CNN을 활용한 램프 점등 이미지 균일도 정량화 방안 연구

이수현(Soohyeon Lee),이수봉(Soobong Lee) 한국자동차공학회 2022 한국자동차공학회 학술대회 및 전시회 Vol.2022 No.11

The lighting image of a car lamp is one of the important factors influencing the exterior design of a car, but it is difficult to find an evaluation method for quantitatively evaluating the lighting image. If there is a method of quantitatively evaluating how uniform the lighting image of the lamp is, it will be easy to manage the quality of the lighting image. This study attempted to determine whether it is possible to quantitatively evaluate the lighting image data of a car lamp using a deep learning algorithm that learned a good lighting image and a poor lighting image of a car lamp using deep learning of image data. CNN algorithms were used for deep learning of image data, and learning was conducted using images of various automobile position and tail lamp lighting. The learned algorithm classified good lighting images and poor lighting image data with a probability of more than 90% for the new lighting image data, and was able to provide lighting uniformity scores for the image data. The method of learning deep learning algorithms could help develop evaluation methods that are difficult to develop in conventional ways or quantitative evaluation methods for complex data.

한반도의 분단과 통일

이수봉(Lee, Soobong) 총신대학교 평화통일연구소 2021 평화통일연구 Vol.3 No.1

‘통일에 대한 논의는 분단에서 출발한다.’는 분석은 대부분의 학자들이 공감하고 있다. 이 말의 의미는 통일은 분단의 문제를 극복하지 않고 이루어질 수 없다는 의미이며, 동시에 적극적 의미에서 통일이라는 긍정적 가치를 실현하는 의미가 있다고 할 것이다. 그러므로 본고에서는 분단과 통일에 대해 논하면서, 통일로 나아가는 이 시대의 한반도가 나가야 할 방향을 모색하고자 한다. 사회 역사적 측면에서 분단은 국민의 분단, 영토의 분단, 주권의 분단으로 이루어졌고, 분단은 외적 원인과 내적 원인으로 나누어 분석하였다. 분단의 책임을 외적 원인에만 돌리는 것은 운명론으로 흐를 가능성이 있어 주의할 필요가 있고, 내적 원인을 보완하여 분단에 대한 민족의 책임과 역사적 주체성을 확립할 필요가 있다. 통일은 분단시대의 통일 개념과 통일을 향한 전환기의 통일 개념과 통일 후의 통일 개념을 나누어 분석하였다. 통일이라는 같은 단어를 사용하여도, 분단시대의 통일은 점령적 통일이고, 통일시대의 통일은 분단을 청산하고 민족이 하나로 회복하자는 통일이며, 통일 후의 통일은 정의로운 사회 구축에 방점이 있다. 분단과 통일에 대한 이해는 삶의 정황 속에서 통일 담론을 형성하는 데 기여할 것으로 기대한다. ‘The discussion of reunification begins with division.’ The analysis is shared by most scholars. This means that unification cannot be achieved without overcoming the problems of division. At the same time, this means realizing the positive value of unification. Therefore, we will discuss division and unification in this paper. And the researchers want to find a way for unification on the Korean Peninsula. In terms of social and historical aspects, the division consisted of the division of the people, the division of territory, and the division of sovereignty. The division was analyzed by dividing it into the causes of the trace and the internal causes. Attributing the responsibility of division to external causes will lead to fatalism. It is necessary to establish a national responsibility and a historical self-reliance for division by supplementing internal causes from the cause of division. Unification was analyzed by dividing the concept of unification in the divided period, of unification in the unifying era and of unification after unification. Even using the same word, the unification in the divided era is an occupying unification. The unification in the unifying era is aimed at ending division and restoring the nation as one. The unification after reunification has a focus on building a just society. It is expected that understanding of division and unification will contribute to forming a discourse on unification in the context of life in the Korean Peninsula.

천리안위성 2A호 위성영상을 위한 영상융합기법의 비교평가

이수봉,최재완,Lee, Soobong,Choi, Jaewan 한국측량학회 2022 한국측량학회지 Vol.40 No.4

기후변화 감시에 위성 자료 활용을 위해 GCOS (Global Climate Observing System)는 시공간 해상도, 시간 변화에 따른 안정성, 불확실도 등의 요구사항을 제시하고 있다. 천리안위성 2A호의 경우, 센서의 한계로 인해 산출물들이 공간해상도 조건에 충족하지 못하는 경우가 많다. 따라서 본 연구에서는 영상융합 기법들을 천리안위성 2A호 영상에 적용하여 산출물 생성 시 활용될 수 있는 최적의 기법을 찾고자 한다. 이를 위해 CS (Component Substitution), MRA (Multiresolution Analysis), VO (Variational Optimization), DL (Deep Learning)에 포함되는 총 6가지 영상융합 기법을 활용하였다. DL의 경우 합성적(Synthesis) 특성 기반 방법을 훈련자료 구축에 사용하였다. 합성적 특성 기반 방법의 과정은 PAN (Panchromatic)과 MS (Multispectral) 영상의 공간해상도 차이만큼 두 영상의 해상도를 낮춰 융합 영상을 생성한 후 원본 MS 영상과 비교한다. 합성적 특성 기반 방법은 공간해상도를 저하시킨 PAN 영상과 MS 영상 간 기하 특성이 같아야 사용자가 원하는 수준의 융합 영상을 제작할 수 있다. 하지만, 훈련자료 구축 시 비유사성이 존재하기에 이를 최소화하는 방법으로 무작위 비율을 활용한 PSGAN 모델(PSGAN_RD)을 추가로 활용하였다. 융합 영상의 검증은 일관성(consistency) 및 합성적 특성 기반 정성적, 정량적 분석을 수행하였다. 분석 결과, 영상융합 알고리즘 중 GSA가 공간 유사도를 나타내는 평가지수에서 가장 높은 수치를 보였으며, 분광 유사도를 나타내는 지수들은 PSGAN_RD 모델의 정확도가 가장 높았다. 융합 영상의 공간 및 분광 특성을 모두 고려한다면 PSGAN_RD 모델이 천리안위성 2A호 산출물 제작에 가장 최적일 것으로 판단하였다. In order to detect climate changes using satellite imagery, the GCOS (Global Climate Observing System) defines requirements such as spatio-temporal resolution, stability by the time change, and uncertainty. Due to limitation of GK-2A sensor performance, the level-2 products can not satisfy the requirement, especially for spatial resolution. In this paper, we found the optimal pan-sharpening algorithm for GK-2A products. The six pan-sharpening methods included in CS (Component Substitution), MRA (Multi-Resolution Analysis), VO (Variational Optimization), and DL (Deep Learning) were used. In the case of DL, the synthesis property based method was used to generate training dataset. The process of synthesis property is that pan-sharpening model is applied with Pan (Panchromatic) and MS (Multispectral) images with reduced spatial resolution, and fused image is compared with the original MS image. In the synthesis property based method, fused image with desire level for user can be produced only when the geometric characteristics between the PAN with reduced spatial resolution and MS image are similar. However, since the dissimilarity exists, RD (Random Down-sampling) was additionally used as a way to minimize it. Among the pan-sharpening methods, PSGAN was applied with RD (PSGAN_RD). The fused images are qualitatively and quantitatively validated with consistency property and the synthesis property. As validation result, the GSA algorithm performs well in the evaluation index representing spatial characteristics. In the case of spectral characteristics, the PSGAN_RD has the best accuracy with the original MS image. Therefore, in consideration of spatial and spectral characteristics of fused image, we found that PSGAN_RD is suitable for GK-2A products.

대비감도함수를 이용한 자동차용 헤드램프 노면 균일도 평가 연구

이용철(Yongcheol Lee),이수봉(Soobong Lee) 한국자동차공학회 2022 한국자동차공학회 학술대회 및 전시회 Vol.2022 No.11

Drivers can feel tired and be obstructed to look forward because of uneven road light. We needed SW to predict and prevent these uneven lights at the design stage, and to test uniformity in the product. In this paper, a contrast sensitivity function was used to develop uniformity test SW of road light of the headlamp. In order to a person to identify an object, there must be a difference in brightness. The contrast threshold to which bright and dark can be identified varies with the spatial frequency. It is called a contrast sensitivity function (CSF) that formula of the relationship between contrast and spatial frequency for human eyes. If there is a difference in brightness with respect to the drivers viewing angle that is less than the contrast threshold, we may feel that the road light is uniform because we cannot distinguish the brightness difference. if the difference in brightness is larger than the contrast threshold, the area of road light would be felt unevenly. The viewing angle can be calculated from the distance information. Contrast can be calculated from illuminance distribution of the road light. The contrast threshold of road light felt unevenly was obtained through experiments. The test subjects checked the position felt unevenly at the real road with HAL and LED headlamp at night. The same lamps were compared through simulation of the spatial frequency and contrast of the position felt unevenly after the photometric scan at testing room. Through this, it was defined that the CSF for road light uniformity test. We developed a program that extracts the illuminance distribution along the horizontal and vertical 0.25° step line from the drivers view and calculates the contrast and spatial frequency. The program marks uneven areas in the road light pattern. As a result, a uniformity visualization program similar to the road uniformity seen in the driver’s view. Through a uniformity visualization program. It is possible to predict and prevent uneven light at the design stage. When a problem occurs in the product, we can easily determine the area that needs to improve.

천리안 위성 자료를 활용한 합성곱 순환 신경망 기반 태풍 최대풍속 산출

이민형(Minhyung Lee),이수봉(Soobong Lee),이정환(Junghwan Lee),한성원(Sung Won Han) 대한산업공학회 2019 대한산업공학회지 Vol.45 No.4

Input and Output for Usage : The Three-Tier Structure to Curriculum for Graduate Interaction Design Program

Aqua C.Y. Chen,이수봉(Lee Soobong),서종환(Seo Jonghwan) 한국디지털디자인학회 2009 디지털디자인학연구 Vol.9 No.4

This paper is about to discuss over its development of interaction technologies what changes has the "interaction design" paradigms brought to the discipline of design education? Started from an overview of a recently implemented Digital Interaction Design Integration Plan of Ming Chuan University in Taiwan where different core competences was integrated from School of Design and Information. 24 design proposals of different interaction design aspects from the departments participated (Product Design Digital Media Commercial and Information Engineering) will be analyzed and explained accordingly. The comparisons of 24 design proposals from this project showed that interaction installations are not equal to interactive product; however interaction design is not interface design because some don't even need any visual interface. This experience poses the questions that over the rapid development of interaction technology what's the influence to the discipline of design education? Moreover what should be the proper curriculum of interaction design program in industrial design education? Through the investigations of established graduate interaction design curricula worldwide (e.g. Designing interaction at RCA London; Designing for Interaction at TU Delft and the other schools in US Europe and Asia) we suggest a framework for planning an interaction design curriculum within graduate art and design education. The main contribution is to set the stage for an urgently needed discussion on how design education needs to adapt to the emerging interaction design contexts we are actually designing for today.

램프 열 성능 예측의 신뢰도 향상을 위한 해석 프로세스 개선

최봉근(Bongkeun Choi),이철우(Chulwoo Lee),고만석(Manseok Ko),이수봉(Soobong Lee),김억(Ok Kim) 한국자동차공학회 2012 한국자동차공학회 학술대회 및 전시회 Vol.2012 No.11

As lamp system is considered as one of the most important components that determine the image of car, most of manufacturers place great emphasis on lamp design for their success in the development of new car. In this research various techniques of advanced CFD analysis were suggested so as to simulate more physical phenomena and improve the accuracy of estimating lamp performance and support the effective success in esthetic design of lamp system after all.