• Journal of electromagnetic engineering and science
• /
• 제18권3호
• /
• pp.206-211
• /
• 2018

In this paper, we proposed a two-level feature vector fusion technique to improve the performance of target classification. The proposed method combines feature vectors of the early-time region and late-time region in the first-level fusion. In the second-level fusion, we combine the monostatic and bistatic features obtained in the first level. The radar cross section (RCS) of the 3D full-scale model is obtained using the electromagnetic analysis tool FEKO, and then, the feature vector of the target is extracted from it. The feature vector based on the waveform structure is used as the feature vector of the early-time region, while the resonance frequency extracted using the evolutionary programming-based CLEAN algorithm is used as the feature vector of the late-time region. The study results show that the two-level fusion method is better than the one-level fusion method.

• 한국콘텐츠학회논문지
• /
• 제7권1호
• /
• pp.48-57
• /
• 2007

눈좌표 검출은 얼굴 인식 및 관련된 응용 분야 등에서 필요한 작업이다. 현재까지 보고된 대부분의 눈좌표 검출 방법은 성공적인 적용을 위해서는 여전히 정확도 및 검출 속도의 개선을 필요로 한다. 본 논문에서는 다중스케일 가버 특징 벡터 모델 기반의 개선된 눈좌표 검출 방법을 제안한다. 제안된 방법은 먼저 다운샘플링된 입력 얼굴 이미지에서 초기 눈좌표에서의 가버 특징 벡터와 해당 스케일의 눈 모델 번치와의 가버젯 유사도를 이용하여 눈좌표를 추정한다. 이후 추정된 눈좌표를 상위 스케일의 얼굴 이미지에서의 눈좌표 초기값으로 취하고 상위 스케일 얼굴 이미지에서 같은 방법으로 눈좌표를 찾으며, 이를 반복적으로 하여 최종적으로 원래 얼굴 이미지에서의 눈좌표를 확정한다. 실험을 통해, 본 논문에서 제안한 다중스케일 가버 특징 벡터 모델 기반 눈좌표 검출 방법이 계산량은 크게 증가시키지 않으면서 기존 연구들에서 보고된 다른 눈좌표 검출 방법에 비해 정확도가 개선된 검출 방법임을 확인하였다.

• 한국인공지능학회지
• /
• 제11권4호
• /
• pp.1-8
• /
• 2023

In this paper, we propose a black ice detection platform framework using Convolutional Neural Networks (CNNs). To overcome black ice problem, we introduce a real-time based early warning platform using CNN-based architecture, and furthermore, in order to enhance the accuracy of black ice detection, we apply a multi-scale dilation convolution feature fusion (MsDC-FF) technique. Then, we establish a specialized experimental platform by using a comprehensive dataset of thermal road black ice images for a training and evaluation purpose. Experimental results of a real-time black ice detection platform show the better performance of our proposed network model compared to conventional image segmentation models. Our proposed platform have achieved real-time segmentation of road black ice areas by deploying a road black ice area segmentation network on the edge device Jetson Nano devices. This approach in parallel using multi-scale dilated convolutions with different dilation rates had faster segmentation speeds due to its smaller model parameters. The proposed MsCD-FF Net(2) model had the fastest segmentation speed at 5.53 frame per second (FPS). Thereby encouraging safe driving for motorists and providing decision support for road surface management in the road traffic monitoring department.

• 한국연소학회:학술대회논문집
• /
• 한국연소학회 1997년도 제15회 KOSCO SYMPOSIUM 논문집
• /
• pp.163-173
• /
• 1997

This article presents an application of a large-scale structural mixing model (Broadwell et al. 1984) to the blowout of turbulent reacting jets discharging perpendicularly into an unconfined cross air-flow. In an analysis of a common stability curve, a plausible explanation can be made that the phenomenon of blowout is related only to the mixing time scale of the two flows. The most notable observation is that the blowout distance is traced at fixed positions at all times according to the velocity ratio R. Measurements of the lower blowout limits in the liftable flame agree qualitatively with the blowout parameter ${\varepsilon}$, proposed by Broadwell et al. Good agreement between the results calculated by a modified blowout parameter ${\varepsilon}^'$ and experimental results confirms the important effect of a large-scale structure in specifying the stabilization feature of blowouts.

• 한국진공학회지
• /
• 제4권S1호
• /
• pp.190-195
• /
• 1995

The feature scale model for selective chemical vapor deopsition process was proposed and the simulation was performed to study the selectivity and uniformity of deposited thin film using Monte Carlo method and string algorithm. The effect of model parameters such as sticking coefficient, aspect ratio, and surface diffusion coefficient on the deposited thin film pattern was improved for lower sticking coefficient and higher aspect ratio. It was revealed that the selectivity loss ascrives to the surface diffusion. Different values of sticking coefficients on Si and on SiO2 surface greatly influenced the deopsited thin film profile. In addition, as the lateral wall angle decreased, the selectively deposited film had improved uniformity except the vicinity of trench wall. The optimum eondition for the most flat selective film deposition pattern is the case with low sticking coefficient and slightly increased surface diffusion coefficient.

• 대한임베디드공학회논문지
• /
• 제19권2호
• /
• pp.73-81
• /
• 2024

Recently, mobile manipulators have been utilized in agriculture industry for weed removal and harvest automation. This paper proposes a domain adaptive fruit detection method for harvest automation, by utilizing OWL-ViT model which is an open-vocabulary object detection model. The vision-language model can detect objects based on text prompt, and therefore, it can be extended to detect objects of undefined categories. In the development of deep learning models for real-world problems, constructing a large-scale labeled dataset is a time-consuming task and heavily relies on human effort. To reduce the labor-intensive workload, we utilized a large-scale public dataset as a source domain data and employed a domain adaptation method. Adversarial learning was conducted between a domain discriminator and feature extractor to reduce the gap between the distribution of feature vectors from the source domain and our target domain data. We collected a target domain dataset in a real-like environment and conducted experiments to demonstrate the effectiveness of the proposed method. In experiments, the domain adaptation method improved the AP50 metric from 38.88% to 78.59% for detecting objects within the range of 2m, and we achieved 81.7% of manipulation success rate.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2004년도 Proceedings of ISRS 2004
• /
• pp.344-346
• /
• 2004

IKONOS 1m satellite imagery is particularly well suited for 3-D feature extraction and 1 :5,000 scale topographic mapping. Because the image line and sample calculated by given RPCs have the error of more than 11m, in order to be able to perform feature extraction and topographic mapping, rational polynomial coefficients(RPCs) camera model that are derived from the very complex IKONOS sensor model to describe the object-image geometry must be refined by several Ground Control Points(GCPs). This paper presents a quantitative evaluation of the geometric accuracy that can be achieved with IKONOS imagery by refining the offset and scaling factors of RPCs using several GCPs. If only two GCPs are available, the offsets and scale factors of image line and sample are updated. If we have more than three GCPs, four parameters of the offsets and scale factors of image line and sample are refined first, and then six parameters of the offsets and scale factors of latitude, longitude and height are updated. The stereo images acquired by IKONOS satellite are tested using six ground points. First, the RPCs model was refined using 2 GCPs and 4 check points acquired by GPS. The results from IKONOS stereo images are reported and these show that the RMSE of check point acquired from left images and right are 1.021m and 1.447m. And then we update the RPCs model using 4 GCPs and 2 check points. The RMSE of geometric accuracy is 0.621 m in left image and 0.816m in right image.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 정보 및 제어부문
• /
• pp.61-63
• /
• 2005

Retreiving indoor scene reference image from database using visual information is important issue in Robot Navigation. Scene matching problem in navigation robot is not easy because input image that is taken in navigation process is affinly distorted. We represent probabilistic framework for the feature matching between features in input image and features in database reference images to guarantee robust scene matching efficiency. By reconstructing probabilistic scene matching framework we get a higher precision than the existing feaure-feature matching scheme. To construct probabilistic framework we represent each image as Gaussian Mixture Model using Expectation Maximization algorithm using SIFT(Scale Invariant Feature Transform).

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2002년도 ICCAS
• /
• pp.96.3-96
• /
• 2002

A scheme of clarifying fault feature in the residual is given for model-based fault detection and diagnosis of control systems. It is based on the residual generation using a robust filter and the noise suppresion in test statistics of the residual by multi-scale discrete wavelet transform. By clarifying the fault feature in the residual, the difficulties of existing model based approaches via adopting a threshold can be overcomed and it has advantage of taking the false alarm and missed detection into acount at the same time, which can make the fault detection and diagnosis easy and correct. To show the effectiveness of our approach, the simulation results are illustrated for a linear syste...

• Journal of Information Processing Systems
• /
• 제19권5호
• /
• pp.673-687
• /
• 2023

Convolutional neural network-based deep learning technology is the most commonly used in image identification, but it requires large-scale data for training. Therefore, application in specific fields in which data acquisition is limited, such as in the military, may be challenging. In particular, the identification of ground weapon systems is a very important mission, and high identification accuracy is required. Accordingly, various studies have been conducted to achieve high performance using small-scale data. Among them, the ensemble method, which achieves excellent performance through the prediction average of the pre-trained models, is the most representative method; however, it requires considerable time and effort to find the optimal combination of ensemble models. In addition, there is a performance limitation in the prediction results obtained by using an ensemble method. Furthermore, it is difficult to obtain the ensemble effect using models with imbalanced classification accuracies. In this paper, we propose a transfer learning-based feature fusion technique for heterogeneous models that extracts and fuses features of pre-trained heterogeneous models and finally, fine-tunes hyperparameters of the fully connected layer to improve the classification accuracy. The experimental results of this study indicate that it is possible to overcome the limitations of the existing ensemble methods by improving the classification accuracy through feature fusion between heterogeneous models based on transfer learning.