• Journal of Korea Multimedia Society
• /
• v.25 no.5
• /
• pp.695-702
• /
• 2022

In this paper, we propose the method of a table structure recognition in various table types for document processing automation. A table with items surrounded by ruled lines are analyzed by detecting horizontal and vertical lines for recognizing the table structure. In case of a table with items separated by spaces, the table structure are recognized by analyzing the arrangement of row items. After recognizing the table structure, the areas of the table items are input into OCR engine and the character recognition result output to a text file in a structured format such as CSV or JSON. In simulation results, the average accuracy of table item recognition is about 94%.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.687-689
• /
• 2021

정맥을 이용한 생체 인식 방법은 신체의 노화에 영향을 받지 않고 높은 사용 편의성과 변조의 위험이 적어 인증 수단으로 폭넓게 활용되고 있다. 그러나 가짜 정맥 영상을 통한 스푸핑 공격 위험이 존재한다. 이러한 문제를 해결하기 위해 실제 정맥 영상과 가짜 정맥 영상을 구분하는 기술이 필요하다. 본 연구에서는 실제 정맥 데이터의 마디와 뼈의 밝기 차이를 이용해 진짜 정맥 영상과 가짜 정맥 영상을 구분하는 기술을 연구했다.

• Annual Conference of KIPS
• /
• 2024.10a
• /
• pp.958-959
• /
• 2024

본 연구의 목적은 Oriented Bounding Box(OBB)를 활용하여 위성 영상에서의 소형 객체 탐지 및 추적 성능을 향상시키고자 한다. 기존 연구에서는 Horizontal Bounding Box(HBB)를 주로 사용하였으며 이는 객체에 대해 부정확한 마진을 많이 남기는 한계가 있었다. 이러한 방식은 객체 탐지에서 낮은 정확도 결과로 이어진다. 본 연구는 SAT-MTB dataset의 Oriented Bounding Box(OBB)를 활용하여 객체의 회전 및 기울기를 정확하게 반영하였다. 본 연구에서 OBB를 반영한 YOLOv8, YOLOv10, LSKNet을 비교한 결과 가장 좋은 성능을 보인 모델은 YOLOv10이었고, Airplane class에서 좋은 성능을 보였다. 이처럼 각 모델의 탐지 정확도를 분석함으로써, 소형 객체 탐지 및 추적 성능 향상에 기여하고자 한다.

• Journal of Sensor Science and Technology
• /
• v.26 no.3
• /
• pp.149-154
• /
• 2017

Detection of rain is one of the essential weather factors that are monitored by automatic weather stations in Korea. In this work, we studied the operation standards required for impedance-based rain detectors in terms of surface temperature and sensitivity, in an effort to establish a qualification procedure for rain detectors. The surface temperature of rain detectors was measured at varying air temperatures from $-30^{\circ}C$ to $20^{\circ}C$, considering the hypothetical presence and absence of rain/snow. In addition, the sensitivity of rain detectors was studied generating artificial raindrops of regular size. The sensitivity was evaluated in terms of the critical number of droplets that triggers the activation of the rain detector. We found that the sensitivity is affected by stationary, horizontal, and vertical droplet deposition methods. The critical number of droplets for the stationary deposition is higher than that for both horizontal and vertical depositions, which provides the maximum limit of droplets required to activate the detector. Based on our experiments considering surface temperature measurements and sensitivity tests, we suggest a revised version of surface temperature and sensitivity requirements for the qualification of impedance-based rain detectors.

• Journal of the Korean Society for Precision Engineering
• /
• v.13 no.2
• /
• pp.133-145
• /
• 1996

This paper presents an indirect cutting force measuring system, which uses the current signals from the AC servo drive units of the horizontal machining center, with its applications to the adaptive regulation of the cutting forces in various milling processes and to the on-line monitoring of tool breakage. A typical model for the feed-drive control system of a horizontal machining center is developed to analyze cutting force measurement from the drive motor. The pulsating milling forces can be measured indirectly within the bandwidth of the current feedback control loop of the feed-drive system. It is shown that the indirectly measured cutting force signals can be used in the adaptive controller for cutting force regulation. The whole scheme has been embedded in the commercial machining center and a series of cutting experiments on the face cutting processes are performed. The adaptive controller reveals reliable cutting force regulating capability against the various cutting conditions. It is also shown that the tool breakage in milling can be detected within one spindle revolution by adaptively filtering the current signals. The effect of the cutter run-out has been considered for the reliable on-line detection of tool breakage.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2006.06a
• /
• pp.239-244
• /
• 2006

Borehole radar reflection surveys were carried out in the horizontal borehole to detect EDZ while constructing the tunnel for the research facility of the nuclear waste disposal in Korea. The horizontal borehole has been bored at a length of 35 m from shelter to be parallel with the tunnel which would be planed. While the tunnel has been constructing with the explosive excavation, the borehole radar reflection surveys carried out 5 times with the interval of 2 or 4 days for monitoring EDZ. The most typical change of the reflection event resulted from the face of the wall of tunnel which had been produced newly by the excavation of the tunnel daily, EDZ has been detected with constructing images of difference between two measurement stages, and also the change of EDZ through the time has been done, which is due to the generation of crack and weakening of the rock strength of the face of the tunnel's wall near previous portion of the face of a blind end of tunnel according to explosive excavation.

• The Journal of the Korea Contents Association
• /
• v.8 no.11
• /
• pp.1-9
• /
• 2008

Thanks to the multimedia technology development, it is possible to show image representations in high quality and to process images in various ways. Anisotropic diffusion as an effective diffusion filtering among many image preprocessing methods and postprocessing methods is used in reduction of speckle noises of ultrasound images, image restoration, edge detection, and image segmentation. However, the conventional anisotropic diffusion based on a cross-kernel causes the following problems. The problem is the concentration of edges in the vertical or horizontal directions. In this paper, a new anisotropic diffusion transform based on directions of gradient is proposed. The proposed method uses the eight directional square-kernel which is an expanded form of the cross-kernel. The proposed method is to select directions of small gradient based on square-kernel. Therefore, the range of proposed diffusion is selected adaptively according to the number of the directions of gradient. Experimental results show that the proposed method can decrease the concentration of edges in the vertical or horizontal directions, remove impulse noise. The image in high quality can be obtained as a result of the proposed method.

• Atmosphere
• /
• v.28 no.2
• /
• pp.223-232
• /
• 2018

The Global-Korean aviation Turbulence Guidance (G-KTG) system is developed using the operational Global Data Assimilation and Prediction System of Korea Meteorological Administration with 17-km horizontal grid spacing. The G-KTG system provides an integrated solution of various clear-air turbulence (CAT) diagnostics and mountain-wave induced turbulence (MWT) diagnostics for low [below 10 kft (3.05 km)], middle [10 kft (3.05 km) - 20 kft (6.10 km)], and upper [20 kft (6.10 km) - 50 kft (15.24 km)] levels. Individual CAT and MWT diagnostics in the G-KTG are converted to a 1/3 power of energy dissipation rate (EDR). 12-h forecast of the G-KTG is evaluated using 6-month period (2016.06~2016.11) of in-situ EDR observation data. The forecast skill is calculated by area under curve (AUC) where the curve is drawn by pairs of probabilities of detection of "yes" for moderate-or-greater-level turbulence events and "no" for null-level turbulence events. The AUCs of G-KTG for the upper, middle, and lower levels are 0.79, 0.69, and 0.63, respectively. Comparison of the upper-level G-KTG with the regional-KTG in East Asia reveals that the forecast skill of the G-KTG (AUC = 0.77) is similar to that of the regional-KTG (AUC = 0.79) using the Regional Data Assimilation and Prediction System with 12-km horizontal grid spacing.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.6
• /
• pp.29-35
• /
• 2010

While diversification of the use of video in the prevalence of cheap video equipment, augmented reality can print additional real-world images and video image. Although many recent advent augmented reality techniques, currently attempting to correct the character recognition is performed. In this paper characters marked with a visual marker recognition, and the color to match the marker color of the characters finds. And, it was shown on the screen by the character recognition. In this paper, by applying the phoneme pattern segmentation algorithm by the horizontal projection, we propose to segment the phoneme to match the six types of Hangul representation. Throughout the experiment sample of phoneme segmentation using augmented reality showed proceeding result at each step, and the experimental results was found to be that detection rate was above 90%.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2012.07a
• /
• pp.87-88
• /
• 2012

Because intra prediction modes in H.264 are determined by the brightness continuity between neighboring blocks, they can be used as a method for extracting edge information in the compression domain. However, if we just consider 9 intra prediction modes in H.264 as 9 different edge directions, we have the following two problems. First, intra prediction modes tend to yield too many edge blocks, generating unnecessary edge information. Second, we may not need all 9 directional edges (including the DC type) in H.264 intra prediction modes. For example, the EHD (edge histogram descriptor) in MPEG-7 defines only 4 directional edge types, namely horizontal, vertical, diagonal (HVD) edges with $0^{\circ}$, $90^{\circ}$, $45^{\circ}$, and $135^{\circ}$. Here, semi-diagonal (SD) edge types with $112.5^{\circ}$, $157.5^{\circ}$, $22.5^{\circ}$, and $67.5^{\circ}$ in the intra prediction modes in H.264 are not used. In this paper. we prepose a method that removes unnecessary edges from the intra prediction modes by utilizing the total average coefficient of 4x4 blocks in each slice and assign SD edges to HVD (horizontal, vertical, diagonal, $0^{\circ}$, $90^{\circ}$, $45^{\circ}$, $135^{\circ}$) edges by the contextual information of the neighboring blocks. Experimental results show that the edges determined by the proposed method in the compression domain are comparable to those of the previous edge detection methods in the spatial domain.