• Journal of the Korean Society for Precision Engineering
• /
• v.10 no.4
• /
• pp.170-179
• /
• 1993

In this study, self-organized neural network is used to solve the vorrespondence problem of the axial stereo image. Edge points are extracted from a pair of stereo images and then the edge points of rear image are assined to the output nodes of neural network. In the matching process, the two input nodes of neural networks are supplied with the coordi- nates of the edge point selected randomly from the front image. This input data activate optimal output node and its neighbor nodes whose coordinates are thought to be correspondence point for the present input data, and then their weights are allowed to updated. After several iterations of updating, the weights whose coordinates represent rear edge point are converged to the coordinates of the correspondence points in the front image. Because of the feature map properties of self-organized neural network, noise-free and smoothed depth data can be achieved.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2021.05a
• /
• pp.40-41
• /
• 2021

The construction industry has the highest occupational accidents/injuries among all industries. Korean government installed surveillance camera systems at construction sites to reduce occupational accident rates. Construction safety managers are monitoring potential hazards at the sites through surveillance system; however, the human capability of monitoring surveillance system with their own eyes has critical issues. Therefore, this study proposed to build a deep learning-based safety monitoring system that can obtain information on the recognition, location, identification of workers and heavy equipment in the construction sites by applying multiple-object tracking with instance segmentation. To evaluate the system's performance, we utilized the MS COCO and MOT challenge metrics. These results present that it is optimal for efficiently automating monitoring surveillance system task at construction sites.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.458-460
• /
• 2020

스마트미터의 기술 발달과 보급으로 인해 전력데이터의 수집이 보다 수월 해짐에 따라 각 시스템에 효율적인 맞춤 서비스 제공을 위한 전력 데이터 분석 기술에 관한 다양한 연구가 활발하게 진행되고 있다. 관련하여 본 논문에서는 숙박업소의 각 방마다 전력소비량을 측정 및 수집하여 전력소비패턴을 분석하고 특징 추출 및 XGBoost 를 이용한 머신러닝 분석방법으로 각 방의 사람 재실 여부를 판별하는 방법을 소개한다. 이와 같은 연구를 통해 추후 숙박업소 혹은 숙박업소를 이용하는 소비자들의 맞춤 서비스 제공에 응용 및 적용 할 수 있다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.248-249
• /
• 2022

In this paper, we propose a method for classifying and tracking objects based on information of multiple users obtained using RGB-D cameras. The 3D information and color information acquired through the RGB-D camera are acquired and information about each user is stored. We propose a user classification and location tracking algorithm in the entire image by calculating the similarity between users in the current frame and the previous frame through the information on the location and appearance of each user obtained from the entire image.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.631-634
• /
• 2008

인터넷을 이용한 사람들의 사회 참여가 확대되면서 다양한 의견(Opinion)들이 급속도로 증가하고 있으며 이러한 의견을 분석하여 유용한 정보로 활용하기 위한 연구가 활발히 진행되고 있다. 그 중에서도 상품리뷰는 기업에서 연구, 개발, 마케팅의 주요 자료로 사용되고 있으며 사용자가 상품의 구매를 결정하는 중요한 요인 중 하나로 작용하고 있다. 본 논문에서는 한국어로 이루어진 상품 리뷰를 분석하여 의견 자질(Feature)을 추출하고 분류(Classification)하는 시스템을 설계하고 구현하였다. 한글 의견 자질 추출을 위하여 먼저 한글 상품 리뷰를 분석하여 의견 사전을 구축하였다. 의견 사전으로는 의견 자질과 의견 어휘, 독립의견어휘, 의견 숙어, 부정어 등의 각기 다른 세부 사전을 구축하여 리뷰 분석 시 단계적으로 적용하여 정확도를 높일 수 있도록 설계하였다. 이렇게 구현된 시스템을 평가하기 위하여 각기 다른 3개의 도메인에서 실제 한국어 리뷰를 수집하여 실험을 수행하였으며 자질 추출에서는 평균 78.86% 정확률, 61.41% 재현율을, 극성 분류에서는 평균 69.46% 정확률, 42.26% 재현율을 나타냈다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.125-128
• /
• 2008

현대사회에서 건강을 해치는 요인으로 흡연, 당뇨, 비만 및 스트레스 등이 있다. 이와 같은 요인들로 순환기질환의 발병이 증가하고 있으며, 특히 심장 질환 사망률이 점차 증가하고 있는 실정이다. 이를 해결하기 위해 본 논문에서는 심장 질환에 대한 조기 진단을 위한 음성 분석학적 특징 요소를 분석하여 결과값을 추출하고자 한다. 이를 위해 본 논문에서는 대전 지역에 거주하고 있는 성인 남성중에서 심장 질환을 앓고 있는 환자들과 심장에 이상이 없는 정상인들로 피실험자 집단을 구성하고 이들의 음성을 수집하여 음성 분석학적 특징 요소들을 추출하고자 한다. 특히 동의보감에서 제시한 심장의 소리를 음성 공학적으로 입증하기 위해 제 5 포먼트와 지터 등의 출력값을 비교, 분석하고자 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.605-607
• /
• 2023

Recent research has highlighted the effectiveness of Deep Learning (DL) techniques in automating the detection of lung sound anomalies. However, the available lung sound datasets often suffer from limitations in both size and balance, prompting DL methods to employ data preprocessing such as augmentation and transfer learning techniques. These strategies, while valuable, contribute to the increased complexity of DL models and necessitate substantial training memory. In this study, we proposed a streamlined and lightweight DL method but effectively detects lung sound anomalies from small and imbalanced dataset. The utilization of 1D dilated convolutional neural networks enhances sensitivity to lung sound anomalies by efficiently capturing deep temporal features and small variations. We conducted a comprehensive evaluation of the ICBHI dataset and achieved a notable improvement over state-of-the-art results, increasing the average score of sensitivity and specificity metrics by 2.7%.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.5
• /
• pp.259-265
• /
• 2023

Pedestrian detection is used to effectively detect pedestrians in various situations based on deep learning. Pedestrian detection has difficulty detecting pedestrians due to problems such as camera performance, pedestrian description, height, and occlusion. Even in the same pedestrian, performance in detecting them can differ according to the height of the pedestrian. The height of general pedestrians encompasses various scales, such as those of infants, adolescents, and adults, so when the model is applied to one group, the extraction of data becomes inaccurate. Therefore, this study proposed a pedestrian detection method that fine-tunes the pedestrian area by Refining Layer and Feature Concatenation to consider various heights of pedestrians. Through this, the score and location value for the pedestrian area were finely adjusted. Experiments on four types of test data demonstrate that the proposed model achieves 2-5% higher average precision (AP) compared to Faster R-CNN and DRPN.

• International Journal of Computer Science & Network Security
• /
• v.23 no.10
• /
• pp.209-213
• /
• 2023

In recent years, the image processing mechanisms are used widely in several medical areas for improving earlier detection and treatment stages, in which the time factor is very important to discover the disease in the patient as possible as fast, especially in various cancer tumors such as the liver cancer. Liver cancer has been attracting the attention of medical and sciatic communities in the latest years because of its high prevalence allied with the difficult treatment. Statistics indicate that liver cancer, throughout world, is the one that attacks the greatest number of people. Over the time, study of MR images related to cancer detection in the liver or abdominal area has been difficult. Early detection of liver cancer is very important for successful treatment. There are few methods available to detect cancerous cells. In this paper, an automatic approach that integrates the intensity-based segmentation and k-means clustering approach for detection of cancer region in MRI scan images of liver.

• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.663-672
• /
• 2023

Most vehicle detection methods have poor vehicle feature extraction performance at night, and their robustness is reduced; hence, this study proposes a night vehicle detection method based on style transfer image enhancement. First, a style transfer model is constructed using cycle generative adversarial networks (cycleGANs). The daytime data in the BDD100K dataset were converted into nighttime data to form a style dataset. The dataset was then divided using its labels. Finally, based on a YOLOv5s network, a nighttime vehicle image is detected for the reliable recognition of vehicle information in a complex environment. The experimental results of the proposed method based on the BDD100K dataset show that the transferred night vehicle images are clear and meet the requirements. The precision, recall, mAP@.5, and mAP@.5:.95 reached 0.696, 0.292, 0.761, and 0.454, respectively.