• International Journal of Internet, Broadcasting and Communication
• /
• 제14권2호
• /
• pp.47-54
• /
• 2022

For safe driving of autonomous vehicles, road damage detection is very important to lower the potential risk. In order to ensure safety while an autonomous vehicle is driving on the road, technology that can cope with various obstacles is required. Among them, technology that recognizes static obstacles such as poor road conditions as well as dynamic obstacles that may be encountered while driving, such as crosswalks, manholes, hollows, and speed bumps, is a priority. In this paper, we propose a method to extract similarity of images and find damaged road images using OpenCV image processing and CNN algorithm. To implement this, we trained a CNN model using 280 training datasheets and 70 test datasheets out of 350 image data. As a result of training, the object recognition processing speed and recognition speed of 100 images were tested, and the average processing speed was 45.9 ms, the average recognition speed was 66.78 ms, and the average object accuracy was 92%. In the future, it is expected that the driving safety of autonomous vehicles will be improved by using technology that detects road obstacles encountered while driving.

• Journal of Information Processing Systems
• /
• 제16권1호
• /
• pp.224-237
• /
• 2020

Recognition of banknote serial number is one of the important functions for intelligent banknote counter implementation and can be used for various purposes. However, the previous character recognition method is limited to use due to the font type of the banknote serial number, the variation problem by the solid status, and the recognition speed issue. In this paper, we propose an aspect ratio based character region segmentation and a convolutional neural network (CNN) based banknote serial number recognition method. In order to detect the character region, the character area is determined based on the aspect ratio of each character in the serial number candidate area after the banknote area detection and de-skewing process is performed. Then, we designed and compared four types of CNN models and determined the best model for serial number recognition. Experimental results showed that the recognition accuracy of each character was 99.85%. In addition, it was confirmed that the recognition performance is improved as a result of performing data augmentation. The banknote used in the experiment is Indian rupee, which is badly soiled and the font of characters is unusual, therefore it can be regarded to have good performance. Recognition speed was also enough to run in real time on a device that counts 800 banknotes per minute.

• 한국멀티미디어학회논문지
• /
• 제22권2호
• /
• pp.136-145
• /
• 2019

The crowded environment of a domestic pig farm is highly vulnerable to the spread of infectious diseases such as foot-and-mouth disease, and studies have been conducted to automatically analyze behavior of pigs in a crowded pig farm through a video surveillance system using a camera. Although it is required to correctly separate occluding pigs for tracking each individual pigs, extracting the boundaries of the occluding pigs fast and accurately is a challenging issue due to the complicated occlusion patterns such as X shape and T shape. In this study, we propose a fast and accurate method to separate occluding pigs not only by exploiting the characteristics (i.e., one of the fast deep learning-based object detectors) of You Only Look Once, YOLO, but also by overcoming the limitation (i.e., the bounding box-based object detector) of YOLO with the test-time data augmentation of rotation. Experimental results with two-pigs occlusion patterns show that the proposed method can provide better accuracy and processing speed than one of the state-of-the-art widely used deep learning-based segmentation techniques such as Mask R-CNN (i.e., the performance improvement over Mask R-CNN was about 11 times, in terms of the accuracy/processing speed performance metrics).

• Geomechanics and Engineering
• /
• 제29권5호
• /
• pp.523-533
• /
• 2022

Several prediction model of penetration rate (PR) of tunnel boring machines (TBMs) have been focused on applying to design stage. In construction stage, however, the expected PR and its trends are changed during tunneling owing to TBM excavation skills and the gap between the investigated and actual geological conditions. Monitoring the PR during tunneling is crucial to rescheduling the excavation plan in real-time. This study proposes a sequential prediction method applicable in the construction stage. Geological and TBM operating data are collected from Gunpo cable tunnel in Korea, and preprocessed through normalization and augmentation. The results show that the sequential prediction for 1 ring unit prediction distance (UPD) is R²≥0.79; whereas, a one-step prediction is R²≤0.30. In modeling algorithm, a gradient boosted regression tree (GBRT) outperformed a least square-based linear regression in sequential prediction method. For practical use, a simple equation between the R² and UPD is proposed. When UPD increases R² decreases exponentially; In particular, UPD at R²=0.60 is calculated as 28 rings using the equation. Such a time interval will provide enough time for decision-making. Evidently, the UPD can be adjusted depending on other project and the R² value targeted by an operator. Therefore, a calculation process for the equation between the R² and UPD is addressed.

• 사상체질의학회지
• /
• 제34권3호
• /
• pp.31-40
• /
• 2022

Objectives Sasang constitutional medicine is a traditional Korean medicine that classifies humans into four constitutions in consideration of individual differences in physical, psychological, and physiological characteristics. In this paper, we proposed a method to classify Taeeum person (TE) and Non-Taeeum person (NTE), Soeum person (SE) and Non-Soeum person (NSE), and Soyang person (ST) and Non-Soyang person (NSY) using a convolutional neural network with only facial images. Methods Based on the convolutional neural network VGG16 architecture, transfer learning is carried out on the facial images of 3738 subjects to classify TE and NTE, SE and NSE, and SY and NSY. Data augmentation techniques are used to increase classification performance. Results The classification performance of TE and NTE, SE and NSE, and SY and NSY was 77.24%, 85.17%, and 80.18% by F1 score and 80.02%, 85.96%, and 72.76% by Precision-Recall AUC (Area Under the receiver operating characteristic Curve) respectively. Conclusions It was found that Soeum person had the most heterogeneous facial features as it had the best classification performance compared to the rest of the constitution, followed by Taeeum person and Soyang person. The experimental results showed that there is a possibility to classify constitutions only with facial images. The performance is expected to increase with additional data such as BMI or personality questionnaire.

• Nuclear Engineering and Technology
• /
• 제55권6호
• /
• pp.2096-2106
• /
• 2023

Rotating machinery is widely applied in important equipment of nuclear power plants (NPPs), such as pumps and valves. The research on intelligent fault diagnosis of rotating machinery is crucial to ensure the safe operation of related equipment in NPPs. However, in practical applications, data-driven fault diagnosis faces the problem of small and imbalanced samples, resulting in low model training efficiency and poor generalization performance. Therefore, a deep convolutional conditional generative adversarial network (DCCGAN) is constructed to mitigate the impact of imbalanced samples on fault diagnosis. First, a conditional generative adversarial model is designed based on convolutional neural networks to effectively augment imbalanced samples. The original sample features can be effectively extracted by the model based on conditional generative adversarial strategy and appropriate number of filters. In addition, high-quality generated samples are ensured through the visualization of model training process and samples features. Then, a deep convolutional neural network (DCNN) is designed to extract features of mixed samples and implement intelligent fault diagnosis. Finally, based on multi-fault experimental data of motor and bearing, the performance of DCCGAN model for data augmentation and intelligent fault diagnosis is verified. The proposed method effectively alleviates the problem of imbalanced samples, and shows its application value in intelligent fault diagnosis of actual NPPs.

• 한국컴퓨터정보학회논문지
• /
• 제27권12호
• /
• pp.29-40
• /
• 2022

본 논문에서는 딥러닝 기법의 하나인 Mask R-CNN과 랜덤포레스트 분류기를 이용해 당뇨병성 망막병증의 병리학적인 특징을 검출하고 분석하여 자동 진단하는 시스템을 연구하였다. 당뇨병성 망막병증은 특수장비로 촬영한 안저영상을 통해 진단할 수 있는데 밝기, 색조 및 명암은 장치에 따라 다를 수 있으며 안과 전문의의 의료적 판단을 도울 인공지능을 이용한 자동진단 시스템 연구와 개발이 가능하다. 이 시스템은 미세혈관류와 망막출혈을 Mask R-CNN 기법으로 검출하고, 후처리 과정을 거쳐 랜덤포레스트 분류기를 이용하여 안구의 정상과 비정상 상태를 진단한다. Mask R-CNN 알고리즘의 검출 성능 향상을 위해 이미지 증강 작업을 실시하여 학습을 진행하였으며 검출 정확도 측정을 위한 평가지표로는 다이스 유사계수와 Mean Accuracy를 사용하였다. 비교군으로는 Faster R-CNN 기법을 사용하였고 본 연구를 통한 검출 성능은 평균 90%의 다이스 계수를 통한 정확도를 나타내었으며 Mean Accuracy의 경우 91% 정확도의 검출 성능을 보였다. 검출된 병리증상을 토대로 랜덤포레스트 분류기를 학습하여 당뇨병성 망막 병증을 진단한 경우 99%의 정확도를 보였다.

• 대한방사선기술학회지:방사선기술과학
• /
• 제45권1호
• /
• pp.41-47
• /
• 2022

The purpose of this study is to apply a deep learning model that can distinguish lung perfusion and lung ventilation images in nuclear medicine, and to evaluate the image classification ability. Image data pre-processing was performed in the following order: image matrix size adjustment, min-max normalization, image center position adjustment, train/validation/test data set classification, and data augmentation. The convolutional neural network(CNN) structures of VGG-16, ResNet-18, Inception-ResNet-v2, and SE-ResNeXt-101 were used. For classification model evaluation, performance evaluation index of classification model, class activation map(CAM), and statistical image evaluation method were applied. As for the performance evaluation index of the classification model, SE-ResNeXt-101 and Inception-ResNet-v2 showed the highest performance with the same results. As a result of CAM, cardiac and right lung regions were highly activated in lung perfusion, and upper lung and neck regions were highly activated in lung ventilation. Statistical image evaluation showed a meaningful difference between SE-ResNeXt-101 and Inception-ResNet-v2. As a result of the study, the applicability of the CNN model for lung scintigraphy classification was confirmed. In the future, it is expected that it will be used as basic data for research on new artificial intelligence models and will help stable image management in clinical practice.

• 마이크로전자및패키징학회지
• /
• 제30권4호
• /
• pp.69-78
• /
• 2023

산업 인공지능의 발달과 함께 반도체의 수요가 크게 증가하고 있다. 시장 수요에 대응하기 위해 패키징 공정에서 자동 결함 검출의 중요성 역시 증가하고 있다. 이에 따라, 패키지의 자동 불량 검사를 위한 딥러닝 기반의 방법론들의 연구가 활발히 이루어 지고 있다. 딥러닝 기반의 모델은 학습을 위해서 대량의 고해상도 데이터를 필요로 하나, 보안이 중요한 반도체 분야의 특성상 관련 데이터의 공유 및 레이블링이 쉽지 않아 모델의 학습이 어려운 한계를 지니고 있다. 또한 고해상도 이미지를 생성하기 위해 상당한 컴퓨팅 자원이 요구되는데, 본 연구에서는 분할정복 접근법을 통해 적은 컴퓨팅 자원으로 딥러닝 모델 학습을 위한 충분한 양의 데이터를 확보하는 방법을 소개한다. 제안된 방법은 높은 해상도의 이미지를 분할하고 각 영역에 조건 레이블을 부여한 후, 독립적인 부분 영역과 경계를 학습시켜, 경계 손실이 일관적인 이미지를 생성하도록 유도한다. 이후, 분할된 이미지를 하나로 통합하여, 최종적으로 모델이 고해상도의 이미지를 생성하도록 구성하였다. 실험 결과, 본 연구를 통해 증강된 이미지들은 높은 효율성, 일관성, 품질 및 범용성을 보였다.

• 한국수자원학회논문집
• /
• 제54권12호
• /
• pp.1233-1242
• /
• 2021

본 연구에서는 비정형 데이터인 사진자료를 이용하여 침수의 발생여부를 판단하는 모델을 개발하였다. 침수분류를 모델 개발을 위하여 CNN기반의 VGG16, VGG19을 이용하였다. 모델을 개발하기 위하여 침수사진과 침수가 발생하지 않은 사진을 웹크롤링 방법을 이용하여 사진을 수집하였다. 웹크롤링 방법을 이용하여 수집한 데이터는 노이즈 데이터가 포함되어 있기 때문에 1차적으로 본 연구와 상관없는 데이터는 소거하였으며, 2차적으로 모델 적용을 위하여 224 × 224로 사진 사이즈를 일괄 변경하였다. 또한 사진의 다양성을 위해서 사진의 각도를 변환하여 이미지 증식을 수행하였으며. 최종적으로 침수사진 2,500장과 침수가 발생하지 않은 사진 2,500장을 이용하여 학습을 수행하였다. 모델 평가결과 모델의 평균 분류성능은 97%로 나타났으며. 향후 본 연구결과를 통하여 개발된 모델을 CCTV관제센터 시스템에 탑재한다면 신속하게 침수피해에 대한 대처가 이루어 질 수 있을 것이라 판단된다.