• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2006.06a
• /
• pp.41-44
• /
• 2006

본 논문은 안면영상의 인증 요소로 사용되는 눈동자영역을 SIFT를 이용하여 추출해내는 방법을 제안하고 있다. 모델이 되는 눈동자영상과 추출 하고자 하는 입력영상의 SET 결과인 Keypoint descriptor를 이용하여 각각의 특징벡터를 구성하고 서로 정합한 후 두 특징 점들 사이에 affine transform이 존재하는지 판단하여 반수 이상에 대응하는 특징 점들에 대해 동일한 affine transform이 존재할 경우 눈동자영역이라고 판단하였다. 실험결과 학습과정이 없으므로 기존의 인식방법 보다 빠르게 영역을 추출하는 결과를 얻을 수 있었다.

• Journal of Information Display
• /
• v.5 no.1
• /
• pp.41-46
• /
• 2004

In this paper, a new stereo 3D moving-target tracking system using the MSE (mean square error) and BPEJTC (binary phase extraction joint transform correlator) algorithms is proposed. A moving target is extracted from the sequential input stereo image by applying a region-based MSE algorithm following which, the location coordinates of a moving target in each frame are obtained through correlation between the extracted target image and the input stereo image by using the BPEJTC algorithm. Through several experiments performed with 20 frames of the stereo image pair with $640{\times}480$ pixels, we confirmed that the proposed system is capable of tracking a moving target at a relatively low error ratio of 1.29 % on average at real time.

• Proceedings of the KIEE Conference
• /
• 1997.07e
• /
• pp.1903-1904
• /
• 1997

This paper was dealt with the experiments of the traverse flow and excited CW $CO_2$ Laser. The discharge and laser output characteristics of CW $CO_2$ laser has been investigated. A stable resonator was adopted for the power extraction from the length of 0.9m discharge region with the gap length of 15mm. The maximum output power of about 100W was obtained at the first experiment with the total filling pressure of 30torr, $CO_2$ : $N_2$ : He = 1:3:16.

• Proceedings of the KIEE Conference
• /
• 1997.07b
• /
• pp.507-509
• /
• 1997

One of the most difficult tasks in the process of car license plate is the extraction of each character from within license plate region. This paper presents a real-time recognition of car licence number using neural network in parking lot. The feature parameters of letters and numbers of license plate are extracted by thinning algorithm. Both feature parameters are used to train neural networks for the image recognition.

• Journal of the Optical Society of Korea
• /
• v.15 no.2
• /
• pp.152-154
• /
• 2011

We propose a novel method which detects convex and concave corners using radial projection. The sum of two neighbors' differences at the local maxima or minima of the radial projection is compared with the angle threshold for detecting corners. In addition, the use of oriented bounding box trees and partial circles makes it possible to detect the corners of complex shapes. The experimental results show that the proposed method can separately detect the convex and concave corners, and that this method is scale invariant.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.6
• /
• pp.91-103
• /
• 1994

As CAD systems are introduced in the filed of engineering design, the necessities for automatic drawing input are increased . In this paper, we propose a method for realizing automatic drawing input by separation of text regions and graphic regions, extraction of line vectors from graphic regions, and recognition of circular arcs and circles from line vectors. Sizes of isolated regions, on a drawing are used for separating text regions and graphic regions. Thinning and maximum allowable error method are used to extract line vectors. And geometric structures of line vectors are analyzed to recognize circular arcs and circles. By processing text regions and graphic regions separately, 30~40% of vector information can be reduced. Recognition of circular arcs and circles can increase the utilization of automatic drawing input function.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.447-448
• /
• 2006

In this paper, we propose tracking algorithm using multi resolution in wavelet transform domain. This algorithm consists of two steps. The first step is feature extraction that is select feature-points using 1-level wavelet transform in ROI (Region of Interest). The other step is feature tracking. Based on multi resolution of wavelet transform, we estimate a displacement between current frame and next frame on the basis of selected feature-points. Experimental results show that the proposed algorithm confirmed a better performance than a centroid tracking and correlation tracking.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.877-878
• /
• 2006

In this paper, we propose a skin lesion detection to develop the system of fluorescence image analysis to identify the fluorescence of topical methyl aminolevulinate(MAL) idduced PpIX in patients with BCC accurately. By fluorescence image analysis we define the border between tumo and tumor-free areas on fluorescence image after topical application of MAL ointment. We excised both the tumor and peri-tumoral areas widely from the 10 patients with BCC, and divided tissue samples into 3 area, such as tumor area, suspected tumor area, tumor-free area, respectively. Our proposed method migt play a role as an adjunctive tool to define the border between tumor and tumor-free areas for Mohs' micrographic surgery.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.13 no.1
• /
• pp.34-42
• /
• 2013

Human faces have been broadly studied in digital image and video processing fields. An appearance-based method, the adaptive boosting learning algorithm using integral image representations has been successfully employed for face detection, taking advantage of the feature extraction's low computational complexity. In this paper, we propose a face-detection postprocessing method that equalizes instantaneous facial regions in an efficient hardware architecture for use in real-time multimedia applications. The proposed system requires low hardware resources and exhibits robust performance in terms of the movements, zooming, and classification of faces. A series of experimental results obtained using video sequences collected under dynamic conditions are discussed.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.04a
• /
• pp.441-445
• /
• 2005

얼굴인식기술이 인증 및 보안을 위한 도구로 활용되고 있지만 입력영상의 상태 즉, 복잡한 배경과 조명환경에 따라 적용할 수 있는 범위가 제약적일 수밖에 없다. 본 논문에서는 이러한 제약을 최소화하기 위한 방법과 좀 더 정확한 얼굴 영역 검출을 위한 기법을 제시한다. 제안된 방법은 움직임에 기반 한 에지 차영상을 이용하여 얼굴 윤곽을 검출한 후 이를 X와 Y축의 프로파일을 이용하여 얼굴영역을 예측한다. 그리고 얼굴의 피부 색상 정보와 특징 구성요소인 눈, 코, 입 등의 특징적인 요소의 에지정보를 이용하여 수직적으로 이를 구분한 후 얼굴인지 아닌지를 판별한다. 제안된 알고리즘은 다양한 배경 및 조명등의 많은 환경적 요인에 따른 입력영상에서도 매우 안정적으로 적용됨을 실험을 통해 확인하였다.