• Journal of Multimedia Information System
• /
• 제5권2호
• /
• pp.87-90
• /
• 2018

Various approaches have been proposed to convert low dynamic range (LDR) to high dynamic range (HDR). Of these approaches, the Multi Scale Fusion (MSF) algorithm based on Laplacian pyramid decomposition is used in many applications and demonstrates its usefulness. However, the pyramid fusion technique has no means for controlling the luminance component because the total number of pixels decreases as the pyramid rises to the upper layer. In this paper, we extract the reflection light of the image based on the Retinex theory and generate the weight map by adjusting the reflection component. This weighting map is applied to achieve an MSF-like effect during image fusion and provides an opportunity to control the brightness components. Experimental results show that the proposed method maintains the total number of pixels and exhibits similar effects to the conventional method.

• 한국멀티미디어학회논문지
• /
• 제19권5호
• /
• pp.837-847
• /
• 2016

Tone mapping for High Dynamic Range(HDR) image provides matching human visual perception between real world scene and displayable devices. Recently, a tone mapping algorithm based on localized gamma correction is proposed. This algorithm is using human visual properties of contrast and colorfulness with background intensity, generating a weight map for gamma correction. However, this method have limitations of controlling enhancement region as well as generating halo artifacts caused by the weight map construction. To overcome aforementioned limitations, proposed algorithm in this paper modifies previous weight map, considering base layer intensity of input luminance channel. By determining enhancement region locally and globally based on base layer intensity, gamma values are corrected accordingly. Therefore, proposed algorithm selectively enhances local brightness and controls strength of edges. Subjective evaluation using z-score shows that our proposed algorithm outperforms the conventional methods.

• 한국식생활문화학회지
• /
• 제31권6호
• /
• pp.634-642
• /
• 2016

Chinese cabbage is produced and consumed as a main material for kimchi and as a staple vegetable in Korea throughout the year. However, due to environmental changes unbalance between supply and demand is repeated annually, requiring development of long-term storage technologies. Chinese cabbages, were harvested, put in plastic boxes, and precooled at $2^{\circ}C$ for 24 hours using forced air precooler. After precooling, Chinese cabbages were MAP-treated with 0.02 mm HDPE film and functional film and stored at low temperature ($1{\pm}0.5^{\circ}C$). The weight-loss rates after 9-weeks of storage were 8.47% in the control group, 4.07% in the HDPE film-treated group, and 3.07% in the functional film-treated group, respectively, suppressing weight loss. Trimming loss rate after 6-weeks of storage was 6.86% in the functional film MAP-treated group and lower than 7.50% in the control group. In the sensory test with 7 points as the limit of commodity, the control group lost it after 6-weeks of storage while the MAP-treated groups retained over 7 points. The functional film MAP-treated group showed over 6 points for processing as kimchi until 9-weeks of storage, proving that Pallet-type MAP storage is effective for extending storage life of spring Chinese cabbage.

• 한국지형공간정보학회:학술대회논문집
• /
• 한국지형공간정보학회 2005년도 아시아 태평양 국제 GSIS 학술발표회
• /
• pp.61-72
• /
• 2005

The landslide risk assessment process consists of hazard risk assessment and vulnerability analysis. landslide hazard risk is location dependent. Therefore, maps and spatial technologies such as GIS are very important components of the risk assessment process. This paper discusses the advantages of using GIS technology in the risk assessment process and illustrates the benefits through case studies of live projects undertaken. The goal of this study is to generate a map of landslide vulnerability map by analysis of static natural factors with GIS. A simple and efficient algorithm is proposed to generate a landslide potentialities map from DEM and existing maps. The categories of controlling factors for landslides, aspect of slope, soil, vegetation are defined. The weight values for landslide potentialities are calculated from AHP method. Slope and slope-direction are extracted from DEM, and soil informations are extracted from digital soil map. Also, vegetation informations are extracted from digital vegetation map. Finally, as overlaying, landslide potentialities map is made out, and it is verified with landslide place.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제52권5호
• /
• pp.277-282
• /
• 2003

In this paper we propose Using Higher Order Neuron on the Supervised Learning Machine of the Kohonen Feature Map. The architecture of proposed model adopts the higher order neuron in the input layer of Kohonen Feature Map as a Supervised Learning Machine. It is able to estimate boundary on input pattern space because or the higher order neuron. However, it suffers from a problem that the number of neuron weight increases because of the higher order neuron in the input layer. In this time, we solved this problem by placing the second order neuron among the higher order neuron. The feature of the higher order neuron can be mapped similar inputs on the Kohonen Feature Map. It also is the network with topological mapping. We have simulated the proposed model in respect of the recognition rate by XOR problem, discrimination of 20 alphabet patterns, Mirror Symmetry problem, and numerical letters Pattern Problem.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2022년도 제66차 하계학술대회논문집 30권2호
• /
• pp.67-70
• /
• 2022

NLP 분야에서 적극 활용되기 시작한 Transformer 모델을 Vision 분야에서 적용하기 시작하면서 object detection과 segmentation 등 각종 분야에서 기존 CNN 기반 모델의 정체된 성능을 극복하며 향상되고 있다. 또한, label 데이터 없이 이미지들로만 자기지도학습을 한 ViT(Vision Transformer) 모델을 통해 이미지에 포함된 여러 중요한 객체의 영역을 검출하는 saliency map을 추출할 수 있게 되었으며, 이로 인해 ViT의 자기지도학습을 통한 object detection과 semantic segmentation 연구가 활발히 진행되고 있다. 본 논문에서는 ViT 모델 뒤에 classifier를 붙인 모델에 일반 학습한 모델과 자기지도학습의 pretrained weight을 사용해서 전이학습한 모델의 시각화를 통해 각 saliency map들을 비교 분석하였다. 이를 통해, 클래스 분류 학습 기반 전이학습이 transformer의 saliency map에 미치는 영향을 확인할 수 있었다.

• 제어로봇시스템학회논문지
• /
• 제15권4호
• /
• pp.451-455
• /
• 2009

This paper provides a comparison of global path planning method in single string by using pulled and pushed SOFM (Self-Organizing Feature Map) which is a method among a number of neural network. The self-organizing feature map uses a randomized small valued initial-weight-vectors, selects the neuron whose weight vector best matches input as the winning neuron, and trains the weight vectors such that neurons within the activity bubble are move toward the input vector. On the other hand, the modified SOFM method in this research uses a predetermined initial weight vectors of the one dimensional string, gives the systematic input vector whose position best matches obstacles, and trains the weight vectors such that neurons within the activity bubble are move toward or reverse the input vector, by rising a pulled- or a pushed-SOFM. According to simulation results one can conclude that the modified neural networks in single string are useful tool for the global path planning problem of a mobile robot. In comparison of the number of iteration for converging to the solution the pushed-SOFM is more useful than the pulled-SOFM in global path planning for mobile robot.

• 한국식품저장유통학회지
• /
• 제6권3호
• /
• pp.264-269
• /
• 1999

생강의 소포장화 유통을 위한 기초연구의 하나로서, 적정 필름의 선정과 결로현상을 방지하기 위하여 두께 0.04, 0.06, 0.08mm의 LDPE 필름에 생강을 저장함과 동시에 0.06mm의 LDPE 필름에 흡습제를 sachet형태와 씨트 형태로 첨가한 다음 생강을 150일 동안 저장하면서 생강의 MA 저장 중 필름 두께 및 흡습제의 첨가방법에 따른 저장 효과를 조사하였다. 흡습제를 첨가하지 않은 MAP방법에서는 필름두께별로 저장 150일 동안 중량 감소율 및 경도의 변화는 거의 차이가 없었으나, 발아율, 부패율 및 환원당 함량에서는 필름두께가 얇은 것일수록 변화속도가 낮게 나타나고 있었다. 또한 0.06mm의 LDPE 필름에 흡습제를 첨가한 MAP방법에서는 흡습제를 첨가하지 않은 동일 두께의 MAP방법보다 중량감소율은 약 3-4배 정도 높게 나타났으나 발아율, 부패율 및 환원당의 증가현상을 억제하는 효과가 3.3, 1.4, 2.3배 정도 각각 억제할 수 있는 효과가 있었으며, 특히 발아율 기준으로 한 저장수명은 약 3배 정도 연장할 수 있는 효과가 있었다.

• 산업기술연구
• /
• 제25권B호
• /
• pp.47-54
• /
• 2005

The goal of this study is to generate a landslide potential map using GIS(Geographic Information System) based method. A simple and efficient algorithm is proposed to generate a landslide potentialities map from DEM(Digital Elevation Model) and existing maps. The categories of controlling factors for landslides, aspect of slope, soil, vegetation are defined. The weight value for landslide potentialities is calculated from AHP(Analytic Hierarchy Process) method. Slope and Slope-direction is extracted from DEM, and soil information is extracted from digital soil map. Also, vegetation information is extracted from digital vegetation map. Finally, as overlaying, landslide potentialities map is made out, and it is compared with landslide place.

• 전자공학회논문지
• /
• 제50권9호
• /
• pp.127-135
• /
• 2013

저조도 환경에서 카메라로 영상을 획득하기 위해 일반적으로 가시광 플래시를 사용하거나 장노출 기법을 사용하게 된다. 그러나 가시광 플래시를 사용할 때 플래시 광에 의한 색 왜곡이나 적목 현상, 눈부심에 의한 거부감을 발생시킨다. 또한 장노출을 사용하게 되면 물체의 움직임에 의한 흔들림 현상이 발생하게 된다. 따라서 최근에는 이러한 단점을 극복하고, 저조도 환경에서 고화질의 영상을 획득하기 위하여 멀티 스팩트럴 플래시(Multi-spectral flash image)를 이용하여 영상을 획득하는 방법이 소개되었다. 이 방법은 가시광과 UV/IR스펙트럼의 다섯 채널을 이용하여 가시광영상의 색 정보와 UV/IR 스팩트럼 영상의 세부정보를 최적화하여 영상을 획득하는 방법이다. 하지만, 픽셀 기반의 최적화 과정에 있어 색 왜곡과 다른 잡음을 발생시키게 된다. 따라서 본 논문에서는 이러한 색 왜곡과 잡음을 개선하기 위해 영역 기반의 가중치 맵을 최적화 방법에 적용하여 색 왜곡을 개선하는 알고리즘을 제안한다. 먼저, 영상에 대하여 Canny 에지 검출 방법을 사용하여 영상의 윤곽을 검출하였다. 이를 가중치 맵으로 최적화방법에 적용함으로, 세부 영역에 대하여 UV/IR 플래시 영상의 정보에 가중치를 부여하고, 평탄한 영역에 대하여 가시광 영상의 색 정보를 가중치를 부여하여 색 왜곡을 개선하였다. 제안한 방법을 평가하기 위하여 실험을 통하여 제안한 방법과 이전방법을 비교하였고, 객관적 평가와 주관적 평가 모두 제안한 방법이 우수한 성능을 나타내었다.