• 한국측량학회지
• /
• 제26권4호
• /
• pp.333-340
• /
• 2008

This paper shows that the use of prior probabilities of the involved classes improve the accuracy of classification in case of poor separability between classes. Three cases of experiments are designed with two LiDAR datasets while considering three different classes (building, tree, and flat grass area). Moreover, random sampling method with human interpretation is used to achieve the approximate prior probabilities in this research. Based on the experimental results, Bayesian classification with the appropriate prior probability makes the improved classification results comparing with the case of non-prior probability when the ratio of prior probability of one class to that of the other is significantly different to 1.0.

• Asia pacific journal of information systems
• /
• 제32권1호
• /
• pp.32-50
• /
• 2022

Technology acceptance is one of the most popular research areas. Rapid developments in technology are making human life more comfortable. However, still most of the rural area has been deprived of benefits of technological advancement. Seventy percent population of India resides in rural area. Leveraging the improved penetration of the internet; mobile friendly population in rural India has been increasingly shopping online in the last few years. e-Governance is one of the important vehicles to provide efficient services to the citizens by Governments. One major obstacle is acceptance of e-Governance platforms by the citizens. Considering the increasing trend of using e-Commerce in rural area, this paper attempts to investigate moderating effect of online shopping experience on intention to use e-Governance portals. We surveyed 365 villagers across Maharashtra: one of the leading states in India. The result confirmed online shopping experience moderates the relationship between: 'perceived security & privacy' and 'attitude'; 'perceived security & privacy' and 'intention to use'; 'Perceived usefulness' and 'attitude'; and, 'attitude' and 'intention to use'. In this study definition of moderating variable 'experience' is unique and different than most of the popular studies. We defined experience as: 'prior use of any application of technology similar to the target application of technology'. Whereas prior studies considered experience as prior experience with target application of the technology.

• 방송공학회논문지
• /
• 제18권2호
• /
• pp.159-166
• /
• 2013

본 논문에서는 영상의 edge 정보를 이용하여 영상의 안개가 짙은 정도에 따라 영역을 분할 후 잘못 예상한 전달량을 보정하는 방법을 제안한다. 안개가 있는 날 얻은 영상은 가시성이 떨어질 뿐만 아니라 영상처리의 다양한 알고리즘을 사용하기 어렵기 때문에 안개가 짙은 정도인 전달량을 이용하여 안개를 제거한다. 안개제거를 위한 대표적인 방법중 하나인 Dark Channel Prior(DCP) 알고리즘은 날씨가 맑은 영상에서 일정 영역 안에서는 어두운 값이 존재하는 특성을 이용하여 안개의 전달량을 예상한다. 하지만 영상의 RGB 신호에서 전 채널 모두 높은 값을 갖고 있는 영역이 전달량을 찾는 영역보다 클 경우 전달량을 잘못 예상하게 된다. 따라서 제안하는 알고리즘을 사용하여 잘못된 전달량 보정을 통해 색이 왜곡되는 부분을 제거함으로서 기존의 알고리즘과 비교하여 영상 내의 색상이 자연스럽게 구현되고 안개가 제거된 결과를 얻었다.

• 한국멀티미디어학회논문지
• /
• 제20권2호
• /
• pp.101-107
• /
• 2017

LabVIEW coding for video dehazing was developed. The dark channel prior proposed by K. He was applied to remove fog based on a single image, and K. B. Gibson's median dark channel prior was applied, and implemented in LabVIEW. In other words, we improved the image processing speed by converting the existing fog removal algorithm, dark channel prior, to the LabVIEW system. As a result, we have developed a real-time fog removal system that can be commercialized. Although the existing algorithm has been utilized, since the performance has been verified real - time, it will be highly applicable in academic and industrial fields. In addition, fog removal is performed not only in the entire image but also in the selected area of the partial region. As an application example, we have developed a system that acquires clear video from the long distance by connecting a laptop equipped with LabVIEW SW that was developed in this paper to a 100~300 times zoom telescope.

• 방송공학회논문지
• /
• 제21권5호
• /
• pp.770-781
• /
• 2016

안개 영상은 색상 정보와 경계선 정보가 줄어들기 때문에 사물의 식별에 어려움이 발생한다. 안개 제거를 위한 대표적인 알고리즘은 Dark Channel Prior(DCP)를 이용한 방식으로 안개 영상의 색상 정보를 이용하여 안개의 전달량을 추정 한 후 안개를 제거한다. 그러나 색상 정보를 바탕의 안개량 추정을 할 경우, 영상 내에 흰색 사물이나 광원과 같이 높은 밝기의 사물이 있는 영역에 대해 안개량 및 전달량을 오추정하는 문제점이 발생한다. 본 논문에서는 안개 영상의 백색 영역을 중심으로 색 정규화를 적용 한 후, 안개량 오추정 영역을 주변 안개값으로 보정하여 안개값 오추정 문제를 개선하였다. 또한 개선된 안개량 추정 결과를 바탕으로 전달량 계산 후 전달량 정련을 통해 안개 영상의 화질을 개선하였다.

• 대한간호학회지
• /
• 제27권1호
• /
• pp.96-108
• /
• 1997

The purpose of this study was to determine the effect of endurance training prior to occurrence of muscle atrophy on the mass, myofibrillar protein content and fiber crossectional area of atrophied hindlimb muscles of rats. Adult female Wistar rats were trained prior to occurrence of muscle atrophy induced by hindlimb suspension. Training began on the 1st day for 10min /day at 15m /min on a 0% grade, training exercise increased daily in time and intensity so that by the 4th week rats were running 60min /day, at 34m /min on a i3.5% grade. Wet weight and relative weight of soleus, plantaris and gastrocnemius muscle decreased significantly after seven days of hindlimb suspension. Wet weight and relative weight of soleus tended to increase and that of plantaris and gastrocnemius tended to decrease in the exercise group as compared to the control group. Myofibrillar protein content of soleus and gastrocnemius tended to increase and that of plantaris tended to decrease in the endurance trained group as compared to the control group. Fiber crossectional area of Type I, II fiber in soleus and plantaris muscle tended to increase in the exercise group as compared to the control group. Wet weight and relative weight of soleus. plantaris and gastrocnemius decreased significantly, myofibrillar protein content of soleus, plantaris and gastrocnemius increased in hindlimb suspended rats following endurance training as compared to the control group. There was no change in fiber type percentage and crossectional area of type I and II fiber in soleus muscle and that of type I and IIfiber in plantaris muscle decreased in the hindlimb suspended rats following endurance training as compared to the control group. Wet weight and relative weight of soleus and plantaris tended to increase, that of gastrocnemius increased significantly, myofibrillar protein content of soleus and plantaris muscle increased significantly and that of gastrocnemius tended to increase in the hindlimb suspended rats following endurance training as compared to sedentary rats following endurance training. Crossectional area of type I fiber of soleus muscle tended to increase. that of type I fiber of plantaris muscle increased significantly and that of type II fiber tended to increase in hindlimb suspended rats following endurance training as compared to sedentary rats following endurance training. The results suggest that endurance training prior to occurrence of muscle atrophy can attenuate the decrease of mass, myofibrillar protein content and fiber crossectional area induced by hindlimb suspension.

• International journal of advanced smart convergence
• /
• 제8권4호
• /
• pp.68-74
• /
• 2019

Along with the deeper architecture in the deep learning approaches, the need for the data becomes very big. In the real problem, to get huge data in some disciplines is very costly. Therefore, learning on limited data in the recent years turns to be a very appealing area. Meta-learning offers a new perspective to learn a model with this limitation. A state-of-the-art model that is made using a meta-learning framework, Meta-SGD, is proposed with a key idea of learning a hyperparameter or a learning rate of the fast adaptation stage in the outer update. However, this learning rate usually is set to be very small. In consequence, the objective function of SGD will give a little improvement to our weight parameters. In other words, the prior is being a key value of getting a good adaptation. As a goal of meta-learning approaches, learning using a single gradient step in the inner update may lead to a bad performance. Especially if the prior that we use is far from the expected one, or it works in the opposite way that it is very effective to adapt the model. By this reason, we propose to add a weight term to decrease, or increase in some conditions, the effect of this prior. The experiment on few-shot learning shows that emphasizing or weakening the prior can give better performance than using its original value.

• 한국환경과학회지
• /
• 제24권8호
• /
• pp.1085-1099
• /
• 2015

The purposes of this study were to suggest the methodology to select prior areas in the environmental pollution survey for livestock excreta (EPSLE) as well as to elucidate the validity of the methodology. In this study, the prior areas in the EPSLE were determined by examining the number of compost facilities categorized according to the three levels of size including the basin, the sub-basin and the watershed, respectively, based on the data from "Annual Nation-wide Pollution Sources Survey (2012)". The results suggested that the list of prior basins were Nakdong, Geum, Youngsan and Han river basins in order. Also, it was examined that the prior sub-basins in the four river basins including Nakdong, Geum, Youngsan and Han rivers were Naesung Stream, Geumgang Gongju, Juam Dam and Namhan Downstream, respectively. The prior watersheds in the sub-basins of Naesung stream, Geumgang Gongju, Juam Dam and Namhan Downstream were Seocheon Downstream, Geum Stream, Gyeombaek Suwipyo and Yanghwa Stream, respectively. The validity of the methodology used in this study was elucidated by analyzing the correlation of the number of compost facilities with the concentrations of T-N and T-P observed in the end-points of sub-basins. The results of correlation analysis showed that the concentrations of T-N and T-P increased with the number of compost facilities. Specifically, there was the stronger correlation between the number of compost facilities and the concentrations of T-N than that for T-P. Consequently, it was proved that the methodology used in this work was valid and rational for the selection of prior areas in environmental pollution survey for EPSLE.

• 물과 미래
• /
• 제17권4호
• /
• pp.273-279
• /
• 1984

An attempt is made to present a method of prediction for typhoons apporaching the Korean-peninsula. The method is based upon the Bayesian theorem to improve the observed (prior) probabilities of typhoons approaching the Korean sea area incorporating conditional probability. A total of 248 typhoons is collected and analyzed to establish prior probability and conditional probability according to the defined procedure. The typhoons used are those which encompassed the western Pacific area to which the Korean-peninsula is subjected. The results of examplary computations suggest that the presented method is promising for predicting approaching typhoons.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 추계학술발표회 논문집
• /
• pp.545-550
• /
• 2004

Chilli reservoir is agricultural purpose reservoir made in 1945. But its bottom is covered with thick accumulated sewage layer, so its original function has failed. This study conducted cadastral state surveying to survey the state of this area prior to dredging the accumulated bottom sewage layer of Chill reservoir. Prior to cadastral state surveying we analyzed the digital map and aerial photographs of study area. In cadastral state surveying, traversed network was constructed on the road and banks near the reservoir, a surveyed map was made by surveying with Auto-reduction EDM Alidade. From the comparison results of the cadastral map and digital map, we found small chilli reservoir is expressed as rice field on the digital map, and it should be corrected. And when we make digital map and topographic map, the errors will be diminished if we use the cadastral map at the same time.