• Journal of KIISE:Databases
• /
• v.30 no.3
• /
• pp.320-335
• /
• 2003

Location-Based Services(LBS) give rise to location-based queries of which results depend on the locations of moving objects. One of important applications of LBS is to examine tracks of continuously moving objects. Moving objects databases need to provide 3-dimensional indexing for efficiently processing range queries on the movement of continuously changing positions. An extension of the 2-dimensional R-tree to include time dimension shows low space utilization and poor search performance, because of high overlap of index nodes and their dead space. To solve these problems, we propose a new R-tree based indexing technique, namely TR-tree. To increase storage utilization, we assign more entries to the past node by using the unbalanced splitting policy. If two nodes are highly overlapped, these nodes are forcibly merged. It is the forced merging policy that reduces the dead space and the overlap of nodes. Since big line segments can also affect the overlap of index nodes to be increased, big line segments should be clipped by the clipping policy when splitting overfull nodes. The TR-tree outperforms the 3DR-tree and TB-tree in all experiments. Particularly, the storage utilization of the TR-tree is higher than the R-tree and R*-tree.

• Journal of the Economic Geographical Society of Korea
• /
• v.4 no.2
• /
• pp.1-22
• /
• 2001

The object of this research is to explain types and factors of locations and regional linkages of manufacturing firms in rural areas using Eumsung County as a case of rural areas. Eumsung County is now vigorously industrializing owing to the government policy dispersing factories in the Capital region. The results of the survey conducted on industrial firms in Eumsung County through questionnaires reveal that individual sites are dominantly preferred by factory owners in comparison to the designated industrial complexes because the former is cheaper in land price. The main factors of industrial location in Eumsung County are cheap land cost, convenient highway accessibility, and the dispersion policy of the Capital region. In terms of the regional linkage, the proportion of the linkage with other regions is larger than that within the local region in case of purchasing raw materials, but the linkage with other regions is lower than that within the local regions in case of selling products. Also most white collar employees in Eumsung County were revealed to reside in and around the Capital region. In order to develop and maintain industrial firms in rural areas so as to avoid "dependency development," efforts for improving of the linkages between the local firms and expanding of social overhead capital are necessary.necessary.

• Journal of Korean Society for Geospatial Information Science
• /
• v.16 no.4
• /
• pp.3-10
• /
• 2008

Cadastral map is a public book that has been composed by continuous parcel having location, number, classification, boundary and an area based on Cadastral Law. A few years ago, cadastral map had been managed by form drawn on 2 dimension plane paper with 7 regular scales. Recently as computer systems are upgrading, cadastral map was able to have a chance to develope one step. Its type has been remade from raster to vector. In result, the cadastral map of vector type becomes to apply variously. Therefore, digital cadastral map has been ready a system to be use with multi-propose by KLIS(Korean Land Information System). In this research, it concretely want presentation of status using land more than original parcel on basic coordination cadastral map and KLIS(Korean Land Information System). The cadastral map is composed as parcel unit was applied by new presentation technique to "Model Research on One Parcel Presentation Technique for Land Status of Cadastral Map". The function of cadastral map on One Parcel Presentation Technique which is not only location relation of possession right and expression of states using land in 28 classifications demonstrated on the cadastral law but also used as foundation data of GIS construct business is developed by lines and classification of parcel to center around public sites of roads, rails, drains and rivers. especially, this research is composed of technique elevation and development of One Parcel Projection Technique of cadastral map in using object of roads among public sites.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.7
• /
• pp.67-76
• /
• 2011

Recently, researches for semantic annotation methods which represent and search objects included in video data, have been briskly activated since video starts to be popularized as types for interactive contents. Different location data occurs at each frame because coordinates of moving objects are changed with the course of time. Saving the location data for objects of every frame is too ineffective. Thus, it is needed to compress and represent effectively. This paper suggests two methods; the first, ontology modeling for moving objects to make users intuitively understandable for the information, the second, to reduce the amount of data for annotating moving objects by using cubic spline interpolation. To verify efficiency of the suggested method, we implemented the interactive video system and then compared with each video dataset based on sampling intervals. The result follows : when we got samples of coordinate less than every 15 frame, it showed that could save up to 80% amount of data storage; moreover, maximum of error deviation was under 31 pixels and the average was less than 4 pixels.

• KIISE Transactions on Computing Practices
• /
• v.22 no.4
• /
• pp.189-194
• /
• 2016

A trajectory is the motion path of a moving object. The advances in IT have made it possible to collect an immeasurable amount of various type of trajectory data from a moving object using location detection devices like GPS. The trajectories of moving objects are widely used in many different fields of research, including the geographic information system (GIS) field. In the GIS field, several attempts have been made to automatically generate digital maps of roads by using the vehicle trajectory data. To achieve this goal, the method to cluster the trajectories on the same road is needed. Usually, the $Fr{\acute{e}}chet$ distance measure is used to calculate the distance between a pair of trajectories. However, the $Fr{\acute{e}}chet$ distance measure requires prolonged calculation time for a large amount of trajectories. In this paper, we presented a fast heuristic algorithm to distinguish whether the trajectories are in close distance or not using the discrete $Fr{\acute{e}}chet$ distance measure. This algorithm trades the accuracy of the resulting distance with decreased calculation time. By experiments, we showed that the algorithm could distinguish between the trajectory within 10 meters and the distant trajectory with 95% accuracy and, at worst, 65% of calculation reduction, as compared with the discrete $Fr{\acute{e}}chet$ distance.

• The KIPS Transactions:PartB
• /
• v.12B no.1 s.97
• /
• pp.47-56
• /
• 2005

A moving object has a various features that its spatial location, shape, and size are changed as time goes. In addition, the moving object has both temporal feature and spatial feature. It is one of the highly interested feature information in video data. In this paper, we propose an efficient content-based multimedia information retrieval system, so tailed ECoMOT which enables user to retrieve video data by using a trajectory information of moving objects in video data. The ECoMOT includes several novel techniques to achieve content-based retrieval using moving objects' trajectories : (1) Muitiple trajectory modeling technique to model the multiple trajectories composed of several moving objects; (2) Multiple similar trajectory retrieval technique to retrieve more similar trajectories by measuring similarity between a given two trajectories composed of several moving objects; (3) Superimposed signature-based trajectory indexing technique to effectively search corresponding trajectories from a large trajectory databases; (4) convenient trajectory extraction, query generation, and retrieval interface based on graphic user interface

• Journal of The Geomorphological Association of Korea
• /
• v.18 no.1
• /
• pp.85-100
• /
• 2011

We were to investigate the geomorphological environment of optimum location in the object of Bronze Age's dwelling site (hereinafter called dwelling site) of total of 205 units confirmed in the 'Cheonan Baekseokdong Relic Group' distributed in hill of the uppermost stream part in the watershed of Jangjae stream in Cheonan, Chungnam Province. To do this, we classified the hill of the object of investigation where dwelling sites were distributed as 8 units of slope micro-landforms and again by combining them with the grade of gradient of 5 units subdivided them into ultra-micro-landforms of total 40 units. On the foundation of this, in the viewpoints of 'gradients of each of micro-landforms' analyzed the 'number of dwelling sites' and 'dwelling site distribution density (measure: number of dwelling sites/1000m²) of 'Cheonan Baekseokdong Relic Group'. As the result, the optimum landform environment where the dwelling sites were located were found to be largely 5 units of ultra-micro-landforms - ① flat land of crest flat, ② gentle slope land of crest flat, ③ flat land of crest slope, ④ gentle slope land of crest slope and ⑤ semi-gentle slope land of crest slope. This analyzed material will be used from now on as basic material which can predict the distribution of dwelling sites of Bronze Age men who dwelled in the watershed of Jangjae stream.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.5
• /
• pp.1075-1084
• /
• 2024

Efforts to expand the installation of devices that assist visually impaired individuals in their mobility are ongoing, but there are significantly fewer devices installed indoors compared to outdoors, causing considerable inconvenience for indoor navigation. Therefore, this paper aims to address these issues by applying the results of machine learning using YOLO(You Only Look Once) to a Raspberry Pi and by researching techniques to reduce errors through the trilateration method of UWB(Ultra-Wideband) sensors, applying it with a Kalman filter. The research results implemented an object recognition algorithm with a comprehensive accuracy of 91.7% using YOLO technology. Based on this object recognition, the direction (left, right, or front) was determined using the distance difference between two ultrasonic sensors set at an angle difference of 15 degrees. A distance of up to 1.5m was accepted through an infrared sensor to output a warning message according to the distance. The distance between the user's tag and the fixed three anchors was measured indoors through a UWB sensor, and the user's location was also measured indoors by linking the distance value with the three-side positioning technique.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.14 no.6
• /
• pp.57-65
• /
• 2018

Conventional RF signal-based indoor localization techniques such as BLE or Wi-Fi based fingerprinting method show considerable localization errors even in small-scale indoor environments due to unstable received signal strength(RSS) of RF signals. Therefore, it is difficult to apply the existing RF-based fingerprinting techniques to large-scale indoor environments such as airports and department stores. In this paper, instead of RF signal we use the geomagnetic sensor signal for indoor localization, whose signal strength is more stable than RF RSS. Although similar geomagnetic field values exist in indoor space, an object movement would experience a unique sequence of the geomagnetic field signals as the movement continues. We use a deep neural network model called the recurrent neural network (RNN), which is effective in recognizing time-varying sequences of sensor data, to track the user's location and movement path. To evaluate the performance of the proposed geomagnetic field based indoor positioning system (IPS), we constructed a magnetic field map for a campus testbed of about $94m{\times}26$ dimension and trained RNN using various potential movement paths and their location data extracted from the magnetic field map. By adjusting various hyperparameters, we could achieve an average localization error of 1.20 meters in the testbed.

• Proceedings of the National Institute of Ecology of the Republic of Korea
• /
• v.3 no.2
• /
• pp.67-72
• /
• 2022

Ecological research relies on the interpretation of large amounts of visual data obtained from extensive wildlife surveys, but such large-scale image interpretation is costly and time-consuming. Using an artificial intelligence (AI) machine learning model, especially convolution neural networks (CNN), it is possible to streamline these manual tasks on image information and to protect wildlife and record and predict behavior. Ecological research using deep-learning-based object recognition technology includes various research purposes such as identifying, detecting, and identifying species of wild animals, and identification of the location of poachers in real-time. These advances in the application of AI technology can enable efficient management of endangered wildlife, animal detection in various environments, and real-time analysis of image information collected by unmanned aerial vehicles. Furthermore, the need for school education and social use on biodiversity and environmental issues using AI is raised. School education and citizen science related to ecological activities using AI technology can enhance environmental awareness, and strengthen more knowledge and problem-solving skills in science and research processes. Under these prospects, in this paper, we compare the results of our early 2013 study, which automatically identified African cichlid fish species using photographic data of them, with the results of reanalysis by CNN deep learning method. By using PyTorch and PyTorch Lightning frameworks, we achieve an accuracy of 82.54% and an F1-score of 0.77 with minimal programming and data preprocessing effort. This is a significant improvement over the previous our machine learning methods, which required heavy feature engineering costs and had 78% accuracy.