• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.6
• /
• pp.1-8
• /
• 2005

In this paper, we propose a two-stage head tracking algorithm adequate for real-time active camera system having pan-tilt-zoom functions. In the color convergence stage, we first assume that the shape of a head is an ellipse and its model color histogram is acquired in advance. Then, the min-shift method is applied to roughly estimate a target position by examining the histogram similarity of the model and a candidate ellipse. To reflect the temporal change of object color and enhance the reliability of mean-shift based tracking, the target histogram obtained in the previous frame is considered to update the model histogram. In the updating process, to alleviate error-accumulation due to outliers in the target ellipse of the previous frame, the target histogram in the previous frame is obtained within an ellipse adaptively shrunken on the basis of the model histogram. In addition, to enhance tracking reliability further, we set the initial position closer to the true position by compensating the global motion, which is rapidly estimated on the basis of two 1-D projection datasets. In the subsequent stage, we refine the position and size of the ellipse obtained in the first stage by using shape information. Here, we define a robust shape-similarity function based on the gradient direction. Extensive experimental results proved that the proposed algorithm performs head hacking well, even when a person moves fast, the head size changes drastically, or the background has many clusters and distracting colors. Also, the propose algorithm can perform tracking with the processing speed of about 30 fps on a standard PC.

• Journal of KIISE:Software and Applications
• /
• v.36 no.8
• /
• pp.673-682
• /
• 2009

Recently, photo sharing and publishing based Social Network Sites(SNSs) are increasingly attracting the attention of academic and industry researches. Unlike the face recognition environment addressed by existing works, face annotation problem under SNSs is differentiated in terms of daily updated images database, a limited number of training set and millions of users. Thus, conventional approach may not deal with these problems. In this paper, we proposed a face annotation method for sharing and publishing photographs that contain faces under a social network service using random projection, non-linear regression and representational state transfer. Our experiments on several databases show that the proposed method records an almost constant execution time with comparable accuracy of the PCA-SVM classifier.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.1387-1392
• /
• 2006

본 논문에서는 SVM (Support Vector Machine)을 기반으로 하여 인체의 뇌 하부구조인 해마에 대한 지능적 형상분석 방법을 제공한다. 일반적으로 의료 영상으로부터 해마의 형상 분석을 하기 위해서는 충분한 임상 데이터를 필요로 한다. 하지만 현실적으로 많은 양의 표본들을 얻는 것이 쉽지 않기 때문에 전문가의 지식을 기반으로 한 작업이 수반되어야 한다. 결국 이러한 요소들이 분석 작업을 어렵게 한다. 의학 기술이 복잡해 지면서 최근의 형상 분석 연구는 점차 통계적 모델을 기반으로 진행되고 있다. 본 연구에서는 해마로부터 고해상도의 매개변수형 모델을 만들어 형상 표현으로 이용하고, 집단간 분류 작업에 SVM 알고리즘을 적용하는 지능적 분석 방법을 구현한다. 우선 메쉬 데이터로부터 물리변형모델 기반의 매개변수 모델을 구축하고, PDM (point distribution model) 방법을 적용하여 두 집단을 대표하는 평균 모델을 생성한다. 마지막으로 SVM 기반의 이진 분류기를 구축하여 집단간 분류 작업을 수행한다. 구현한 모델링 방법과 분류기의 성능을 평가하기 위하여 본 연구에서는 네 가지 커널 함수 (linear, radial basis function, polynomial, sigmoid)들을 적용한다. 본 논문에서 제시한 매개변수형 모델은 다양한 형태의 의료 데이터로부터 보편적인 3차원 모델을 생성하고, 또한 모델의 전역적, 국부적인 특징들을 복합적으로 표현할 수 있기 때문에 통계적 형상분석에 적합하다. 그리고 SVM 기반의 분류기는 적은 수의 학습 데이터로부터 정상인 해마 집단과 간질 환자 집단간의 정확한 분류를 가능하게 한다.

• Journal of Satellite, Information and Communications
• /
• v.9 no.3
• /
• pp.59-63
• /
• 2014

This paper describes the basic principles of 3D broadcast system and proposes new 3D broadcast technology that reduces the amount of data by applying CS(Compressed Sensing). Differences between Sampling theory and the CS technology concept were described. Recently proposed CS algorithm AMP(Approximate Message Passing) and CoSaMP(Compressive Sampling Matched Pursuit) were described. This paper compared an accuracy between two algorithms and a calculation time that image data compressed and restored by these algorithms. As result determines a low complexity algorithm for 3D broadcast system.

• Proceedings of the KSME Conference
• /
• 2001.06e
• /
• pp.536-541
• /
• 2001

A high resolution digital cinematic Particle Image Velocimetry(PIV) has been developed. The system consists of a high speed CCD camera, a continuous Ar-ion laser and a computer with camera controller. To improve the spatial resolution, we adopt a Recursive Technique for velocity interrogation. At first, we obtain a velocity vector for a larger interrogation window size based on the conventional two-frame cross-correlation PIV analysis using the FFT algorithm. Based on the knowing velocity information, more spatially resolved velocity vectors are obtained in the next iteration step with smaller interrogation windows. The correct velocity vector at the first step is found to be critical, so we apply a Multiple Correlation Validation(MCV) technique in order to decrease the spurious vectors. The MCV technique turns out to improve SNR(Signal to Noise Ratio) of the correlation table. The developed cinematic PIV method has been applied to the measurement of the unsteady flow characteristics of a Rushton turbine mixer. A total of 3,245 instantaneous velocity vectors were successfully obtained with 4 ms time resolution. The acquired spatial resolution corresponds the performance of the conventional high resolution digital PIV system using a $1K{\times}1K$ CCD camera.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.43 no.2 s.308
• /
• pp.11-19
• /
• 2006

In this paper, we propose a method to estimate camera extrinsic parameter based on projective and permutation invariance point features. Because feature informations in previous research is variant to c.:men viewpoint, extraction of correspondent point is difficult. Therefore, in this paper, we propose the extracting method of invariant point features, and new matching method using similarity evaluation function and Graham search method for reducing time complexity and finding correspondent points accurately. In the calculation of camera extrinsic parameter stage, we also propose two-stage motion parameter estimation method for enhancing convergent degree of LM algorithm. In the experiment, we compare and analyse the proposed method with existing method by using various indoor images to demonstrate the superiority of the proposed algorithms.

• Journal of Digital Convergence
• /
• v.13 no.12
• /
• pp.209-215
• /
• 2015

In the past, there have been many theory and algorithms for vehicle tracking. But the time complexity of many feature point matching methods for vehicle tracking are exponential. Also, object segmentation and detection algorithms presented for vehicle tracking are exhaustive and time consuming. Therefore, we present the fast and efficient two stages method that can efficiently track the many moving vehicles on the road. The first detects the vehicle plate regions and extracts the feature points of vehicle plates. The second associates the feature points between frames using dynamic programming.

• Journal of Environmental Impact Assessment
• /
• v.10 no.1
• /
• pp.39-47
• /
• 2001

본 연구는 지리정보시스템 및 원격탐사기법을 응용하여 북한의 자연환경을 조사하기 위한 기초연구로서 수행되었으며, 과거 약 20년 동안의 평양 및 남포지역의 산림면적의 변화 및 경관구조 변화측면에서의 산림 파편화 양상을 조사하였다. 조사자료로는 Landsat MSS 및 TM 영상의 NDVI값을 이용하였으며, 보다 정확한 피복분류를 위하여 변형된 Cluster-Busting 알고리즘을 활용하여 산림과 비산림지역으로 단순화시켜 분석하였다. 경관구조의 변화를 살피기 위해서 조각밀도, 형태 및 핵심내부지역의 면적 등의 경관지수(Landscape Indices)를 활용하였다. 분석과정을 거쳐서 도출된 결론은 다음과 같다. 첫째, Cluster-busting 방법을 활용한 토지피복 분류결과 87.3%의 총 분류 정확도를 얻었으며, Binary Map을 이용한 변화감지(Change Detection)기법 또한 그 결과가 정확한 것으로 판단되었다. 둘째, '79년에서 '98년에 이르는 기간동안, 평양의 경우 '79년 산림면적의 15%, 그리고 남포지역의 경우 14%가 감소하였다. 셋째, 경관지수를 이용하여 북한 산림의 파편화 변화를 조사한 결과 산림조각의 개수는 늘어나고 조각의 평균면적 및 핵심내부면적은 감소하였으며 조각크기의 다양성 또한 낮아졌다. 산림조각 형태지수 또한 매우 증가하였는데 이러한 결과들은 평양 및 남포지역의 산림조각이 파편화되고 그 형태 또한 불규칙적이며 복잡하게 변화하였음을 보여주고 있다.

• Journal of Broadcast Engineering
• /
• v.1 no.2
• /
• pp.108-117
• /
• 1996

In this paper we propose a Wavelet- based compression algorithm of HDTV video signals for the purpose of recording HDTV signals in the digital VTR. Comparing to the DCT- based compression method, which only yields unrecognizable DCT coefficients, the low frequency components of Wavelet coefficients maintain recognizable spatial domain information. So, it is more suitable for various VTR operations such as editing and multi-speed mode operations. Also, the adopted Wavelet filter can be Implemented with simple shift operations, which can reduce the computational complexities substantially. The quality of reconstructed HDTV signals with a 4:1 compression ratio turns out to be good enough for the studio use.

• Journal of Broadcast Engineering
• /
• v.23 no.3
• /
• pp.431-436
• /
• 2018

In this paper, a simple but effective algorithm to reduce the computational complexity of thumbnail generation and to improve image quality without aliasing artifacts is proposed. For the high speed decoding, the proposed algorithm performs partial decoding per $4{\times}4$ boundary in TU(Transform Unit), and preforms TU boundary in PU(Prediction Unit). The proposed method defines the weights based on intra prediction directions and estimates the thumbnail pixel by using that weights. this method remains thumbnail extraction time and improves thumbnail image quality compared with conventional algorithms.