• Journal of Korean Society of Forest Science
• /
• v.110 no.2
• /
• pp.155-164
• /
• 2021

Many studies have been conducted on developing automatic plant identification algorithms using machine learning to various plant features, such as leaves and flowers. Unlike other plant characteristics, barks show only little change regardless of the season and are maintained for a long period. Nevertheless, barks show a complex shape with a large variation depending on the environment, and there are insufficient materials that can be utilized to train algorithms. Here, in addition to the previously published bark image dataset, BarkNet v.1.0, images of barks were collected, and a dataset consisting of 53 tree species that can be easily observed in Korea was presented. A convolutional neural network (CNN) was trained and tested on the dataset, and the factors that interfere with the model's performance were identified. For CNN architecture, VGG-16 and 19 were utilized. As a result, VGG-16 achieved 90.41% and VGG-19 achieved 92.62% accuracy. When tested on new tree images that do not exist in the original dataset but belong to the same genus or family, it was confirmed that more than 80% of cases were successfully identified as the same genus or family. Meanwhile, it was found that the model tended to misclassify when there were distracting features in the image, including leaves, mosses, and knots. In these cases, we propose that random cropping and classification by majority votes are valid for improving possible errors in training and inferences.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.6
• /
• pp.409-418
• /
• 2021

As the amount of construction for aerial photography increases, the need for automation of quality inspection is emerging. In this study, an experiment was performed to classify or detect clouds in aerial photos using deep learning techniques. Also, classification and detection were performed by including satellite images in the learning data. As algorithms used in the experiment, GoogLeNet, VGG16, Faster R-CNN and YOLOv3 were applied and the results were compared. In addition, considering the practical limitations of securing erroneous images including clouds in aerial images, we also analyzed whether additional learning of satellite images affects classification and detection accuracy in comparison a training dataset that only contains aerial images. As results, the GoogLeNet and YOLOv3 algorithms showed relatively superior accuracy in cloud classification and detection of aerial images, respectively. GoogLeNet showed producer's accuracy of 83.8% for cloud and YOLOv3 showed producer's accuracy of 84.0% for cloud. And, the addition of satellite image learning data showed that it can be applied as an alternative when there is a lack of aerial image data.

• Korean Journal of Heritage: History & Science
• /
• v.51 no.2
• /
• pp.22-37
• /
• 2018

Unmanned aerial vehicles (UAV) have attracted considerable attention both at home and abroad. The UAV is equipped with a camera that shoots images, which is advantageous for access to areas where archaeological investigations are not possible. Moreover, it is possible to acquire three-dimensional spatial image information by modeling the terrain through aerial photographing, and it is possible to specify the interpretation of the terrain of the survey area. In addition, if we understand the change of the terrain through comparison with past aerial photographs, it will be very helpful to grasp the existence of the ruins. The terrain modeling for searching these remains can be divided into two parts. First, we acquire the aerial photographs of the current terrain using the drone. Then, using image registration and post-processing, we complete the image-joining and terrain-modeling using past aerial photographs. The completed modeled terrain can be used to derive several analytical results. In the present terrain modeling, terrain analysis such as DSM, DTM, and altitude analysis can be performed to roughly grasp the characteristics of the change in the form, quality, and micro-topography. Past terrain modeling of aerial photographs allows us to understand the shape of landforms and micro-topography in wetlands. When verified with actual findings and overlapping data on the modelling of each terrain, it is believed that changes in hill shapes and buried Microform can be identified as helpful when used in low-flying applications. Thus, modeling data using aerial photographs is useful for identifying the reasons for the inability to carry out archaeological surveys, the existence of terrain and ruins in a wide area, and to discuss the preservation process of the ruins. Furthermore, it is possible to provide various themes, such as cadastral maps and land use maps, through comparison of past and present topographical data. However, it is certain that it will function as a new investigation methodology for the exploration of ruins in order to discover archaeological cultural properties.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2003.04a
• /
• pp.383-386
• /
• 2003

이 연구는 주변국의 국가지도제작기관을 분석하고, 측량 및 지도제작을 위한 수치제작 체계 즉, 수치자료기반을 제작하기 위한 체계를 분석하고자 하는 것

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.5 no.2
• /
• pp.109-120
• /
• 1985

This thesis is a study on multiple close range photogrammetry, and the purpose of this study is to develop the most accurate adjustment method of three dimensional object coordinates. This was achieved by comparing the standard errors of actual data to the computed values from 2 photos and multiple photos. The conventional methods for multiple photos have been analyzed by using geometric model formation. But in this study, the equation of collinearity condition which has been applied to aerial photogrammetry was derived to be a basic principle of close range photogrammetry, and the algorithm for analyzing multiple photos was developed using simultaneous bundle adjustment. The method used in this study, showed more homogeneous accuracy in coordinate and more consistent variance of error than those of conventional methods. It was found that the cases using 3, 4, and 5 photos were more accurate than using 2 photos; the accuracies were improved to 15%, 35%, and 50%, for each case. Thus this study is expected to be useful in measuring the geometry of historic monuments and other structures requiring high accuracy. Also the combined case of multiple photos is considered to be effective for the precise analysis of the objects which are difficult to measure for obstacles.

• Proceedings of the KSRS Conference
• /
• 2007.03a
• /
• pp.92-96
• /
• 2007

전라남도 국민참여형 위성영상 블로그 시스템은 2006년도 행정자치부 자치단체정보화지원사업의 일환으로 구축된 것으로서， 사용자가 고해상도 위성영상 위에 직접 자신의 글，사진， 동영상을 올리고，이를 다른 사용자나 시스템과 공유할 수 있도록 구축된 시스템이다. 본 시스템이 궁극적으로 구현하고자 했던 과제는 다음과 같다. (1)지리정보 혹은 기타 정보의 고정성을 탈피하고，(2)지리정보와 멀티미디어 자료의 효율적 통합을 통해 geoUCC를 구축하며，(3)사용자의 참여를 통한 지리정보 제공 방식의 쌍방향성을 구현하여，(4) 사용자를 일방적 자료 활용자에서 자료 생성자로 발전시키고 궁극적으로는 자치단체와 국민 간 새로운 방식의 소통 방식을 생성하는 것이다. 본 시스템 구축의 결과로서 본 시스템에서 일반 국민에게 서비스되는 내용은 다음과 같다. (1) 전라남도 전역의 1m급 고해상도 위성영상의 대국민 서비스，(2)전라남도 전역에 대한 지번 단위까지의 주소 검색 서비스，(3)웹 상에서의 실시간 3차원 GIS 서비스，(4)Web2.0의 철학을 반영한 RSS, Trackback 및 Tag 검색 기능，(5)사용자가 직접 고해상도 위성영상 위에 자신의 글， 사진，동영상，홈페이지 등을 올릴 수 있는 "위성영상 블로그 쓰기 기능"，(6)국내 거의 모든 포털 사이트의 블로그와 홈페이지를 지원하는 시스템 연계(퍼가기) 기능 등이다. 결론적으로 본 시스템의 사용자는 고해상도 위성영상 기반의 지도 위에 자신의 여행기，사진， 동영상을 올림으로써 자신만의 사용자제작콘텐츠(UCC)를 생성할 수 있을 뿐만 아니라， 이를 다른 사용자와 공유， 활용함으로써 실질적인 쌍방향，참여형 지리정보시스템을 이용하게 된 것이다.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.25 no.3
• /
• pp.191-198
• /
• 2007

According to the survey data during the Japanese Occupation Period, the length of South Korea's coastline is about 11,542km, including the coastlines of mainland and islands. To accurately revise/renew this coastline data through site survey, it will cost great money and time. Also, various development projects such as reclamation works on public waters, constructions of ports/harbors, etc. This paper used aerial photographs, satellite image data and GPS survey data with certain intervals to monitor the change in coastal areas of Songieong, Haeundae, Kwanganri, Songdo and Dadaepo. The local area subjected for this research was limited to areas near Busan. The specific contents of this research include. Launching qualitative/time series analyses on the change of coastal areas using aerial photographs, satellite image data and RTK-GPS surveys.

• Journal of the Korean Geophysical Society
• /
• v.9 no.4
• /
• pp.371-376
• /
• 2006

Digital high resolution cameras are widely available, and are increasingly use in digital close-range photogrammetry. And photogrammetry instruments are developing rapidly and the precision is improving continuously. The building of 3D terrains of high precision are possible and the calculation of the areas or the earthwork volumes have high precision due to the development of the technique of the spatial information system using computer. Using the digital camera which has capacity of keeping numerical value by itself and easy carrying, we analyze the positioning error according to various change of photographing condition. Also we try to find a effective method of acquiring basis data for 3D modeling of high-accuracy in pixel degree through digital close-range photogrammetry with bundle adjustment for digital elevation model generation.

• Journal of Korean Society for Geospatial Information Science
• /
• v.5 no.2 s.10
• /
• pp.207-220
• /
• 1997

This study aims to develope the production method of digital image map of high capable utiliy and terrain interpretability using aerial photo and Geospatial Information System. Theory and efficient practical method was studied to generate tile digital image map with low-cost personal computer system using the merging procedure of raster scanned aerial photo and vector topographic map. Determination theory of ground coordinates, digital image processing, production of digital elevation model was reviewed. And some chariteristics of digital image map, image collection method and significant concepts of digital image processing was studied. Also input and output way of image data to generate the digital image nap, production method of orthophoto map using aerial photo through digital differential rectification was studied. As the result, digital image map was produced and analyzed through the above mentioned procedures.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.9 no.2
• /
• pp.9-18
• /
• 1991

A diagram has been annexed to the spot inspection record of the traffic accident were drawn up with measuring-tape heretofor. However, conventional reduction method have been usually contained the various problems according to the social development. To solve these problems, it was performed that photogrammetric method are measured 3-D coordinate of needed points and precised plotting of traffic accident spot investigation. Also, they have produced some good effects on speedy spot investigation and settlement of traffic stagnation. Because semi-metric camera and MR2 system are utilize for low cost, simple operation and almost unnecessary for control point surveying, these fittings being at the close-range photogrammetry will be extremely powerful over versatile fields like traffic accident in police.