• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.11a
• /
• pp.161-162
• /
• 2016

비슷한 코너의 모양을 가지는 다수의 글자가 포함된 문서 영상을 인식하는 일은 쉽지 않다. 일반적으로 성능이 우수하다고 알려진 SIFT 알고리즘은 코너를 기반으로 특징을 기술하는 알고리즘이기 때문에 각 글자가 비슷한 코너의 모양을 가지는 문서 영상 인식에서는 좋은 성능을 발휘하지 못한다. 반면, LLAH 는 각 단어의 크기를 알아내어 가우시안 필터와 이진화를 통해 단어를 하나의 점으로 나타내고 각 점과 점 사이의 기하 관계를 기술자로 표현하기 때문에 문서의 단어에서 점이 일관되게 추출된다면 좋은 인식 성능을 발휘한다. 그러나, 영상에서 단어의 크기를 알아내는 작업은 계산 측면에서 많은 비용을 필요로 한다. 이에 본 논문에서는 LLAH 를 사용하기 전에 반복적인 가우시안 필터와 이진화를 적용하여 단어의 크기를 알지 못하는 상황에서도 스케일에 강인하게 문서 영상을 인식할 수 있는 알고리즘을 제안한다.

• Journal of Internet Computing and Services
• /
• v.5 no.3
• /
• pp.57-62
• /
• 2004

In this paper, an efficient speaker identification based on robust vector quantizationprincipal component analysis (VQ-PCA) is proposed to solve the problems from outliers and high dimensionality of training feature vectors in speaker identification, Firstly, the proposed method partitions the data space into several disjoint regions by roust VQ based on M-estimation. Secondly, the robust PCA is obtained from the covariance matrix in each region. Finally, our method obtains the Gaussian Mixture model (GMM) for speaker from the transformed feature vectors with reduced dimension by the robust PCA in each region, Compared to the conventional GMM with diagonal covariance matrix, under the same performance, the proposed method gives faster results with less storage and, moreover, shows robust performance to outliers.

• Journal of Korea Multimedia Society
• /
• v.16 no.11
• /
• pp.1281-1287
• /
• 2013

In order to prevent illegal video copies, many video fingerprints have been developed. Video fingerprints should be robust from various video transformations and have high discriminative powers. In general, video fingerprints are generated from three feature spaces such as luminance, gradient, and DCT coefficients. However, there is a few study for the robustness and discrimination according to feature spaces. Thus, we analyzed the property of each feature space by video copy detion task with the robustness and the discrimination of video fingerprints. We generated three video fingerprints from these feature spaces using a same algorithm. In our test, a video fingerprint. based on DCT coefficient outperformed others because the discrimination of it was higher.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.4
• /
• pp.47-52
• /
• 2011

Constant false alarm rate (CFAR) is an automatic detection algorithm for active sonar system. Among several CFAR algorithms, ordered statistics (OS) CFAR has the best performance over cell averaging (CA), smallest of (SO), greatest of (GO) algorithms at non-homogeneous environments. However, OS CFAR has the disadvantage of bad detection performance in multiple target conditions. We suggest an ordered statistics ratio (OSR) CFAR algorithm that is robust to multiple target environments. The proposed and conventional schemes are compared with computer simulations.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2019.05a
• /
• pp.203-205
• /
• 2019

위성항법시스템의 안정적인 정보 제공의 요구사항은 점차 증가하고 있지만, 의도적인 전파교란 및 자연환경에 의하여 성능이 저하된다. eLoran 시스템은 위성항법시스템의 단점을 보완하기위한 대표적인 지상파항법시스템으로 고출력 신호로 의도적인 전파교란에 강인한 장점이 있다. eLoran 시스템에서 사용자는 E-field 또는 H-field 2가지 종류의 수신 안테나를 사용 환경에 따라 적합한 것을 사용한다. 접지가 필요없고 상대적으로 주변 전자장비의 잡음에 강인한 H-field안테나는 두 개의 루프로 구성되어 루프간의 위상과 이득차이로 일정한 원형의 지향성을 가지지 못한다. 그러므로 수신한 신호의 방향에 따라 측정치의 변화가 발생하므로 H-field 안테나를 사용시에는 수신하는 신호의 방향에 따른 오차를 제거해야한다. 본 논문에서는 H-field 안테나와 송신국간의 기하학적 방향오차에 따른 오차를 제거하기 위한 후처리 필터를 제안하였다. eLoran 모의 신호생성기를 활용하여 오차를 분석하고 모델링하여 제거하는 기법을 개발하였다. 제안한 기법을 검증하기 위하여 시뮬레이션과 차량실험을 통하여 오차를 확인하고 제거하여 성능향상을 확인하였다.

• Journal of Broadcast Engineering
• /
• v.27 no.6
• /
• pp.944-947
• /
• 2022

Due to the coronavirus pandemic, the wearing of a mask has been increasing worldwide; thus, the importance of image analysis on masked face images has become essential. Although head pose estimation can be applied to various face-related applications including driver attention, face frontalization, and gaze detection, few studies have been conducted to address the performance degradation caused by masked faces. This study proposes a new data augmentation that synthesizes the masked face, depending on the face image size and poses, which shows robust performance on BIWI benchmark dataset regardless of mask-wearing. Since the proposed scheme is not limited to the specific model, it can be utilized in various head pose estimation models.

• Journal of Digital Convergence
• /
• v.14 no.10
• /
• pp.271-276
• /
• 2016

In this paper, we propose robust face and eye detection algorithm under changing environmental condition such as lighting and pose variations. Generally, the eye detection process is performed followed by face detection and variations in pose and lighting affects the detection performance. Therefore, we have explored face detection based on Modified Census Transform algorithm. The eye has dominant features in face area and is sensitive to lighting condition and eye glasses, etc. To address these issues, we propose a robust eye detection method based on Gabor transformation and Features from Accelerated Segment Test algorithms. Proposed algorithm presents 27.4ms in detection speed with 98.4% correct detection rate, and 36.3ms face detection speed with 96.4% correct detection rate for eye detection performance.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.45 no.3
• /
• pp.123-134
• /
• 2008

In this paper, we propose a real-tin e face recognition system that is robust under various lighting conditions. Th Modified Census Transform algorithm that is insensitive to illumination variations is employed to extract local structure features. In a practical face recognition system, acquired images through a camera are likely to be blurred and some of them could be side face images, resulting that unacceptable performance could be obtained. To improve stability of a practical face recognition system, we propose a real-time algorithm that rejects unnecessary facial picture and makes use of recognition consistency between successive frames. Experimental results on the Yale database with large illumination variations show that the proposed approach is approximately 20% better than conventional appearance-based approaches. We also found that the proposed real-time method is more stable than existing methods that produces recognition result for each frame.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.5 no.4
• /
• pp.101-110
• /
• 1995

In this paper, a DC servo motor position control system based on acceleration control is proposed. The proposed control system consists of an acceleration controller and an auto-tuqing fuzzy PID controller. The auto-tuning fuzzy PID controller provides corrections for an acceleration reference to remove the effect of parametric uncertainties. And it comprises of the expert system which performs the automatic tuning of the PID controller parameters and the conventional PID controller. Expermental results demonstrate strate thi~tth e proposed overall control system has robust properties and good control performances with regard to unmeasurable disturbances and parameter variations. Therefore, the proposed control scheme enhances the applicability of an acceleration control approach and especially performs accurate position control under such an operating environment that model uncertainties exist and/or load, etc. change significantly.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.4
• /
• pp.528-533
• /
• 2010

This paper studied the training methods less affected by the emotional variation for the development of the robust speech recognition system. For this purpose, the effect of emotional variation on the speech signal and the speech recognition system were studied using speech database containing various emotions. The performance of the speech recognition system trained by using the speech signal containing no emotion is deteriorated if the test speech signal contains the emotions because of the emotional difference between the test and training data. In this study, it is observed that vocal tract length of the speaker is affected by the emotional variation and this effect is one of the reasons that makes the performance of the speech recognition system worse. In this paper, a training method that cover the speech variations is proposed to develop the emotionally robust speech recognition system. Experimental results from the isolated word recognition using HMM showed that propose method reduced the error rate of the conventional recognition system by 28.4% when emotional test data was used.