• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.401-410
• /
• 2021

Popular in discriminant classification analysis, k-nearest neighbor classification methods have limitations that do not reflect the local characteristic of the data, considering only the number of fixed neighbors. Considering the local structure of the data, the adaptive nearest neighbor method has been developed to select the number of neighbors. In the analysis of high-dimensional data, it is common to perform dimension reduction such as random projection techniques before using k-nearest neighbor classification. Recently, an ensemble technique has been developed that carefully combines the results of such random classifiers and makes final assignments by voting. In this paper, we propose a novel discriminant classification technique that combines adaptive nearest neighbor methods with random projection ensemble techniques for analysis on high-dimensional data. Through simulation and real-world data analyses, we confirm that the proposed method outperforms in terms of classification accuracy compared to the previously developed methods.

• Journal of Digital Contents Society
• /
• v.16 no.4
• /
• pp.605-613
• /
• 2015

This paper addresses the problem to make sparse the projection matrix in pattern recognition method. Recently, the size of computer program is often restricted in embedded systems. It is very often that developed programs include some constant data. For example, many pattern recognition programs use the projection matrix for dimension reduction. To improve the recognition performance, very high dimensional feature vectors are often extracted. In this case, the projection matrix can be very big. Recently, RSR(roated sparse regression) method[1] was proposed. This method has been proved one of the best algorithm that obtains the sparse matrix. We propose three methods to improve the RSR; outlier removal, sampling and elastic net RSR(E-RSR) in which the penalty term in RSR optimization function is replaced by that of the elastic net regression. The experimental results show that the proposed methods are very effective and improve the sparsity rate dramatically without sacrificing the recognition rate compared to the original RSR method.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.499-502
• /
• 2012

문양이나 패턴이 없는 민 무늬의 물체에는 동일점 정합에 의해 3 차원 계측을 하는 스테레오 정합방식을 적용할 수 없다. 본 논문에서는 난수 발생 함수로 제작한 랜덤 칼라 패턴을 대상물체에 투영하여, 대상 물체 표면에 특징적인 문양을 인위적으로 생성시키는 것에 의해 민 무늬의 물체를 스테레오 정합법으로 측정하는 방식을 제안한다. 투사된 패턴으로 자체 문양을 지니게 된 물체를 스테레오 카메라로 촬영하였고, 동일점 정합은 전역 스테레오 정합 방식의 일종인 TRW 방식에 의한 컬러 매칭 방식을 사용하였다. 제안된 방식은 원형의 흰색 석고상 3 차원 계측에 적용되었고, 안정적이고 정확한 스테레오 정합 계측 결과를 보였다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.37-40
• /
• 2004

본 논문에서는 POSC 기반의 3D 메쉬 워터마킹 방법을 제안하였다. 제안한 방법에서는 3D 메쉬를 두 가지의 제약 조건 집합으로 수렴 조건을 만족할 때까지 반복 투영한다. 이들 집합은 워터마크를 삽입하기 위한 강인성 집합 및 비가시성 집합으로 구성된다. 원 모델없이 워터마크를 추출하기 위하여 제안한 방법에서는 워터마크가 삽입되는 위치 정보 및 결정치를 이용한다. 실험 결과로부터 제안한 방법이 메쉬 간단화, 절단, 아핀 변환, 및 랜덤 잡음 첨가 등의 공격에 우수한 강인성을 가짐을 확인하였다.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.10
• /
• pp.48-58
• /
• 1998

We present a method that can recognize face images using independent component analysis that is used mainly for blind sources separation in signal processing. We assumed that a face image can be expressed as the sum of a set of statistically independent feature images, which was obtained by using independent component analysis. Face recognition was peformed by projecting the input image to the feature image space and then by comparing its projection components with those of stored reference images. We carried out face recognition experiments with a database that consists of various varied face images (total 400 varied facial images collected from 10 per person) and compared the performance of our method with that of the eigenface method based on principal component analysis. The presented method gave better results of recognition rate than the eigenface method did, and showed robustness to the random noise added in the input facial images.

• Korean Journal of Optics and Photonics
• /
• v.33 no.6
• /
• pp.267-274
• /
• 2022

In this paper, we report the design and fabrication of binary diffractive optical elements (DOEs) for random-dot-pattern projection for Schlieren imaging. We selected the binary phase level and a pitch of 10 ㎛ for the DOE, based on cost effectiveness and ease of manufacture. We designed the binary DOE using an iterative Fourier-transform algorithm with binary phase optimization. During initial optimization, we applied a computer-generated pseudorandom dot pattern of uniform intensity as a target pattern, and found significant intensity nonuniformity across the field. Based on the evaluation of the initial optimization, we weighted the target random dot pattern with Gaussian profiles to improve the intensity uniformity, resulting in the improvement of uniformity from 52.7% to 90.8%. We verified the design performance by fabricating the designed binary DOE and a beam projector, to which the same was applied. The verification confirmed that the projector produced over 10,000 random dot patterns over 430 mm × 430 mm at a distance of 5 meters, as designed, but had a slightly less uniformity of 84.5%. The fabrication errors of the DOE, mainly edge blurring and spacing errors, were strong possibilities for the difference.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2005.05a
• /
• pp.848-853
• /
• 2005

3D geometric modeling of an object of interest has been intensively investigated in many fields including CAD/CAM and computer graphics. Traditionally, CAD and geometric modeling tools are widely used to create geometric models that have nearly the same shape of 3D real objects or satisfy designers intent. Recently, with the help of the reverse engineering (RE) technology, we can easily acquire 3D point data from the objects and create 3D geometric models that perfectly fit the scanned data more easily and fast. In this paper, we present 3D geometric model generation based on a stereo vision system (SVS) using random pattern projection. A triangular mesh is considered as the resulting geometric model. In order to obtain reasonable results with the SVS-based geometric model generation, we deal with many steps including camera calibration, stereo matching, scanning from multiple views, noise handling, registration, and triangular mesh generation. To acquire reliable stere matching, we project random patterns onto the object. With experiments using various random patterns, we propose several tips helpful for the quality of the results. Some examples are given to show their usefulness.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.11C
• /
• pp.1592-1599
• /
• 2004

In this paper, we proposed the 3D mesh watermarking using projection onto convex sets (POCS). 3D mesh is projected iteratively onto two constraint convex sets until it satisfy the convergence condition. These sets consist of the robustness set and the invisibility set that designed to embed watermark Watermark is extracted without original mesh by using the decision values and the index that watermark is embedded. Experimental results verified that the watermarked mesh have the robustness against mesh simplification, cropping, affine transformation, and vertex randomization as well as the invisibility.

• Journal of Digital Convergence
• /
• v.11 no.12
• /
• pp.345-352
• /
• 2013

In this paper, a depth estimate method is proposed using GPU(Graphics Processing Unit) in monocular image. a monocular image is a 2D image with missing 3D depth information due to the camera projection and we used a monocular cue to recover the lost depth information by the projection present. The proposed algorithm uses an energy function which takes a variety of cues to create a more generalized and reliable depth map. But, a processing time is late because energy function is defined from the various monocular cues. Therefore, we propose a depth estimate method using GPGPU(General Purpose Graphics Processing Unit). The objective effectiveness of the algorithm is shown using PSNR(Peak Signal to Noise Ratio), a processing time is decrease by 61.22%.

• Journal of the Institute of Convergence Signal Processing
• /
• v.12 no.4
• /
• pp.242-249
• /
• 2011

In this paper, the correspondence matching method of stereo images was proposed by means of sampling projective transformation matrix in planar region of scene. Though this study is based on RANSAC, it does not use uniform distribution by random sampling in RANSAC, but use multi non-uniform computed from difference in positions of feature point of image or templates matching. The existing matching method sampled that the correspondence is presumed to correct by use of the condition which the correct correspondence is almost satisfying, and applied RANSAC by matching the correspondence into one to one, but by sampling in stages in multi probability distribution computed for image in the proposed method, the correct correspondence of high probability can be sampled among multi correspondence candidates effectively. In the result, we could obtain many correct correspondence and verify effectiveness of the proposed method in the simulation and experiment of real images.