• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.5
• /
• pp.407-414
• /
• 2020

SVM (Support vector machine) analysis was performed after applying a deep learning technique based on an Inception-based model (GoogLeNet). The accuracy of automatic image classification was analyzed using an SVM with multiple virtual grid sizes. Six classes were selected from a standard land cover map. Cars were added as a separate item to increase the classification accuracy of roads. The virtual grid size was 2-5 m for natural areas, 5-10 m for traffic areas, and 10-15 m for building areas, based on the size of items and the resolution of input images. The results demonstrate that automatic classification accuracy can be increased by adopting an integrated approach that utilizes weighted virtual grid sizes for different classes.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.15 no.4
• /
• pp.93-100
• /
• 2005

In this paper, we describe biometric watermarking techniques for secure user verification on the remote, multimodal biometric system employing both fingerprint and face information, and compare their effects on verification accuracy quantitatively. To hide biometric data with watermarking techniques, we first consider possible two scenarios. In the scenario 1, we use a fingerprint image as a cover work and hide facial features into it. On the contrary, we hide fingerprint features into a facial image in the Scenario 2. Based on the experimental results, we confirm that the Scenario 2 is superior to the Scenario 1 in terms of the verification accuracy of the watermarking image.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.42 no.3
• /
• pp.28-36
• /
• 2010

Today We have three full series of the annals of Joseon Dynasty. Taebaeksan volume which is one of them shows comparatively good maintenance but Taebaeksan volume also decomposed in patches. The size of cover and inner paper were increased in late Joseon Dynasty, but the weight, thickness and basis weight bears no relation to each other. The oxidation index of paper in annals of pre-periods of King Sunjo shows very high value. After the annals of King Sunjo, the oxidation index of paper is very similar to today's Hanji. L.a.b values of inner paper in the Taebaeksan volume are similar generally. The average length of fiber using image analysis soft ware is 8.08~9.92 mm, that like immediate value of measurement. The air fraction of fiber's range is 1.20~2.12%. It lower than today's Hanji. After considering the factors of the length of fiber, air fraction and FE-SEM, we reached that no difference between raw material of inner paper in the Taebaeksan volume.

• Journal of Korean Society of Rural Planning
• /
• v.12 no.1 s.30
• /
• pp.75-83
• /
• 2006

본 연구에서는 최대우도법과 인공신경망 모형에 의해 카테고리 분류를 수행하고 각각의 분류 성능을 비교 평가하였다. 인공신경망 모형은 오류역전파 알고리즘을 이용한 것으로서 학습을 통한 은닉층의 최적노드수를 결정하여 카테고리 분류를 수행하도록 하였다. 인공신경망 최적 모형은 입력층의 노드수가 7개, 은닉층의 최적노드수가 18개, 그리고 출력층의 노드수가 5개인 것으로 구성하였다. 위성영상은 1996년에 촬영된 Landsat TM-5 영상을 사용하였고, 최대우도법과 인공신경망 모형에 의한 카테고리 분류를 위하여 각각의 카테고리에 대한 분광특성을 대표하는 지역을 절취하였다. 분류 정확도는 인공신경망 모형에 의한 방법이 90%, 최대우도법이 83%로서, 인공신경망 모형의 분류 성능이 뛰어난 것으로 나타났다. 카테고리 분류 항목인 토지 피복 상태에 따른 분류는 두 가지 방법에서 밭과 주거지의 분류오차가 큰 것으로 나타났다. 특히, 최대우도법에 의한 밭에서의 태만오차는 62.6%로서 매우 큰 값을 보였다. 이는 밭이나 주거지의 특성이 위성영상 촬영시기에 따라 나지의 형태로 분류되거나 산림, 또는 논으로도 분류되는 경향이 있기 때문인 것으로 보인다. 차후에 카테고리 분류를 위한 각각의 클래스의 보조적인 정보를 추가한다면, 카테고리 분류 향상이 이루어질 것으로 기대된다.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.598-601
• /
• 2006

Satellite imagery may contain large regions covered with atmospheric aerosol. A high-resolution satellite imagery affected by non-homogenous aerosol cover should be processed for land cover study and perform the radiometric calibration that will allow its future application for Korea Multi-Purpose Satellite (KOMPSAT) data. In this study, aerosol signal was separated from high resolution satellite data based on the reflectance separation method. Since aerosol removal has a good sensitivity over bright surface such as man-made targets, aerosol optical thickness (AOT) retrieval algorithm could be used. AOT retrieval using Look-up table (LUT) approach for utilizing the transformed image to radiometrically compensate visible band imagery is processed and tested in the correction of satellite scenery. Moderate Resolution Imaging Spectroradiometer (MODIS), EO-1/HYPERION data have been used for aerosol correction and AOT retrieval with different spatial resolution. Results show that an application of the aerosol detection for HYPERION data yields successive aerosol separation from imagery and AOT maps are consistent with MODIS AOT map.

• Weed & Turfgrass Science
• /
• v.6 no.1
• /
• pp.55-60
• /
• 2017

Shade affects turf quality by reducing light for photosynthesis. The shade tolerance of the tropical grasses, Digitaria longiflora and Axonopus compressus were evaluated against Stenotaphrum secundatum (St. Augustinegrass). The grasses were established under shade structures that provide 0%, 50%, 75% or 90% shade level for 30 days. A suite of leaf traits, recorded from similar leaf developmental stage, displayed distinct responses to shade conditions. Leaf length, relative to control, increased in all three species as shade level increased. The mean leaf extension rate was lowest in St. Augustinegrass (80.42%) followed by A. compressus (84.62%) and D. longiflora (90.78%). The higher leaf extension rate in D. longiflora implied its poor shade tolerance. Specific leaf area (SLA) increased in all species with highest mean SLA increase in D. longiflora ($348.55cm^2mg^{-1}$)followed by A. compressus ($286.88cm^2mg^{-1}$) and St. Augustinegrass ($276.28cm^2mg^{-1}$). The highest SLA increase in D. longiflora suggested its lowest performance under shade. The percent green cover, as estimated by digital image analysis, was lowest in D. longiflora (53%) under 90% shade level compared to both species. The relative shade tolerance of the three turfgrasses could be ranked as St. Augustinegrass > A. compressus > D. longiflora.

• Proceedings of the KSRS Conference
• /
• 1998.09a
• /
• pp.128-133
• /
• 1998

To assess the flood damages and to provide necessary information for preventing future catastrophe, it is necessary to appraise the inundated area with more accurate and rapid manner. This study attempts to evaluate the potential of satellite synthetic aperture radar (SAR) data for mapping of flood inundated area in southern part of Korea. JERS L-band SAR data obtained during the summer of 1997 were used to delineate the inundated areas. In addition, Landsat TM data were also used for analyzing the land cover condition before the flooding. Once the two data sets were co-registered, each data was separately classified. The water surface areas extracted from the SAR data and the land cover map generated using the TM data were overlaid to determine the flood inundated areas. Although manual interpretation of water surfaces from the SAR image seems rather simple, the computer classification of water body requires clear understanding of radar backscattering behavior on the earth's surfaces. It was found that some surface features, such as rice fields, runaway, and tidal flat, have very similar radar backscatter to water surface. Even though satellite SAR data have a great advantage over optical remote sensor data for obtaining imagery on time and would provide valuable information to analyze flood, it should be cautious to separate the exact areas of flood inundation from the similar features.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.17 no.4
• /
• pp.411-420
• /
• 1999

Coastal development and reclamation work make environment of shore destroy, such as rapid change of land use and destruction of wet-land and ocean ecosystem. Therefore new technique to detect change have been needed. This study designed new change detection method and applied to study area. The change detection image and quantitative change area by each classes are calculated. Also, this study can use the basic idea-determination data for coastal development and city plan as the sense of sight by changed images that changed from any land-cover to any land-cover between two dates.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.891-893
• /
• 2003

This study uses the images from the Corona ‘spy’ satellite, which have been declassified in November 2002 and available on Internet order for a very low cost. The image used dates from 1973 and has about 6m panchromatic characteristics. Along with a Landsat5TM of 1990 and Aster of 2001, a temporal range of about 30 years is achieved. A simple classification of the area was processed and crosschecked manually from the available recent toposheets of Thailand. Results show the development of human infrastructure in the Protected Island of Koh Chang in Thailand, from 1973 to date. Specific human locations are identified linked either to tourism development, or to villages of fishermen. Scope for using Corona in land cover changes on a longer time period than usual satellite images is possible. Some classification issues coming from the sensor have to be taken into account. Accuracy assessment is also an issue because of the age of the sensor.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.5
• /
• pp.2348-2365
• /
• 2018

QR codes as public patent are widely used to acquire the information in various fields. However, it faces security problem when delivering the privacy message by QR code. To overcome this weakness, we propose a secret hiding scheme by improving exploiting modification direction to protect the private message in QR code. The secret messages will be converted into octal digit stream and concealed to the cover QR code by overwriting the cover QR code public message bits. And the private messages can be faithfully decoded using the extraction function. In our secret hiding scheme, the QR code public message still can be fully decoded publicly from the marked QR codes via any standard QR Code reader, which helps to reduce attackers' curiosity. Experiments show that the proposed scheme is feasible, with high secret payload, high security protection level, and resistant to common image post-processing attacks.