• Korean Journal of Remote Sensing
• /
• v.38 no.5_3
• /
• pp.939-951
• /
• 2022

The most important factor in designing autonomous driving systems is to recognize the exact location of the vehicle within the surrounding environment. To date, various sensors and navigation systems have been used for autonomous driving systems; however, all have limitations. Therefore, the need for high-definition (HD) maps that provide high-precision infrastructure information for safe and convenient autonomous driving is increasing. HD maps are drawn using three-dimensional point cloud data acquired through a mobile mapping system (MMS). However, this process requires manual work due to the large numbers of points and drawing layers, increasing the cost and effort associated with HD mapping. The objective of this study was to improve the efficiency of HD mapping by segmenting semantic information in an MMS point cloud into six classes: roads, curbs, sidewalks, medians, lanes, and other elements. Segmentation was performed using various machine learning techniques including random forest (RF), support vector machine (SVM), k-nearest neighbor (KNN), and gradient-boosting machine (GBM), and 11 variables including geometry, color, intensity, and other road design features. MMS point cloud data for a 130-m section of a five-lane road near Minam Station in Busan, were used to evaluate the segmentation models; the average F1 scores of the models were 95.43% for RF, 92.1% for SVM, 91.05% for GBM, and 82.63% for KNN. The RF model showed the best segmentation performance, with F1 scores of 99.3%, 95.5%, 94.5%, 93.5%, and 90.1% for roads, sidewalks, curbs, medians, and lanes, respectively. The variable importance results of the RF model showed high mean decrease accuracy and mean decrease gini for XY dist. and Z dist. variables related to road design, respectively. Thus, variables related to road design contributed significantly to the segmentation of semantic information. The results of this study demonstrate the applicability of segmentation of MMS point cloud data based on machine learning, and will help to reduce the cost and effort associated with HD mapping.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.22 no.4
• /
• pp.287-298
• /
• 2020

It is predicted that future climate warming will occur, and the subtropical climate zone currently confined to the south coast of Korea will gradually rise north. The shift of climate zone implies a change in area for cultivating crops. This study aimed to evaluate the current and future status of climate zones based on the high-resolution climate data of South Korea to prepare adaptation measures for cultivating crops under changing agricultural climate conditions. First, the climatic maps of South and North Korea were produced by using the high-resolution monthly maximum and minimum daily temperature and monthly cumulative precipitation produced during the past 30 years (1981-2010) covering South and North Korea. Then the climate zones of the Korean Peninsula were classified based on the Köppen climate classification. Second, the changes in climate zones were predicted by using the corrected monthly climate data of the Korean Peninsula (grid resolution 30-270m) based on the RCP8.5 scenario of the Korea Meteorological Administration. Köppen climate classification was applied based on the RCP8.5 scenario, the temperature and precipitation of the Korean Peninsula would continue to increase and the climate would become simpler. It was predicted that the temperate climate, appearing in the southern region of Korea, would be gradually expanded and the most of the Korean Peninsula, excluding some areas of Hamgkyeong and Pyeongan provinces in North Korea, would be classified as a temperate climate zone between 2071 and 2100. The subarctic climate would retreat to the north and the Korean Peninsula would become warmer and wetter in general.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.1163-1175
• /
• 2021

In autonomous driving, road markings are an essential element for object tracking, path planning and they are able to provide important information for localization. This paper presents an approach to update and measure road surface markers with HD maps as well as matching using inverse perspective mapping. The IPM removes perspective effects from the vehicle's front camera image and remaps them to the 2D domain to create a bird-view region to fit with HD map regions. In addition, letters and arrows such as stop lines, crosswalks, dotted lines, and straight lines are recognized and compared to objects on the HD map to determine whether they are updated. The localization of a newly installed object can be obtained by referring to the measurement value of the surrounding object on the HD map. Therefore, we are able to obtain high accuracy update results with very low computational costs and low-cost cameras and GNSS/INS sensors alone.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.5
• /
• pp.186-201
• /
• 2021

As autonomous driving technology advances, the accuracy of the vehicle position is important for recognizing the environments around driving. Map-matching localization techniques based on high definition (HD) maps have been studied to improve localization accuracy. Because conventional map-matching techniques estimate the vehicle position based on an HD map reference dataset representing the center of the lane, the estimated position does not reflect the deviation of the lateral distance within the lane. Therefore, this paper proposes a localization system based on the reference lateral position dataset extracted using image processing and HD maps. Image processing extracts the driving lane number using inverse perspective mapping, multi-lane detection, and yellow central lane detection. The lane departure method estimates the lateral distance within the lane. To collect the lateral position reference dataset, this approach involves two processes: (i) the link and lane node is extracted based on the lane number obtained from image processing and position from GNSS/INS, and (ii) the lateral position is matched with the extracted link and lane node. Finally, the vehicle position is estimated by matching the GNSS/INS local trajectory and the reference lateral position dataset. The performance of the proposed method was evaluated by experiments carried out on a highway environment. It was confirmed that the proposed method improves accuracy by about 1.0m compared to GNSS / INS, and improves accuracy by about 0.04m~0.21m (7~30%) for each section when compared with the existing lane-level map matching method.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.13 no.1
• /
• pp.20-27
• /
• 2011

High-definition, geo-referenced digital climate maps can be produced by applying watershed-specific modules to adjust synoptic observations for local effects including cold air drainage. Since there is no information available on North Korean watersheds, existing geospatial technology for digital climate mapping cannot be transferred to North Korea. We applied a watershed extraction algorithm based on ArcHydro to the North Korean portion of ASTER GDEM and utilized geographical information on major rivers and mountains to adjust the products. Proposed hydrologic zoning system for North Korean watersheds consists of 21 river basins, 93 stream basins and 885 catchments. Combined with the existing 840 South Korean hydrologic units, we now have a complete set of 1,725 catchments which may serve a framework for digital climate modeling across whole land area of the Korean Peninsula.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.620-624
• /
• 2009

In this paper, we present a new camera system combining a high-quality 3-D scanner and hybrid camera system to generate a multiview video-plus-depth. In order to get the 3-D video using the hybrid camera system and 3-D scanner, we first obtain depth information for background region from the 3-D scanner. Then, we get the depth map for foreground area from the hybrid camera system. Initial depths of each view image are estimated by performing 3-D warping with the depth information. Thereafter, multiview depth estimation using the initial depths is carried out to get each view initial disparity map. We correct the initial disparity map using a belief propagation algorithm so that we can generate the high-quality multiview disparity map. Finally, we refine depths of the foreground boundary using extracted edge information. Experimental results show that the proposed depth maps generation method produces a 3-D video with more accurate multiview depths and supports more natural 3-D views than the previous works.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.14 no.4
• /
• pp.161-169
• /
• 2012

The RCP 8.5 scenario based temperature outlook (12.5 km resolution) was combined with high-definition gridded temperature maps (30 m grid spacing) across the Korean Peninsula in order to reclassify the cold hardiness zone for winter barley, a promising grain crop in the future under warmer winter conditions. Reference maps for the January minimum and mean temperature were prepared by applying the watershed-specific geospatial climate prediction schemes to the synoptic observations from 1981 to 2010 across North and South Korea. Decadal changes in the January minimum and mean temperatures projected by a regional version of RCP8.5 climate change scenario were prepared for the 2011-2100 period at 12.5 km grid spacing and were subsequently added to the reference maps, producing the 30 m resolution temperature surfaces for 9 decades from 2011 to 2100. A criterion for threshold temperature to grow winter barley safely in Korea was applied to the future temperature surfaces and the resulting maps were used to predict the production potential of 3 cultivar groups for the 9 future decades under the projected temperature conditions. By 2020s, hulled barley cultivars could be grown safely at the southern part of North Korea as well as the mountainous Gangwon province. Furthermore, most of South Korean rice paddies will be safe for growing naked barley after harvesting rice. Also, dual cropping systems such as 'winter-barley after rice' could be possible at most of the North Korean rice paddies by 2040s. Additional grain production in North Korea could increase up to 4 million tons per year if dual cropping systems can be fully operated, i.e., winter barley after rice at all lowlands and winter barley after maize or potato at all uplands.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.13 no.1
• /
• pp.35-40
• /
• 2011

While high-definition precipitation maps with a 270 m spatial resolution are available for South Korea, there is little information on geospatial availability of precipitation water for the famine - plagued North Korea. The restricted data access and sparse observations prohibit application of the widely used PRISM (Parameter-elevation Regressions on Independent Slopes Model) to North Korea for fine-resolution mapping of precipitation. A hybrid method which complements the PRISM grid with a sub-grid scale elevation function is suggested to estimate precipitation for remote areas with little data such as North Korea. The fine scale elevation - precipitation regressions for four sloping aspects were derived from 546 observation points in South Korea. A 'virtual' elevation surface at a 270 m grid spacing was generated by inverse distance weighed averaging of the station elevations of 78 KMA (Korea Meteorological Administration) synoptic stations. A 'real' elevation surface made up from both 78 synoptic and 468 automated weather stations (AWS) was also generated and subtracted from the virtual surface to get elevation difference at each point. The same procedure was done for monthly precipitation to get the precipitation difference at each point. A regression analysis was applied to derive the aspect - specific coefficient of precipitation change with a unit increase in elevation. The elevation difference between 'virtual' and 'real' surface was calculated for each 270m grid points across North Korea and the regression coefficients were applied to obtain the precipitation corrections for the PRISM grid. The correction terms are now added to the PRISM generated low resolution (~2.4 km) precipitation map to produce the 270 m high resolution map compatible with those available for South Korea. According to the final product, the spatial average precipitation for entire territory of North Korea is 1,196 mm for a climatological normal year (1971-2000) with standard deviation of 298 mm.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.13 no.1
• /
• pp.28-34
• /
• 2011

This study was carried out to estimate monthly mean of daily maximum and minimum temperature across North Korea at a 30 m grid spacing for a climatological normal year (1971-2000) and the 4 decadal averages (1971-1980, 1981-1990, 1991-2000, and 2001-2010). A geospatial climate interpolation method, which has been successfully used to produce the so-called 'High-Definition Digital Climate Maps' (HD-DCM), was used in conjunction with the 27 North Korean and 17 South Korean synoptic data. Correction modules including local effects of cold air drainage, thermal belt, ocean, solar irradiance and urban heat island were applied to adjust the synoptic temperature data in addition to the lapse rate correction. According to the final temperature estimates for a normal year, North Korean winter is expected colder than South Korean winter by $7^{\circ}C$ in average, while the spatial mean summer temperature is lower by $3^{\circ}C$ than that for South Korea. Warming trend in North Korea for the recent 40 years (1971-2010) was most remarkable in spring and fall, showing a 7.4% increase in the land area with 15 or higher daily maximum temperature for April.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.3
• /
• pp.155-166
• /
• 2019

Road infrastructure has increased explosively due to economic development after industrialization and at present road infrastructure is being changed and increased by construction of new roads and maintenance and expansion of existing roads. Such road infrastructure should support safe driving. Road management plays an important role in safe driving. The purpose of this dissertation is to verify predictability of dangerous sections by analyzing road geometrical structure such as surface irregularity and superelevation for some sections in Central Inland Expressway by MMS and present ways of managing roads using MMS. Having analyzed surface irregularity of roads by using MMS, it was found that over 50 percent of all eight sections, targets of this study need betterments and for superelevation, over 50 percent of two sections goes against superelevation standard. Targets of this study are sections that accidents occurred frequently based on history of past accidents and predictability of dangerous sections can be verified through analysis of road geometrical structure using MMS. Using MMS data created by construction of high definition maps which are being undergone for all roads and methods proposed by this study will help investigate dangerous sections efficiently according to road environment. A result of MMS can be used for maintenance of road furniture.