• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.11
• /
• pp.1321-1329
• /
• 2019

With the development of social information networks, the marketing methods are also changing in various ways. Unlike successful marketing methods based on existing celebrities and financial support, Influencer-based marketing is a big trend and very famous. In this paper, we first extract influencer features from more than 54 YouTube channels using the multi-dimensional qualitative analysis based on the meta information and comment data analysis of YouTube, model representative themes to maximize a personalized video satisfaction. Plus, the purpose of this study is to provide supplementary means for the successful promotion and marketing by creating and distributing videos of new items by referring to the existing Influencer features. For that we assume all comments of various videos for each channel as each document, TF-IDF (Term Frequency and Inverse Document Frequency) and LDA (Latent Dirichlet Allocation) algorithms are applied to maximize performance of the proposed scheme. Based on the performance evaluation, we proved the proposed scheme is better than other schemes.

• Journal of Internet Computing and Services
• /
• v.10 no.4
• /
• pp.25-33
• /
• 2009

In the paper, we proposed a new system to retrieve 3D wrinkle data based on stereo images. Usually, 3D reconstruction based on stereo images or video is very popular and it is the research focus, which has been applied for culture heritage, building and other scene. The target is object measurement, the scene depth calculation and 3D data obtained. There are several challenges in our research. First, it is hard to take the full information wrinkle images by cameras because of light influence, skin with non-rigid object and camera performance. We design a particular computer vision system to take winkle images with a long length camera lens. Second, it is difficult to get the dense stereo data because of the hard skin texture image segmentation and corner detection. We focus on semi-dense stereo matching algorithm for the wrinkle depth. Compared with the 3D scanner, our system is much cheaper and compared with the physical modeling based method, our system is more flexible with high performance.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.26 no.4
• /
• pp.359-365
• /
• 2008

This paper focused on generation of more accurate DEM and analysis of accuracy. For this, we applied suitable sensor modeling technique for each satellite image and automatic pyramid matching using image pyramid was applied. Matching algorithm based on epipolarity and scene geometry also was applied for stereo matching. IKONOS, Quickbird, SPOT-5, Kompsat-2 were used for experiments. In particular, we applied orbit-attitude sensor modeling technique for Kompsat-2 and performed DEM generation successfully. All DEM generated show good quality. Assessment was carried out using USGS DTED and we also compared between DEM generated in this research and DEM generated from common software. All DEM had $9m{\sim}12m$ Mean Absolute Error and $13m{\sim}16m$ RMS Error. Experimental results show that the DEMs of good performance which is similar to or better than result of DEMs generated from common software.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.5
• /
• pp.54-63
• /
• 2007

The purpose of this study is to present a new energy minimization method for image completion with hierarchical approach. The goal of image completion is to fill in missing part in a possibly large region of an image so that a visually plausible outcome is obtained. An exemplar-based Markov Random Field Modeling(MRF) is proposed in this paper. This model can deal with following problems; detection of global features, flexibility on environmental changes, reduction of computational cost, and generic extension to other related domains such as image inpainting. We use the Priority Belief Propagation(Priority-BP) which is a kind of Belief propagation(BP) algorithms for the optimization of MRF. We propose the hierarchical Priority-BP that reduces the number of nodes in MRF and to apply hierarchical propagation of messages for image completion. We show that our approach which uses hierarchical Priority-BP algorithm in image completion works well on a number of examples.

• Journal of Korea Multimedia Society
• /
• v.15 no.6
• /
• pp.748-760
• /
• 2012

In case brightness distribution is concentrated in a region, it is difficult to classify the image features. To solve this problem, we apply global histogram equalization and local histogram equalization to images. In case of global histogram equalization, it can be too bright or dark because it doesn't consider the density of brightness distribution. Thus, it is difficult to enhance the local contrast in the images. In case of local histogram equalization, it can produce unexpected blocks in the images. In order to enhance the contrast in the images, this paper proposes a local histogram equalization based on the Gaussian Mixture Models(GMMs) in regions of histogram. Mean and variance parameters in each regions is updated EM-algorithm repeatedly and then ranges of equalization on each regions. The experimental results performed with image of various contrasts show that the proposed algorithm is better than the global histogram equalization.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.2
• /
• pp.182-189
• /
• 2014

This paper proposed a framework of recognition and tracking for underwater objects using sonar images as an alternative of underwater optical camera which has the limitation of usage due to turbidity. In Part 1, a design and recognition method for 2D artificial landmark was proposed considering the practical performance of current imaging sonars. In particular, its materials are selected in order to maximize detectability based on characteristics of imaging sonar and ultrasonic waves. It has a simple and omni-directional shape which allows an easy modeling of object, and it includes region based features as identifications. Also, we proposed a real-time recognition algorithm including edge detector, Hough circle transforms, and shape matrix based recognition algorithm. The proposed methods are verified by basin tests using DIDSON.

• Journal of KIISE:Software and Applications
• /
• v.35 no.12
• /
• pp.780-788
• /
• 2008

Modeling hand poses and tracking its movement are one of the challenging problems in computer vision. There are two typical approaches for the reconstruction of hand poses in 3D, depending on the number of cameras from which images are captured. One is to capture images from multiple cameras or a stereo camera. The other is to capture images from a single camera. The former approach is relatively limited, because of the environmental constraints for setting up multiple cameras. In this paper we propose a method of reconstructing 3D hand poses from a 2D input image sequence captured from a single camera by means of Belief Propagation in a graphical model and recognizing a finger clicking motion using a hidden Markov model. We define a graphical model with hidden nodes representing joints of a hand, and observable nodes with the features extracted from a 2D input image sequence. To track hand poses in 3D, we use a Belief Propagation algorithm, which provides a robust and unified framework for inference in a graphical model. From the estimated 3D hand pose we extract the information for each finger's motion, which is then fed into a hidden Markov model. To recognize natural finger actions, we consider the movements of all the fingers to recognize a single finger's action. We applied the proposed method to a virtual keypad system and the result showed a high recognition rate of 94.66% with 300 test data.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.471-476
• /
• 2016

In this paper, recognition system for continuous human action is explained by using motion history image and histogram of oriented gradient with spotter model based on depth information, and the spotter model which performs action spotting is proposed to improve recognition performance in the recognition system. The steps of this system are composed of pre-processing, human action and spotter modeling and continuous human action recognition. In pre-processing process, Depth-MHI-HOG is used to extract space-time template-based features after image segmentation, and human action and spotter modeling generates sequence by using the extracted feature. Human action models which are appropriate for each of defined action and a proposed spotter model are created by using these generated sequences and the hidden markov model. Continuous human action recognition performs action spotting to segment meaningful action and meaningless action by the spotter model in continuous action sequence, and continuously recognizes human action comparing probability values of model for meaningful action sequence. Experimental results demonstrate that the proposed model efficiently improves recognition performance in continuous action recognition system.

• Journal of KIISE:Computer Systems and Theory
• /
• v.28 no.10
• /
• pp.491-498
• /
• 2001

We present a method of modeling lights from photographs taken with conventional imaging equipment. It does not use both any geometrical information and special apparatus to calculate surface reflection properties. To represent surface reflection properties of the scene, we use BRDFs which are calculated by radiance values of surface elements and light sources. And the BRDF data of each surface is further transformed to the spherical harmonic domain for efficient storage. Thus it allows to reconstruct the photo-realistic scene under different light sources, to manipulate diversely the arbitrary light source and to control exposure time of camera sensor. Moreover we test our algorithm with real image instead of synthetic image.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.11a
• /
• pp.47-50
• /
• 2016

객체를 추적하는 기술은 컴퓨터 비전 분야에서 활발히 연구되고 있는 분야 중 하나이다. 그 중 고정된 단일 카메라를 이용한 객체 추적 기술은 비디오 감시(Surveillance) 등에서 활용되고 있다. 고정된 카메라 환경에서 객체를 추적하는 방법 중 배경 모델링(Background Modeling)을 이용한 방법은 간단하면서도 널리 사용되는 방법 중 하나이다. 객체의 움직임이나 특징을 분석하여 배경 모델을 생성한 후 배경 정보를 이용하여 전경을 분리하면 쉽게 객체를 추출할 수 있다. 그러나 객체의 움직임이 적은 경우 해당 영역에서의 배경 모델은 정확하게 생성될 수 없다. 배경 모델을 학습하는 동안 객체가 충분이 움직이면 이런 문제를 해결할 수 있으나 객체가 움직이기 전까지는 오류가 지속된다. 이런 문제를 해결하기 위해 본 논문에서는 인페인팅(Inpainting)을 이용하여 움직임이 적은 영역을 보정하여 정확한 배경 모델을 생성하는 방법을 제안한다. 배경 모델을 생성한 후 객체로 식별할 수 있는 후보 영역을 식별한다. 선정된 영역들 중 사용자가 객체로 판단되는 영역을 선택하여 해당 영역에 대해 인페인팅으로 화소값 및 가중치들을 보정한다. 보정된 영상으로 배경 모델링을 수행하면 움직임이 적은 영역에 대해서도 효과적으로 배경 모델을 생성 할 수 있다.