• Journal of the Society of Naval Architects of Korea
• /
• v.51 no.3
• /
• pp.203-211
• /
• 2014

To raise the industrial competitiveness in the field of ship-building, it is crucially important that the yard should use production facilities and working space effectively. Among the related works, the management of tremendous blocks' number, the limited area of assembly shops and inefficient personnel and facility management still need to be improved in terms of being exposed to a lot of problems. To settle down these conundrums, the various strategies of block arrangement on the assembly floors have been recently presented and in the results, have increasingly began to be utilized in practice. However, it is a wonder that the sampled or approximated block shapes which usually are standardized projections or the geometrically convex contour only have been prevailed until now. In this study, all parts including the panel, stiffeners, outer shells, and all kinds of outfitting equipment are first extracted using the Volume Primitive plug-in module from the ship customized CAD system and then, the presented system constructs a simpler and more compact ship data structure and finally generates the novel projected contours for the block arrangement system using the adaptive concave hull algorithm.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.6
• /
• pp.1356-1363
• /
• 2005

In the gyrocompass system, the use of the fiber optic gyroscope(FOG) makes this traditional system considerably attractive because it has strong points in terms of weight, power, warming-up time, and cost. In this paper, a novel digital north-finding method based upon an FOG, which can be applied to the gyrocompass system, is proposed. The analytical model for the earth signal of the FOG is described, and the earth signals passed through lock-in amplifiers are modeled. Additionally, a north-finding algorithm using two lock-in amplifier outputs is developed, and the proposed method is organized by the developed algorithm. Simulation results are included to verify the performance of the proposed method.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.1123-1126
• /
• 2005

여러 임베디드 시스템 운영체제 중에서 임베디드 리눅스는 다양한 오픈 소스 S/W를 사용할 수 있고, 다양한 임베디드 시스템에 이식할 수 있다는 장점 때문에 널리 사용되고 있다. 하지만, 임베디드 리눅스는 리눅스의 기본 접근제어 메커니즘인 임의적 접근제어(Discretionary Access Control, DAC) 기법을 그대로 사용하고 있어서 사용자의 Identity가 도용 당하거나 Trojan Horse와 같은 프로그램이 설치될 경우, 접근제어가 효력을 상실하게 된다는 결점을 가지고 있다. 더욱 문제가 되는 것은 DAC의 특성상, 프로세스가 필요 이상의 과도한 특권을 가지고 실행되며, 그 결과 잘못된 프로세스가 그 자신과 관계 없는 프로그램이나 운영체제의 커널마저 손상시키는 결과를 낳을 수 있다는 것이다. 이에 따라 보다 강건한 접근제어 메커니즘에 대한 연구의 필요성이 대두되고 있다. 본 논문에서는 임베디드 리눅스 운영체제의 접근제어 메커니즘이 가지고 있는 보안적 결점에 대해서 알아보고, 이 결점을 보완하기 위해 타입 강제(Type Enforcement, TE) 기법을 사용함으로써, 임베디드 시스템에 적합하면서 강력한 접근제어를 제공할 수 있는 안전한 임베디드 리눅스 시스템에 대한 설계 모델을 보여주고자 한다.

• Smart Media Journal
• /
• v.9 no.2
• /
• pp.92-98
• /
• 2020

This paper proposes a vehicle license plate recognition system using light weight deep learning models without high-end server. The proposed license plate recognition system consists of 3 steps: [license plate detection]-[character area segmentation]-[character recognition]. SSD-Mobilenet was used for license plate detection, ResNet with localization was used for character area segmentation, ResNet was used for character recognition. Experiemnts using Samsung Galaxy S7 and LG Q9, accuracy showed 85.3% accuracy and around 1.1 second running time.

• Journal of the Korean Society of Industry Convergence
• /
• v.25 no.6_3
• /
• pp.1285-1291
• /
• 2022

In this paper, we propose a low-cost, low-power embedded environment-based deep learning lightweight model for input images to recognize laundry management codes. Laundry franchise companies mainly use barcode recognition-based systems to record laundry consignee information and laundry information for laundry collection management. Conventional laundry collection management systems using barcodes require barcode printing costs, and due to barcode damage and contamination, it is necessary to improve the cost of reprinting the barcode book in its entirety of 1 billion won annually. It is also difficult to do. Recognition performance is improved by applying the VGG model with 7 layers, which is a reduced-transformation of the VGGNet model for number recognition. As a result of the numerical recognition experiment of service parts drawings, the proposed method obtained a significantly improved result over the conventional method with an F1-Score of 0.95.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.2
• /
• pp.51-58
• /
• 2023

This paper proposes a deep learning method for estimating the heart rate from facial videos. Our proposed method estimates remote photoplethysmography (rPPG) signals to predict the heart rate. Although there have been proposed several methods for estimating rPPG signals, most previous methods can not be utilized in low-power single board computers due to their computational complexity. To address this problem, we construct a lightweight student model and employ a knowledge distillation technique to reduce the performance degradation of a deeper network model. The teacher model consists of 795k parameters, whereas the student model only contains 24k parameters, and therefore, the inference time was reduced with the factor of 10. By distilling the knowledge of the intermediate feature maps of the teacher model, we improved the accuracy of the student model for estimating the heart rate. Experiments were conducted on the UBFC-rPPG dataset to demonstrate the effectiveness of the proposed method. Moreover, we collected our own dataset to verify the accuracy and processing time of the proposed method on a real-world dataset. Experimental results on a NVIDIA Jetson Nano board demonstrate that our proposed method can infer the heart rate in real time with the mean absolute error of 2.5183 bpm.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.6
• /
• pp.319-325
• /
• 2023

In this study, we proposed a method for hyperparameter optimization in the building and training of a deep learning model designed to process point cloud data collected by a millimeter-wave radar system. The primary aim of this study is to facilitate the deployment of a baseline model in resource-constrained IoT devices. We evaluated a RadHAR baseline deep learning model trained on a public dataset composed of point clouds representing five distinct human activities. Additionally, we introduced a coarse-to-fine hyperparameter optimization procedure, showing substantial potential to enhance model efficiency without compromising predictive performance. Experimental results show the feasibility of significantly reducing model size without adversely impacting performance. Specifically, the optimized model demonstrated a 3.3% improvement in classification accuracy despite a 16.8% reduction in number of parameters compared th the baseline model. In conclusion, this research offers valuable insights for the development of deep learning models for resource-constrained IoT devices, underscoring the potential of hyperparameter optimization and model size reduction strategies. This work contributes to enhancing the practicality and usability of deep learning models in real-world environments, where high levels of accuracy and efficiency in data processing and classification tasks are required.

• Journal of IKEEE
• /
• v.28 no.3
• /
• pp.397-404
• /
• 2024

This paper presents the design and implementation results of a Raspberry-Pi-based embedded system with an FPGA accelerator that can detect and classify objects using an FMCW radar sensor for preventing door collision accidents in vehicles. The proposed system performs a radar sensor signal processing and a deep learning processing that classifies objects into bicycles, automobiles, and pedestrians. Since the CNN algorithm requires substantial computation and memory, it is not suitable for embedded systems. To address this, we implemented a lightweight deep learning model, BNN, optimized for embedded systems on an FPGA, and verified the results achieving a classification accuracy of 90.33% and an execution time of 20ms.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.9
• /
• pp.444-452
• /
• 2024

This study proposes a deep learning architecture optimized for fire detection derived through Layer Importance Evaluation. In order to solve the problem of unnecessary complexity and operation of the existing Convolutional Neural Network (CNN)-based fire detection system, the operation of the inner layer of the model based on the weight and activation values was analyzed through the Layer Importance Evaluation technique, the layer with a high contribution to fire detection was identified, and the model was reconstructed only with the identified layer, and the performance indicators were compared and analyzed with the existing model. After learning the fire data using four transfer learning models: Xception, VGG19, ResNet, and EfficientNetB5, the Layer Importance Evaluation technique was applied to analyze the weight and activation value of each layer, and then a new model was constructed by selecting the top rank layers with the highest contribution. As a result of the study, it was confirmed that the implemented architecture maintains the same performance with parameters that are about 80% lighter than the existing model, and can contribute to increasing the efficiency of fire monitoring equipment by outputting the same performance in accuracy, loss, and confusion matrix indicators compared to conventional complex transfer learning models while having a learning speed of about 3 to 5 times faster.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.4
• /
• pp.58-66
• /
• 2009

Recently the emerging demands of the light-video encoder promotes lots of research efforts on DVC (Distributed Video Coding). As an appropriate video compression method, DVC has been studied, and Wyner-Ziv (WZ) video compression is its one representative structure. The WZ encoder splits the image into two kinds of frames, one is key frame which is compressed by conventional intra coding, and the other is WZ frame which is encoded by WZ coding. The WZ decoder decodes the key frame first, and estimates the WZ frame using temporal correlation between key frames. Estimated WZ frame (Side Information) cannot be the same as the original WZ frame due to the absence of the WZ frame information at decoder. As a result, the difference between the estimated and original WZ frames are regarded as virtual channel noise. The WZ frame is reconstructed by removing noise in side information. Therefore precise noise estimation produces good performance gain in WZ video compression by improving error correcting capability by channel code. But noise cannot be estimated precisely at WZ decoder unless there is good WZ frame information, and generally it is estimated from the difference of corresponding key frames. Also the estimated noise is limited by comparing with frame level noise to reduce the uncertainty of the estimation method. However these methods cannot provide good noise estimation for every frame or each bit plane. In this paper, we propose a noise nodel selection method which chooses a better noise model for each bit plane after generating candidate noise models. Experimental result shows PSNR gain up to 0.8 dB.