• Journal of the Korea Society of Computer and Information
• /
• v.16 no.3
• /
• pp.99-107
• /
• 2011

The feature extraction of asterias amurensis by using patterns is difficult to extract all the concave and convex features of asterias amurensis nor classify concave and convex. Concave and convex as important structural features of asterias amurensis are the features which should be found and the classification of concave and convex is also necessary for the recognition of asterias amurensis later. Accordingly, this study suggests the technique to extract the features of concave and convex, the main features of asterias amurensis. This technique classifies the concave and convex features by using the multi-directional linear scanning and form the candidate groups of the concave and convex feature points and decide the feature points of the candidate groups and apply convex hull algorithm to the extracted feature points. The suggested technique efficiently extracts the concave and convex features, the main features of asterias amurensis by dividing them. Accordingly, it is expected to contribute to the studies on the recognition of asterias amurensis in the future.

• Journal of the Institute of Convergence Signal Processing
• /
• v.6 no.1
• /
• pp.15-22
• /
• 2005

In this paper, we propose a face recognition system by using the CCD color image. We first get the face candidate image by using YCbCr color model and adaptive skin color information. And we use it initial curve of active contour model to extract face region. We use the Eye map and mouth map using color information for extracting facial feature from the face image. To obtain center point of Log-polar image, we use extracted facial feature from the face image. In order to obtain feature vectors, we use extracted coefficients from DCT and wavelet transform. To show the validity of the proposed method, we performed a face recognition using neural network with BP learning algorithm. Experimental results show that the proposed method is robuster with higher recogntion rate than the conventional method for the rotation and scale variant.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.11
• /
• pp.137-146
• /
• 2015

In this paper, we propose a pothole detection method using various features of intensity and motion. Segmentation, decision steps of pothole detection are processed according to the values which are derived from feature characteristics. For segmentation using intensity, we use a binarization method using histogram to separate pothole region from background. For segmentation using motion, we filter using high pass filter and get standard deviation value. This value is divided by regression value according to camera environment such as photographing angle, height, velocity, etc. We get binary image by histogram based binarization. For decision, candidate regions are decided whether pothole or not using comparison of candidate and background's features. Experimental results show that our proposed pothole detection method has better results than existing methods and good performance in discrimination between pothole and similar patterns.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.2
• /
• pp.79-86
• /
• 2024

In this paper, a novel visual tracking method which can utilize the context of object regions is presented. Conventional methods have the inherent problem of treating all candidate regions independently, where the tracker could not successfully discriminate regions with similar appearances. This was due to lack of contextual modeling in a given scene, where all candidate object regions should be taken into consideration when choosing a single region. The goal of the proposed method is to encourage feature exchange between candidate regions to improve the discriminability between similar regions. It improves upon conventional methods that only consider a single region, and is implemented by employing the MLP-Mixer model for enhanced feature exchange between regions. By implementing channel-wise, inter-region interaction operation between candidate features, contextual information of regions can be embedded into the individual feature representations. To evaluate the performance of the proposed tracker, the large-scale LaSOT dataset is used, and the experimental results show a competitive AUC performance of 0.560 while running at a real-time speed of 65 fps.

• Spatial Information Research
• /
• v.4 no.2
• /
• pp.131-146
• /
• 1996

The purpose of this study is to extract the main locational factors to affect the location of fastfood stores in Kangnam -Gu, Seoul by using Geographic Information Systems. The Franchise system, which came to be employed in full scale since 1990, now enjoying the booming period, especially fast food industry. The procedure of this research has four steps. First, the spatial distribution of fastfood chain stores in Seoul is analyzed by the land use and road map. Second, the spatial variations of fastfood stores in dong districts of Seoul is explained by multiple regression analysis. Third, the main locational factors to affect the location of fast food stores in Kangnam -Gu are extracted by demand and supply sides using GIS technologies. Finally, the potentiallocational zones where are selected by extracted locational factors are compared with the actual distribution of f astf ood stores in Kangnam - Gu. In terms of demand side, the main locational factors include commercial and service facilities, sub¬ways and bus stops which have a lot of pedestrians, and large apartment districts that have high popu¬lation densities. In terms of supply sides, the main locational factors include land use types and land value, accessibility. After fast food chain stores of Kangnam -Gu are overlaid final potentiallocational zones extracted by demand and supply sides of locational factors, it can be identified that over 80 % of fastfood stores are located in the potentiallocational zone. In conclusion, this study shows that spatial analysis functions may potentially be improved using GIS technologies. However there are still difficulties for the locational analysis for service industries to col¬lect appropriate data in terms of computerized base maps as well as consumer behavior and store characteristics itself.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.1
• /
• pp.94-104
• /
• 2007

Iris recognition is biometric technology which uses a unique iris pattern of user in order to identify person. In the captured iris image by conventional iris recognition camera, it is often the case with eyelid occlusion, which covers iris information. The eyelids are unnecessary information that causes bad recognition performance, so this paper proposes robust algorithm in order to detect eyelid. This research has following three advantages compared to previous works. First, we remove the detected eyelash and specular reflection by linear interpolation method because they act as noise factors when locating eyelid. Second, we detect the candidate points of eyelid by using mask in limited eyelid searching area, which is determined by searching the cross position of eyelid and the outer boundary of iris. And our proposed algorithm detects eyelid by using parabolic hough transform based on the detected candidate points. Third, there have been many researches to detect eyelid, but they did not consider the rotation of eyelid in an iris image. Whereas, we consider the rotation factor in parabolic hough transform to overcome such problem. We tested our algorithm with CASIA Database. As the experimental results, the detection accuracy were 90.82% and 96.47% in case of detecting upper and lower eyelid, respectively.

• Journal of Internet Computing and Services
• /
• v.10 no.6
• /
• pp.99-111
• /
• 2009

The result of Image segmentation, an indispensable process in image processing, significantly affects the analysis of an image. Despite the significance of image segmentation, it produces some problems when the variation of pixel values is large, or the boundary between background and an object is not clear. Also, these problems occur frequently when many objects in an image are placed very close by. In this paper, when the shape of objects in an image is circular, we proposed an algorithm which segment an each object in an image using the geometric characteristic of circular shape. The proposed algorithm is composed of 4 steps. First is the boundary edge extraction of whole object. Second step is to find the candidate points for further segmentation using the boundary edge in the first step. Calculating the representative circles using the candidate points is the third step. Final step is to draw the line connecting the overlapped points produced by the several erosions and dilations of the representative circles. To verify the efficiency of the proposed algorithm, the algorithm is compared with the three well-known cell segmentation algorithms. Comparison is conducted by the number of segmented region and the correctness of the inner segment line. As the result, the proposed algorithm is better than the well-known algorithms in both the number of segmented region and the correctness of the inner segment line by 16.7% and 21.8%, respectively.

• Journal of Astronomy and Space Sciences
• /
• v.22 no.4
• /
• pp.393-408
• /
• 2005

We developed and compared two automatic algorithms for moving object detections in the YSTAR-NEOPAT sky survey program. One method, called starlist comparison method, is to identify moving object candidates by comparing the photometry data tables from successive images. Another method, called image subtraction method, is to identify the candidates by subtracting one image from another which isolates sources moving against background stars. The efficiency and accuracy of these algorithms have been tested using actual survey data from the YSTAR-NEOPAT telescope system. For the detected candidates, we performed eyeball inspection of animated images to confirm validity of asteroid detections. Main conclusions include followings. First, the optical distortion in the YSTAR-NEOPAT wide-field images can be properly corrected by comparison with USNO-B1.0 catalog and the astrometric accuracy can be preserved at around 1.5 arcsec. Secondly, image subtraction provides more robust and accurate detection of moving objects. For two different thresholds of 2.0 and $4.0\sigma$, image subtraction method uncovered 34 and 12 candidates and most of them are confirmed to be real. Starlist comparison method detected many more candidates, 60 and 6 for each threshold level, but nearly half of them turned out to be false detections.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.11a
• /
• pp.178-181
• /
• 2013

본 논문에서는 SVC와 MVC의 부호화 구조를 결합하여 구현된 스케일러블 다시점 비디오 부호화의 움직임 추정 기법과 DPB를 위한 GPB 기반의 RPL (reference picture list) 설계를 제안한다. 제안된 움직임 추정 기법에서는 부호화 과정에서 필요한 예측 부호화의 성능 향상을 위해서 서로 다른 시점 (view)의 픽처 정보를 참조픽처의 후보로서 사용한다. 또한, B픽처 예측의 경우 HEVC에서 사용하는 HEVC GPB 기술을 통해 참조화면에서 두 개의 움직임 벡터를 활용한다. 제안된 움직임 예측 구조에 의해서 압축된 비디오 데이터의 크기를 감소시켜 압축 효율을 증대시킬 수 있다. 다양한 실험을 통해서 제안된 예측 구조를 적용함으로써 스케일러블 다시점 비디오 부호화에서의 압축 효율의 향상을 얻어낼 수 있음을 확인하였다.

• Proceedings of the KIEE Conference
• /
• 2006.07d
• /
• pp.2083-2084
• /
• 2006

본 논문에서는 조명에 변화에 강인하고 기존의 퍼지 색상 필터보다 정확하고 빠른 얼굴 감지 알고리즘 이용하여 얼굴을 인식하고 얼굴로부터 특징점(눈, 눈썹, 입)틀을 추출하고 추출된 특징점을 이용하여 감성을 판별하는 방법을 제안한다. 향상된 얼굴 인식 기술이란 퍼지 색상 필터의 단점이 영상의 크기와 성능에 따라 처리속도가 느려지는 것을 보완하기 위하여 최소한의 규칙을 사용하여 얼굴 후보 영역을 선별 적용하여 얼굴영역을 추출하는 기법을 말한다. 이렇게 추출된 얼굴영역에서 감정이 변화 할 때 가장 두드러지게 변화를 나타내는 눈, 눈썹 그리고 입의 특징점을 이용하여 감성을 분류한다.