• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.6
• /
• pp.692-696
• /
• 2017

The shape of press components is becoming increasingly complex due to customer demands, process shortening and cost savings. In addition, the stability of the pressing process frequently varies during mass production due to the influence of many factors. In order to ensure the process stability, it is necessary to establish a process in which reproducibility is realized in tolerance, which is sufficient for advance study of shape, material, press, mold and lubrication. However, unforeseen changes in process parameters cause disruptions in production line shutdowns and production planning. In this paper, we introduce a method to monitor a real time process by applying a sensor to a press mold. A non-contact type sensor for measuring the flow of a sheet material and an example of an experiment using the optical sensor which is highly applicable to mass production are presented. An optical sensor was installed in a cylindrical drawing mold to test its potential application while changing the material, blank holder force, and drawing ratio. We also quantitatively determined that the flow of other sheet materials was quantified locally using a square drawing die and that the measured value was always smaller than the drawing depth due to the material elongation. Finally, we propose a field that can be used by attaching the sensor to the press mold. We hope that the consequent cost reduction will contribute to increasing global mold competitiveness.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.51.2-51.2
• /
• 2019

Multi-wavelength observations become very popular in astronomy. Even though there are some correlations among different sensor images, it is not easy to translate from one to the other one. In this study, we apply a deep learning method for image-to-image translation, based on conditional generative adversarial networks (cGANs), to solar images. To examine the validity of the method for scientific data, we consider several different types of pairs: (1) Generation of SDO/EUV images from SDO/HMI magnetograms, (2) Generation of backside magnetograms from STEREO/EUVI images, (3) Generation of EUV & X-ray images from Carrington sunspot drawing, and (4) Generation of solar magnetograms from Ca II images. It is very impressive that AI-generated ones are quite consistent with actual ones. In addition, we apply the convolution neural network to the forecast of solar flares and find that our method is better than the conventional method. Our study also shows that the forecast of solar proton flux profiles using Long and Short Term Memory method is better than the autoregressive method. We will discuss several applications of these methodologies for scientific research.

• The Journal of Korea Robotics Society
• /
• v.16 no.1
• /
• pp.49-55
• /
• 2021

Recently, a smart farm technology is drawing attention as an alternative to the decline of farm labor population problems due to the aging society. Especially, there is an increasing demand for automatic harvesting system that can be commercialized in the market. Pre-harvest crop detection is the most important issue for the harvesting robot system in a real-world environment. In this paper, we proposed a real-time tomato instance tracking algorithm by using deep learning and probability models. In general, It is hard to keep track of the same tomato instance between successive frames, because the tomato growing environment is disturbed by the change of lighting condition and a background clutter without a stochastic approach. Therefore, this work suggests that individual tomato object detection for each frame is conducted by YOLOv3 model, and the continuous instance tracking between frames is performed by Kalman filter and probability model. We have verified the performance of the proposed method, an experiment was shown a good result in real-world test data.

• Journal of the Korean Society of Laryngology, Phoniatrics and Logopedics
• /
• v.33 no.2
• /
• pp.77-82
• /
• 2022

Hunminjeongeum was made by the Great King Sejong, and composed of 17 consonant and 11 vowel letters. All the 28 letters were made according to the shape of vocal organ or space at the point of articulation for each letters. This review article focused on phonetic and phoniatric consideration for explanation of the designs of the middle vowel letters, especially three main vowel letters [ • (天, heaven), ㅡ (地, earth), ㅣ (人, human)] using video-fluoroscopic evaluation as well as computed tomography scanning, etc. During articulating / • / sound, a ball-like space at frontal portion of the oral cavity was found, tongue was contracted, and sound was deep (舌縮而聲深). During /ㅡ/ sound, a flat air space between oral tongue and hard palate was created. Tongue was slightly contacted neither deep nor shallow (舌小縮而聲不深不淺). During /ㅣ/ sound, tongue was not contacted and Sound is light (舌不縮而聲淺). Tongue was moved forward making longitudinal oro-pharyngeal air space. So, I'd like to suggest that we had better change the explanation drawing from a philosophical modeling to a more scientific modeling from real vocal tract space modeling during articulating middle vowels of Hunminjeongeum.

• Smart Media Journal
• /
• v.13 no.8
• /
• pp.16-25
• /
• 2024

Recently, there has been a growing emphasis on identifying both known and unknown diseases in plant disease recognition. In this task, a model trained only on images of known classes is required to classify an input image into either one of the known classes or into an unknown class. Consequently, the capability to recognize unknown diseases is critical for model deployment. To enhance this capability, we are considering three factors. Firstly, we propose a new logits-based scoring function for unknown scores. Secondly, initial experiments indicate that a compact feature space is crucial for the effectiveness of logits-based methods, leading us to employ the AM-Softmax loss instead of Cross-entropy loss during training. Thirdly, drawing inspiration from the efficacy of transfer learning, we utilize a large plant-relevant dataset, PlantCLEF2022, for pre-training a model. The experimental results suggest that our method outperforms current algorithms. Specifically, our method achieved a performance of 97.90 CSA, 91.77 AUROC, and 90.63 OSCR with the ResNet50 model and a performance of 98.28 CSA, 92.05 AUROC, and 91.12 OSCR with the ConvNext base model. We believe that our study will contribute to the community.

• Hwankyungkyoyuk
• /
• v.16 no.2
• /
• pp.39-50
• /
• 2003

Secondary school environmental education is offered mostly in extra-curricular activities, not included in the school curriculum. Therefore, drawing out its curriculum has been left to teachers' discretion. Teachers' beliefs in EE have much more influence on the selection of content of EE than any other psychological factors. It follows that the reality of school EE depends on what teachers believes about EE and which content of EE they select according to their beliefs. This study is to research the realities of the secondary school EE: examine teachers' beliefs about EE and analyse environmental views in their environment-related programs; examine whether teachers' programs reflect their beliefs and environmental views; identify the factors which teachers inhibit teachers from their expression of these beliefs. To identify what perspective was reflected to these programs, this study employs environmentalism. Environmentalism can be classified into 3 categories: environmental management; deep ecology; social ecology. Data were gathered by means of in-depth interviews with five teachers. To show teachers' beliefs clearly, the analysis of the data was conducted on the basis of 5 categories: 'ecological sensitivity', 'value education', 'behaviors', 'issues', 'perception of a cause of environmental problems, related to social structures'. These belifs reflect deep ecological perspective and social ecology. But these perspectives did not accord with those in programs. Most EE programs included only the deep ecological perspective. They didn't reflect their beliefs that students should perceive the cause of environmental problems in the social structure. The factors which teachers find as inhibiting or supporting their expression of these beliefs were analysed as following: (1) EE requests the change of curriculum and methods of instruction; (2) teachers should acquire 'environmental consciousness' and environmentrelated knowledge; (3) programs requests an administrative and financial support.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1949-1955
• /
• 2021

With the recent COVID-19 situation, countries are seriously feeling the need for medical personnel and their technologies. PDepending on the aging society, the number of medical staff is actually decreasing, and in order to solve this problem, research is needed to replace the part that does not require high expertise among actual medical practices performed by doctors. This paper describes and proposes actual research methods related to unmanned medical systems that use various deep learning image processing-based technologies to check the recovery status applicable to rehabilitation areas where medical staff should face patients directly. The proposed method replaces passive calculations such as a protractor or a method of drawing a line in a photograph, which is the method used for actual motion comparison. Since it is performed in real time, it helps to diagnose quickly, and it is easy for medical staff to provide necessary information because data on the degree of match of motion performance can be checked.

• Korean Journal of Materials Research
• /
• v.14 no.1
• /
• pp.1-8
• /
• 2004

Laser weldability of aluminized steels for deep drawing application has been investigated. Test coupons for Nd:YAG laser welding and $CO_2$ laser welding were prepared trom the commercial steels. According to the test results, total penetration and back bead width of aluminized steels were sensitive to the welding conditions. Bead width at the half thickness of the overlap joint, however, was rather constant. Laser weldability of aluminized steels was superior to that of zinc coated steel. Weld microstructure revealed that overlap zone adjacent to the fusion line was filled with coated materials, which was thought to be desirable to protect weld from crevice corrosion. The aluminum coated materials was also found in the weld metal. Practically no spattering was observed in the laser welding of aluminized steels even when the welding was performed without joint gap. In the welding of zinc coated steel, however, spattering was so severe that it was difficult to get the acceptable weld. Bead quality of aluminized steel laser weld was smooth and stable.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2009.10a
• /
• pp.254-258
• /
• 2009

In this paper, the low frequency vibration forming system is developed for micro-patterns formation on the metal substrate. many researchers have studied about micro-forming technologies such as micro deep drawing, press forming, forging, extrusion etc. for the formation of precise micro-patterns on the surface of metal substrates, multi-step forming process must be used to improve qualifies of the deformed patterns. Since the low frequency vibration forming system could easily deform the surface of metal substrates, several steps of multi-step forming process should be removed by using the low frequency vibration forming system. In order to find optimal process conditions, we have carried out low frequency vibration forming process with varying the vibration frequency from 110Hz to 500Hz.

• Journal of the Korean Society for Precision Engineering
• /
• v.4 no.1
• /
• pp.74-82
• /
• 1987

딥 드로오잉 시의 소재누르기판의 힘을 펀치의 변위 혹은 힘의 임의의 함수로서 제어 시킬 수 있는 서브프레스장치 및 유압장치를 제작하고 제어용 프로그램을 작성하여 시 험하였다. 제작한 장치를 인스트론 시험기에 설치하고 제어용 프로그램을 IBM퍼스널 컴퓨터를 사용하여 실행시킨 결과, 기계 제어 부분은 어셈블러언어로, 계산 부분은 고급 언어로 프로그램을 작성하여 링크시킨 후에 컴파일링하여 사용하는것이 제어의 응답성 및 정확도를 향상시킬 수 있음이 판명되었다. 또한, 소재누르기판의 힘의 제어에 필요한 이론 및 예비 결과들도 제시하였다.