• 한국정보처리학회논문지
• /
• 제5권8호
• /
• pp.2041-2049
• /
• 1998

소프트웨어의 품질을 향상시키며, 기존코드의 결함식별을 용이하게 하는 방법으로 프로그램의 후상태 종속성 분석을 통한 프로그램 ？살 및 유지보수지원 기법을 제안한다. 결함을 식별하고 분석하기에 이해도가 중요시 되는 교정유지보수를 위해서, 기존 코드의 분석 및 세그먼트화를 후상태 종속성모형(PSDG)을 이용하여 정적분할과 동적분할 및 의미분할의 장점을 살린 코드분할로 수행한다. 분할의 원리는 기존코드의 상태 종속성을 추적하여 그래프로 모형화한 후, 조각화(Clustering)와 강조분할(Highlighting)을 통해서 프로그램을 분할한다. PSDG 모형화의 결과로 비효율적인 프로그램 결함코드(Deadcode)의 식별 및 제거가 가능하며, 관련 프로그램 문장들을 일반화할 수 있고, 상태전이도 모형과의 확장연계로 분석 및 설계의 문서로 이용될 수 있다.

• 정보처리학회논문지B
• /
• 제17B권3호
• /
• pp.207-214
• /
• 2010

본 논문에서는 실제 도로환경의 신호등 및 속도표지판 영역 검출 및 인식 방법을 제안하였다. 밝기정보 및 HIS 컬러모델에기반한 색상정보를 이용하여 신호등을 인식하였다. 또한 HSI 컬러정보로부터 적색강도를 추정함으로써 속도 표지판을 검출하였다. 표지판의 경사여부를 판단하여 시계방향, 반시계방향으로 각각 표지판을 회전시켜 기울기를 보정한 후 인식을 행함으로써 인식률을 제고하였다. 도로환경의 동영상을 대상으로 인식을 행한 결과 신호등과 속도표지판이 혼합된 영상에서도 매우 강건한 인식 결과를 보인다.

• 인터넷정보학회논문지
• /
• 제22권3호
• /
• pp.37-43
• /
• 2021

본 논문에서는 모바일 환경에서 실시간으로 전면과 후면 카메라의 객체를 인식하여 객체 픽셀의 영역을 분할하고 이미지 처리를 통해 합성하는 방법을 연구하였다. 이를 위해 Apple사의 iOS에서 제공하는 듀얼 카메라에 DeepLabV3 머신러닝 모델을 적용하여 객체를 분할하였다. 또한 이미지 합성 및 후처리를 위해 Apple사의 코어 이미지와 코어 그래픽 라이브러리를 이용하여 영역의 배경 제거 및 합성 방식을 제안하고 구현하였다. 또한, 이전 연구에 비해 CPU 사용량을 개선하였고 깊이와 DeepLabV3의 처리 속도를 비교하여 처리 결과에 영향을 주는 요소를 분석하였다. 마지막으로 이 두 방식을 활용한 카메라 애플리케이션을 개발하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권5호
• /
• pp.1778-1797
• /
• 2021

Document image binarization is an important pre-processing step in document analysis and archiving. The state-of-the-art models for document image binarization are variants of encoder-decoder architectures, such as FCN (fully convolutional network) and U-Net. Despite their success, they still suffer from three limitations: (1) reduced feature map resolution due to consecutive strided pooling or convolutions, (2) multiple scales of target objects, and (3) reduced localization accuracy due to the built-in invariance of deep convolutional neural networks (DCNNs). To overcome these three challenges, we propose an improved semantic segmentation model, referred to as DP-LinkNet, which adopts the D-LinkNet architecture as its backbone, with the proposed hybrid dilated convolution (HDC) and spatial pyramid pooling (SPP) modules between the encoder and the decoder. Extensive experiments are conducted on recent document image binarization competition (DIBCO) and handwritten document image binarization competition (H-DIBCO) benchmark datasets. Results show that our proposed DP-LinkNet outperforms other state-of-the-art techniques by a large margin. Our implementation and the pre-trained models are available at https://github.com/beargolden/DP-LinkNet.

• 드라이브 ㆍ 컨트롤
• /
• 제19권4호
• /
• pp.104-109
• /
• 2022

For autonomous driving research using AI, datasets collected from road environments play an important role. In other countries, various datasets such as CityScapes, A2D2, and BDD have already been released, but datasets suitable for the domestic road environment still need to be provided. This paper analyzed and verified the dataset reflecting the Korean driving environment. In order to verify the training dataset, the class imbalance was confirmed by comparing the number of pixels and instances of the dataset. A similar A2D2 dataset was trained with the same deep learning model, ConvNeXt, to compare and verify the constructed dataset. IoU was compared for the same class between two datasets with ConvNeXt and mIoU was compared. In this paper, it was confirmed that the collected dataset reflecting the driving environment of Korea is suitable for learning.

• Smart Structures and Systems
• /
• 제30권1호
• /
• pp.17-34
• /
• 2022

For steel structures, fatigue cracks are critical damage induced by long-term cycle loading and distortion effects. Vision-based crack detection can be a solution to ensure structural integrity and performance by continuous monitoring and non-destructive assessment. A critical issue is to distinguish cracks from other features in captured images which possibly consist of complex backgrounds such as handwritings and marks, which were made to record crack patterns and lengths during periodic visual inspections. This study presents a parametric study on image-based crack identification for orthotropic steel bridge decks using captured images with complicated backgrounds. Firstly, a framework for vision-based crack segmentation using the atrous convolution-based Deeplapv3+ network (ACDN) is designed. Secondly, features on crack images are labeled to build three databanks by consideration of objects in the backgrounds. Thirdly, evaluation metrics computed from the trained ACDN models are utilized to evaluate the effects of obstacles on crack detection results. Finally, various training parameters, including image sizes, hyper-parameters, and the number of training images, are optimized for the ACDN model of crack detection. The result demonstrated that fatigue cracks could be identified by the trained ACDN models, and the accuracy of the crack-detection result was improved by optimizing the training parameters. It enables the applicability of the vision-based technique for early detecting tiny fatigue cracks in steel structures.

• 방송공학회논문지
• /
• 제26권7호
• /
• pp.868-876
• /
• 2021

Deploying deep neural network model training performs remarkable performance in the fields of Object detection and Instance segmentation. To train these models, features are first extracted from the input image using a backbone network. The extracted features can be reused by various tasks. Research has been actively conducted to serve various tasks by using these learned features. In this process, standardization discussions about encoding, decoding, and transmission methods are proceeding actively. In this scenario, it is necessary to analyze the response characteristics of features against various distortions that may occur in the data transmission or data compression process. In this paper, experiment was conducted to inject various distortions into the feature in the object recognition task. And analyze the mAP (mean Average Precision) metric between the predicted value output from the neural network and the target value as the intensity of various distortions was increased. Experiments have shown that features are more robust to distortion than images. And this points out that using the feature as transmission means can prevent the loss of information against the various distortions during data transmission and compression process.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2024년도 제69차 동계학술대회논문집 32권1호
• /
• pp.45-46
• /
• 2024

본 논문에서는 Self-Attention 기반 딥러닝 기법인 Dense Prediction Transformer(DPT) 모델을 MVTec Anomaly Detection(MVTec AD) 데이터셋에 적용하여 실제 산업 제품 이미지 내 이상 부분을 분할하는 연구를 진행하였다. DPT 모델의 적용을 통해 기존 Convolutional Neural Network(CNN) 기반 이상 탐지기법의 한계점인 지역적 Feature 추출 및 고정된 수용영역으로 인한 문제를 개선하였으며, 실제 산업 제품 데이터에서의 이상 분할 시 기존 주력 기법인 U-Net의 구조를 적용한 최고 성능의 모델보다 1.14%만큼의 성능 향상을 보임에 따라 Self-Attention 기반 딥러닝 기법의 적용이 산업 제품 이상 분할에 효과적임을 입증하였다.

• Journal of information and communication convergence engineering
• /
• 제15권3호
• /
• pp.187-192
• /
• 2017

Autonomous driving vehicle research demands complex road and lane understanding such as lane departure warning, adaptive cruise control, lane keeping and centering, lane change and turn assist, and driving under complex road conditions. A fast and robust road lane detection subsystem is a basic but important building block for this type of research. In this paper, we propose a method that performs road lane detection from black box input. The proposed system applies Random Sample Consensus to find the best model of road lanes passing through divided regions of the input image under HSV color model. HSV color model is chosen since it explicitly separates chromaticity and luminosity and the narrower hue distribution greatly assists in later segmentation of the frames by limiting color saturation. The implemented method was successful in lane detection on real world on-board testing, exhibiting 86.21% accuracy with 4.3% standard deviation in real time.

• Wind and Structures
• /
• 제21권6호
• /
• pp.709-726
• /
• 2015

In this study, a novel vision-based bolt-loosening monitoring technique is proposed for bolted joints connecting tubular steel segments of the wind turbine tower (WTT) structure. Firstly, a bolt-loosening detection algorithm based on image processing techniques is developed. The algorithm consists of five steps: image acquisition, segmentation of each nut, line detection of each nut, nut angle estimation, and bolt-loosening detection. Secondly, experimental tests are conducted on a lab-scale bolted joint model under various bolt-loosening scenarios. The bolted joint model, which is consisted of a ring flange and 32 sets of bolt and nut, is used for simulating the real bolted joint connecting steel tower segments in the WTT. Finally, the feasibility of the proposed vision-based technique is evaluated by bolt-loosening monitoring in the lab-scale bolted joint model.