• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.602-605
• /
• 2021

최근 인공지능 연구가 활발히 진행되고 있는 가운데 국내외에서 오픈 데이터셋을 제공하고 있어 기술개발이 가속화되고 있다. 데이터셋은 지도학습을 위한 학습데이터로 라벨링 데이터를 포함하고 있어 다양한 라벨링 기능이 적용된 도구 개발이 필요하다. 본 논문에서는 의료영상의 라벨링 데이터를 정교하고 빠르게 생성하기 위한 라벨링 웹 애플리케이션에 대해서 기술한다. 이를 구현하기 위해서 Back Projection, Grabcut 기법을 이용한 반자동 방식과 기계학습 모델을 통해서 예측한 자동 방식의 라벨링 기능을 구현하였다. 이와 관련하여 라벨링 기능별 수행 결과를 근감소증 진단을 위한 영상 라벨링 수행결과와 정량분석 결과를 보였다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.5
• /
• pp.287-294
• /
• 2013

Binary connected-component labeling is widely used in the fields of the image processing and the computer vision. Many kinds of labeling techniques have been developed, and two-scan is known as the fastest method among them. Traditionally pixel-based scan masks have been used for the first stage of the two-scan. Recently, block-based labeling techniques were introduced by C. Grana et. al. and L. He et. al. They are faster than pixel-based labeling methods. In this paper, we propose a new binary connected-component labeling technique with block-based labels and a pixel-based scan mask. The experimental results with various images show that the proposed method is faster than the He's which is known as the fastest method currently. The amount of performance enhancement is averagely from 3.9% to 22.4% according to the sort of the images.

• Journal of Digital Convergence
• /
• v.11 no.4
• /
• pp.259-266
• /
• 2013

Binary connected-components labeling, which is widely used in the field of the pattern recognition, has been researched for a long time as one of the basic image processing techniques. Two-scan algorithm has been mainly used in the researches of the connected-components labeling. Recently, for the first scan in the two-scan algorithm, block-based labeling approaches have been used and reported as the fastest methods. In this paper, a new efficient scan mask for connected-components labeling with a block-based labeling approach is proposed. Labeling with the new pixel-based scan mask is more efficient than any other existing method. The results of the experiments show that the proposed method is faster than the existing fastest method.

• Journal of Nutrition and Health
• /
• v.43 no.2
• /
• pp.197-206
• /
• 2010

Now a days, people eat outside of the home more and more frequently. Menu labeling can help people make more informed decisions about the foods they eat and help them maintain a healthy diet. This study was conducted to develop menu labeling system using Nutri-API (Nutrition Analysis Application Programming Interface). This system offers convenient user interface and menu labeling information with printout format. This system provide useful functions such as new food/menu nutrients information, retrieval food semantic service, menu plan with subgroup and nutrient analysis informations and print format. This system provide nutritive values with nutrient information and ratio of 3 major energy nutrients. MLS system can analyze nutrients for menu and each subgroup. And MLS system can display nutrient comparisons with DRIs and % Daily Nutrient Values. And also this system provide 6 different menu labeling formate with nutrient information. Therefore it can be used by not only usual people but also dietitians and restaurant managers who take charge of making a menu and experts in the field of food and nutrition. It is expected that Menu Labeling System (MLS) can be useful of menu planning and nutrition education, nutrition counseling and expert meal management.

• Progress in Medical Physics
• /
• v.24 no.2
• /
• pp.108-118
• /
• 2013

Currently, an arterial spin labeling (ASL) magnetic resonance imaging (MRI) technique does not routinely used in clinical studies to measure perfusion in brain because optimization of imaging protocol is required to obtain optimal perfusion signals. Therefore, the objective of this study was to investigate changes of perfusion-weighed signal intensities with varying several parameters on a pulsed arterial spin labeling MRI technique obtained from a 3T MRI system. We especially evaluated alternations of ASL-MRI signal intensities on special brain areas, including in brain tissues and lobes. The signal targeting with alternating radiofrequency (STAR) pulsed ASL method was scanned on five normal subjects (mean age: 36 years, range: 29~41 years) on a 3T MRI system. Four parameters were evaluated with varying: 1) the labeling gap, 2) the labeling delay time, 3) the labeling thickness, and 4) the slice scan order. Signal intensities were obtained from the perfusion-weighted imaging on the gray and white matters and brain lobes of the frontal, parietal, temporal, and occipital areas. The results of this study were summarized: 1) Perfusion-weighted signal intensities were decreased with increasing the labeling gap in the bilateral gray matter areas and were least affected on the parietal lobe, but most affected on the occipital lobe. 2) Perfusion-weighted signal intensities were decreased with increasing the labeling delay time until 400 ms, but increased up to 1,000 ms in the bilateral gray matter areas. 3) Perfusion-weighted signal intensities were increased with increasing the labeling thickness until 120 mm in both the gray and white matter. 4) Perfusion-weighted signal intensities were higher descending scans than asending scans in both the gray and white matter. We investigated changes of perfusion-weighted signal intensities with varying several parameters in the STAR ASL method. It should require having protocol optimization processing before applying in patients. It has limitations to apply the ASL method in the white matter on a 3T MRI system.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.87-89
• /
• 2013

기존의 영상 라벨러는 이중 스캔 연결 성분 라벨링 방식이 주로 사용되어 왔다. 그러나, 이중 스캔 연결 성분 라벨링 방식은 별도의 라벨 병합 구간 때문에 실시간 처리와 다채널 처리가 어렵다. 최근 단일 스캔 연결 성분 라벨링 방법들이 제안되었지만, 이들 역시 스캔 횟수는 줄일 수 있으나 별도의 라벨 병합 구간을 제거하지는 못하였다. 본 논문에서는 별도의 라벨 병합 구간을 갖지 않는 단일 스캔 연결 성분 라벨러 구조를 제안한다. 제안하는 구조는 라인 스캔과 동시에 라벨 병합이 이루어지기 때문에 고속 동작하고, 예상 가능한 프레임 처리율을 가지며, 다채널 처리가 가능하다. 제안하는 라벨러 구조는 FPGA로 설계되었고 동작이 검증되었으며, 성능 실험을 통해 고속 동작과 다채널 영상 처리에 적합함을 확인하였다.

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.465-466
• /
• 2015

전 세계적으로 에너지 수요의 증가에 따라 에너지를 효율적으로 이용해야 한다는 인식이 확산되고 있다. 에너지 절약과 환경보호를 위해 세계 각국에서는 에너지효율제품의 보급화를 위해 에너지 효율기준 및 라벨링 제도를 시행하고 있다. 본 논문에서는 개발도상국인 에티오피아 선정하여 구체적인 사례를 적용하여 보았다. 효율적인 에너지 관리를 위한 제도로 최저소비효율기준(MEPS : Minimum Energy Performance Standard)과 에너지소비효율등급라벨표시(이하 '라벨링')에 대한 내용을 살펴보고 국내, 해외선진국, 개발도상국 등 여러 국가의 운영 현황을 조사하였다. 그리고 에티오피아에 적용하기 위해 여러 상황을 고려한 맞춤 라벨링을 설계해 보고자 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.322-325
• /
• 2022

삼킴 장애 환자는 뇌졸중, 치매, 외상성 뇌손상, 파킨슨병, 암이 주요 원인으로 급속히 증가하고 있다. 특히 고령화 사회가 되면서 더욱 삼킴 장애 환자는 늘어날 것으로 전망하고 있다. 고령 환자의 삼킴 이상의 진단을 위해 가장 많이 사용하고 있는 검사법으로는 비디오 조영 삼킴 검사(VFSS)이다. VFSS는 진단에 있어서 숙련된 전문의가 필요하기 때문에 대학병원 급에서 주로 시행하며, 고령 환자에게는 분석 결과를 상담받을 때까지 오랜 시간을 소요해야하는 문제점들이 있다. 본 논문에서는 삼킴 장애 진단을 위한 의료영상 관리 및 라벨링 시스템에 대해서 기술한다. 이를 구현하기 위해 서버에서 대용량 멀티프레임 영상을 성능 저하 없이 핸들링 하고 라벨링 데이터 생성을 위한 라벨링 툴을 구현하였다. 차후 라벨링 데이터를 생성하고 학습을 통하여 삼킴 장애 진단을 위한 인공지능 모델을 개발하고자 한다.

• Journal of Environmental Policy
• /
• v.13 no.3
• /
• pp.109-133
• /
• 2014

With a growing concern about climate change and green house gases mitigation, carbon labeling policy has been launched in several countries as an environmental policy which connects low carbon production to low carbon consumption. This research aims to propose a model that explains consumers' attitude and brand loyalty toward carbon labeling products. This model specifies the consumer's psychological processes by which consumer values, such as autonomy and environmental values, affect carbon labeling product and corporate images and finally form brand loyalty toward carbon labeling products. Panel data were collected in two separate surveys and analyzed using a structural equation technique. Results are summarized as follows. First, consumers' autonomy value(AV) positively affects locus of control(LC) and corporate image(CI). Second, consumers' environmental value(EV) positively influences perceived consumer effectiveness(PCE), which in turn has a negative effect on perceived barriers(PB). Perceived barriers finally affect product image(PI) negatively. Third, both corporate image and product image have causal relationships with brand loyalty. Our results suggest that carbon labeling policy contributes not only to the reduction of greenhouse gases but also to the increase of consumers' attitude and brand loyalty toward carbon labeling products. This research also provides governments with directions for efficient environmental policy and firms with guidance on effective marketing strategies about carbon labeling.

• The Monthly Technology and Standards
• /
• s.93
• /
• pp.19-19
• /
• 2009

미 농무부는 생물기반제품의 우선구매 프로그램을 통하여 생물기반제품의 이용을 촉진하고, 자발적인 라벨링 프로그램은 더 많은 제품정보를 원하는 소비자에게 도움이 될 것이다.