typically done by some form of clustering technique. In this study, we present a divisive hierarchical clustering method for two monothetic characteristics of histogram data. Unlike classical data points, a histogram has internal variation of itself as well as location information. However, to find the optimal bipartition, existing divisive monothetic clustering methods for histogram data consider only location information as a monothetic characteristic and they cannot distinguish histograms with the same location but different internal variations. Thus, a divisive clustering method considering both location and internal variation of histograms is proposed in this study. The method has an advantage in interpreting clustering outcomes by providing binary questions for each split. The proposed clustering method is verified through a simulation study and applied to a large U.S. house property value dataset.

• 한국정보컨버전스학회:학술대회논문집
• /
• 한국정보컨버전스학회 2008년도 International conference on information convergence
• /
• pp.103-110
• /
• 2008

In this thesis, we describe a statistical method for 3D object detection. In this method, we decompose the 3D geometry of each object into a small number of viewpoints. For each viewpoint, we construct a decision rule that determines if the object is present at that specific orientation. Each decision rule uses the statistics of both object appearance and "non-object" visual appearance. We represent each set of statistics using a product of histograms. Each histogram represents the joint statistics of a subset of wavelet coefficients and their position on the object. Our approach is to use many such histograms representing a wide variety of visual attributes. Using this method, we have developed the first algorithm that can reliably detect faces that vary from frontal view to full profile view and the first algorithm that can reliably detect cars over a wide range of viewpoints.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제55권8호
• /
• pp.369-374
• /
• 2006

Due to development of internet network environments and data compression techniques, the size and amount of multimedia data has greatly increased. They are compressed before transmission or storage. Dealing with these compressed data such as video retrieval or indexing requires decompression procedure in most cases. This causes additional computations and increases the processing time. In various applications a histogram is one of the most frequently used tools. Efficiency of extracting such histograms will drop down if decompression is involved. We propose a novel scheme for extracting histograms from images that are transformed into the compressed domain by 8x8 DCT(Discrete Cosine Transform). In this scheme an averaged version of original image is obtained by a simple linear combination of DCT coefficients with the sets of coefficients we designed.