• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.5
• /
• pp.1240-1250
• /
• 2010

In this paper, an eye tracking method is presented by using on iterated spatial moment adapting weighted gray level that can accurately detect and track user's eyes under the complicated background. The region of face is detected by using Haar-like feature before extracting region of eyes to minimize an region of interest from the input picture of CCD camera. And the region of eyes is detected by using eigeneye based on the eigenface of Principal component analysis. Also, feature points of eyes are detected from darkest part in the region of eyes. The tracking of eyes is achieved correctly by using iterated spatial moment adapting weighted gray level.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10b
• /
• pp.739-741
• /
• 2004

얼굴인식 기술 분야에 있어서 Haar-like feature를 이용한 얼굴 검출 알고리즘은 많은 관련 알고리즘 중에 매우 빠른 트레이닝 시간과 처리속도 향상의 장점을 가지고 있다 그러므로 특히 동영상에서의 얼굴 검출에서 유용하게 쓰일 수 있다. 이러한 방법으로 검출된 얼괄 데이터는 HMM(Hidden Markov Model)알고리즘을 이용하여 이미 트레이닝된 얼굴 데이터베이스와의 비교를 통해 얼굴인식에 있어서 가장 확률이 높은 사람을 본인의 얼굴로 인증하는 신원 확인 시스템을 구현할 수 있게 된다. 신원 확인 시스템에 있어서 얼굴 검출 율이나 신원 확인 성공률은 모두 학습 과정에 의해 결정되기 때문에 얼마나 많은 학습을 효율적으로 하느냐에 따라 성능이 좌우된다. 이러한 시스템은 카메라에 얼굴을 보여주는 것만으로 신원 확인이 가능하기 때문에 번거로운 신원 확인 과정을 거쳐야 하는 다른 시스템 구조에 비해 매우 편리한 기능을 제공할 수 있다.

• Proceedings of the KIEE Conference
• /
• 2009.07a
• /
• pp.1855_1856
• /
• 2009

본 논문은 Haar-like 마스크와 유사한 특징을 갖지만 두 사각형 영역의 크기와 위치를 제한하지 않는 분리된 두 사각 특징 마스크를 이용한 Adaboost 기반 얼굴검출 알고리즘을 제안한다. 기존의 Haar-like 특징이 단순히 두 사각 영역의 화소값들의 차를 구함으로써 계산이 용이하나 인접한 두 사각 영역으로 한정함으로써 고품질 특징을 얻기 어렵다. 이런 Haar-like 특징마스크의 내재된 문제점을 개선하기 위해, 제안하는 특징 마스크는 다양한 크기와 분리된 두 사각 영역을 갖는 형태로 고품질의 특징을 얻는다. 고품질의 특징은 Adaboost 알고리즘의 약 분류기(weak classifier)의 성능을 학습단계부터 높여 전반적으로 얼굴 검출 알고리즘의 성능을 향상시킨다. 제안하는 분리된 두 사각 특징을 이용한 adaboost 기반 얼굴검출 기법의 우수성을 다양한 실험을 통해 검증하였다.

• Annual Conference of KIPS
• /
• 2015.10a
• /
• pp.735-736
• /
• 2015

최근 잇따른 보안사고의 발생주기가 짧아지고 그 피해는 점점 심각해져만 가고 있다. 이에 맞춰 여러 대응방안이 나오고 있지만 새로운 취약점은 계속해서 발견되고 있다. 그에 대응하여 개인을 식별할 새로운 기술인 보안과 관련하여 영상처리기술이 사용되고 있으며 현재도 활발히 연구중에 있다. 본 논문은 현재 사용되는 얼굴인식 알고리즘인 Adaboost-CAMShift 그리고 Adaboost-Haar-like Feature의 기술들을 비교 분석 하고 소개하는 것을 목표로 한다.

• Annual Conference of KIPS
• /
• 2014.04a
• /
• pp.861-864
• /
• 2014

본 논문은 혀 미각 영역별 분석을 통해 신체의 이상 여부에 대한 정보를 제공하는 설진 진단 시스템의 첫 단계로 얼굴 영상에서 혀 영역을 검출하는 실험을 통하여 미각 영역별 분석의 기반을 다진다. 제안하는 알고리즘은 혀 영상을 획득한 후, Haar-like Feature를 이용하여 혀를 검출한다. 검출된 혀 영역은 HSV컬러모델의 특징을 이용하여 이진화 한 후, Connected Component Labeling을 이용하여 혀 영역 분리한다. 한방병원의 환자들의 혀 사진 100장을 이용하여 90%의 검출률을 확인하였다.

• Journal of IKEEE
• /
• v.18 no.4
• /
• pp.536-543
• /
• 2014

Terrain classification is still a challenging issue in image processing, especially with high resolution satellite images. The well-known obstacles include low accuracy in the detection of targets, especially for the case of man-made structures, such as buildings and roads. In this paper, we present an efficient approach to classify and detect building footprints, foliage, grass and road from high resolution grayscale satellite images. Our contribution is to build a strong classifier using AdaBoost based on a combination of co-occurrence and Haar-like features. We expect that the inclusion of Harr-like feature improves the classification performance of the man-made structures, since Haar-like feature is extracted from corner features and rectangle features. Also, the AdaBoost algorithm selects only critical features and generates an extremely efficient classifier. Experimental result indicates that the classification accuracy of AdaBoost classifier is much higher than that of the conventional classifier using back propagation algorithm. Also, the inclusion of Harr-like feature significantly improves the classification accuracy. The accuracy of the proposed method is 98.4% for the target detection and 92.8% for the classification on high resolution satellite images.

• Journal of IKEEE
• /
• v.22 no.1
• /
• pp.46-52
• /
• 2018

The problem of traffic light detection and recognition has recently become one of the most important topics in various researches on autonomous driving. Most algorithms are based on colors to detect and recognize traffic light signals. These methods have disadvantage in that the recognition rate is lowered due to the change of the color of the traffic light, the influence of the angle, distance, and surrounding illumination environment of the image. In this paper, we propose machine learning based detection and recognition algorithm using shape information to solve these problems. Unlike the existing algorithms, the proposed algorithm detects and recognizes the traffic signals based on the morphological characteristics of the traffic lights, which is advantageous in that it is robust against the influence from the surrounding environments. Experimental results show that the recognition rate of the signal is higher than those of other color-based algorithms.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.4
• /
• pp.377-385
• /
• 2015

This paper proposes a novel forward vehicle tracking algorithm based on the WMIL(Weighted Multiple Instance Learning) equipped with a particle filter. In the proposed algorithm Haar-like features are used to train a vehicle object detector to be tracked and the location of the object are obtained from the recognition result. In order to combine both the WMIL to construct the vehicle detector and the particle filter, the proposed algorithm updates the object location by executing the propagation, observation, estimation, and selection processes involved in particle filter instead of finding the credence map in the search area for every frame. The proposed algorithm inevitably increases the computation time because of the particle filter, but the tracking accuracy was highly improved compared to Ababoost, MIL(Multiple Instance Learning) and MIL-based ones so that the position error was 4.5 pixels in average for the videos of national high-way, express high-way, tunnel and urban paved road scene.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.2
• /
• pp.55-64
• /
• 2010

This paper proposes a face detection method using Free Rectangle feature which possesses a quick execution time and a high efficiency. The proposed mask of Free Rectangle feature is composed of two separable rectangles with the same area. In order to increase the feature diversity, Haar-like feature generally uses a complex mask composed of two or more rectangles. But the proposed feature mask can get a lot of very efficient features according to any position and scale of two rectangles on the feature window. Moreover, the Free Rectangle feature can largely reduce the execution time since it is defined as the only difference of the sum of pixels of two rectangles irrespective of the mask type. Since it yields a quick detection speed and good detection rates on real world images, the proposed face detection method based on Adaboost algorithm is easily applied to detect another object by changing the training dataset.

• The Journal of the Korea Contents Association
• /
• v.10 no.12
• /
• pp.84-90
• /
• 2010

In this paper, we propose a quadratic discriminant analysis based approach for improving the discriminating strength of weak classifiers based on simple Haar-like features that were used in the Viola-Jones object detection framework. Viola and Jones built a strong classifier using a boosted ensemble of weak classifiers. However, their single threshold (or decision boundary) based weak classifier is sub-optimal and too weak for efficient discrimination between object class and background. A quadratic discriminant analysis based approach is presented which leads to hyper-quadric boundary between the object class and background class, thus realizing multiple thresholds based weak classifiers. Experiments carried out for car detection using 1000 positive and 3000 negative images for training, and 500 positive and 500 negative images for testing show that our method yields higher classification performance with fewer classifiers than single threshold based weak classifiers.