• Tunnel and Underground Space
• /
• v.29 no.3
• /
• pp.184-196
• /
• 2019

This study aims to extract a 3D image of micro-cracks generated by hydraulic fracturing tests, using the deep learning method and X-ray computed tomography images. The pixel-level cracks are difficult to be detected via conventional image processing methods, such as global thresholding, canny edge detection, and the region growing method. Thus, the convolutional neural network-based encoder-decoder network is adapted to extract and analyze the micro-crack quantitatively. The number of training data can be acquired by dividing, rotating, and flipping images and the optimum combination for the image augmentation method is verified. Application of the optimal image augmentation method shows enhanced performance for not only the validation dataset but also the test dataset. In addition, the influence of the original number of training data to the performance of the deep learning-based neural network is confirmed, and it leads to succeed the pixel-level crack detection.

• Journal of IKEEE
• /
• v.26 no.3
• /
• pp.468-474
• /
• 2022

In this paper, we propose a method to implement a real-time fire alarm system using deep learning. The deep learning image dataset for fire alarms acquired 1,500 sheets through the Internet. If various images acquired in a daily environment are learned as they are, there is a disadvantage that the learning accuracy is not high. In this paper, we propose a fire image data expansion method to improve learning accuracy. The data augmentation method learned a total of 2,100 sheets by adding 600 pieces of learning data using brightness control, blurring, and flame photo synthesis. The expanded data using the flame image synthesis method had a great influence on the accuracy improvement. A real-time fire detection system is a system that detects fires by applying deep learning to image data and transmits notifications to users. An app was developed to detect fires by analyzing images in real time using a model custom-learned from the YOLO V4 TINY model suitable for the Edge AI system and to inform users of the results. Approximately 10% accuracy improvement can be obtained compared to conventional methods when using the proposed data.

• Journal of Positioning, Navigation, and Timing
• /
• v.12 no.1
• /
• pp.23-35
• /
• 2023

Korea Augmentation Satellite System (KASS) is a satellite-based augmentation system (SBAS) that provides approach procedure with vertical guidance-I (APV-I) level corrections and integrity information to Korea territory. KASS is used to monitor navigation performance in real-time, and this paper introduces the design, implementation, and verification process of mission monitoring (MIMO) in KASS. MIMO was developed in compliance with the Minimum Operational Performance Standards of the Radio Technical Commission for Aeronautics for Global Positioning System (GPS)/SBAS airborne equipment. In this study, the MIMO system was verified by comparing and analyzing the outputs of reference tools. Additionally, the definition and derivation method of accuracy, integrity, continuity, and availability subject to MIMO were examined. The internal and external interfaces and functions were then designed and implemented. The GPS data pre-processing was minimized during the implementation to evaluate the navigation performance experienced by general users. Subsequently, tests and verification methods were used to compare the obtained results based on reference tools. The test was performed using the KASS dataset, which included GPS and SBAS observations. The decoding performance of the developed MIMO was identical to that of the reference tools. Additionally, the navigation performance was verified by confirming the similarity in trends. As MIMO is a component of KASS used for real-time monitoring of the navigation performance of SBAS, the KASS operator can identify whether an abnormality exists in the navigation performance in real-time. Moreover, the preliminary identification of the abnormal point during the post-processing of data can improve operational efficiency.

• Journal of the Korea Institute of Building Construction
• /
• v.23 no.2
• /
• pp.197-207
• /
• 2023

Concrete is a widely used material due to its excellent compressive strength and durability. However, depending on the surrounding environment and the characteristics of the materials used in the construction, various defects may occur, such as cracks on the surface and subsidence of the structure. The detects on the surface of the concrete structure occur after completion or over time. Neglecting these cracks may lead to severe structural damage, necessitating regular safety inspections. Traditional visual inspections of concrete walls are labor-intensive and expensive. This research presents a deep learning-based semantic segmentation model designed to detect cracks in concrete walls. The model addresses surface defects that arise from aging, and an image augmentation technique is employed to enhance feature extraction and generalization performance. A dataset for semantic segmentation was created by combining publicly available and self-generated datasets, and notable semantic segmentation models were evaluated and tested. The model, specifically trained for concrete wall fracture detection, achieved an extraction performance of 81.4%. Moreover, a 3% performance improvement was observed when applying the developed augmentation technique.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.1149-1161
• /
• 2021

With a trend of the utilization of computer vision for satellite images, cloud detection using deep learning also attracts attention recently. In this study, we conducted a U-Net cloud detection modeling using SPARCS (Spatial Procedures for Automated Removal of Cloud and Shadow) Cloud Dataset with the image data augmentation and carried out 10-fold cross-validation for an objective assessment of the model. Asthe result of the blind test for 1800 datasets with 512 by 512 pixels, relatively high performance with the accuracy of 0.821, the precision of 0.847, the recall of 0.821, the F1-score of 0.831, and the IoU (Intersection over Union) of 0.723. Although 14.5% of actual cloud shadows were misclassified as land, and 19.7% of actual clouds were misidentified as land, this can be overcome by increasing the quality and quantity of label datasets. Moreover, a state-of-the-art DeepLab V3+ model and the NAS (Neural Architecture Search) optimization technique can help the cloud detection for CAS500 (Compact Advanced Satellite 500) in South Korea.

• Journal of Multimedia Information System
• /
• v.6 no.2
• /
• pp.81-86
• /
• 2019

Deep learning is emerging as one of the best tool in processing data related to medical imaging. In our research work, we have proposed a deep learning based framework CNN (Convolutional Neural Network) for the classification of dysplastic tissue images. The CNN has classified the given images into 4 different classes namely normal tissue, mild dysplastic tissue, moderate dysplastic tissue and severe dysplastic tissue. The dataset under taken for the study consists of 672 tissue images of epithelial squamous layer of oral cavity captured out of the biopsy samples of 52 patients. After applying the data pre-processing and augmentation on the given dataset, 2688 images were created. Further, these 2688 images were classified into 4 categories with the help of expert Oral Pathologist. The classified data was supplied to the convolutional neural network for training and testing of the proposed framework. It has been observed that training data shows 91.65% accuracy whereas the testing data achieves 89.3% accuracy. The results produced by our proposed framework are also tested and validated by comparing the manual results produced by the medical experts working in this area.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.1
• /
• pp.147-165
• /
• 2021

In this work, we empirically evaluated the efficiency of the recent EfficientNetB0 model to identify and diagnose malaria parasite infections in blood smears. The dataset used was collected and classified by relevant experts from the Lister Hill National Centre for Biomedical Communications (LHNCBC). We prepared our samples with minimal image transformations as opposed to others, as we focused more on the feature extraction capability of the EfficientNetB0 baseline model. We applied transfer learning to increase the initial feature sets and reduced the training time to train our model. We then fine-tuned it to work with our proposed layers and re-trained the entire model to learn from our prepared dataset. The highest overall accuracy attained from our evaluated results was 94.70% from fifty epochs and followed by 94.68% within just ten. Additional visualization and analysis using the Gradient-weighted Class Activation Mapping (Grad-CAM) algorithm visualized how effectively our fine-tuned EfficientNetB0 detected infections better than other recent state-of-the-art DCNN models. This study, therefore, concludes that when fine-tuned, the recent EfficientNetB0 will generate highly accurate deep learning solutions for the identification of malaria parasites in blood smears without the need for stringent pre-processing, optimization, or data augmentation of images.

• Journal of Multimedia Information System
• /
• v.9 no.2
• /
• pp.75-86
• /
• 2022

The existing image quality assessment (IQA) datasets have a small number of samples. Some methods based on transfer learning or data augmentation cannot make good use of image quality-related features. A No Reference (NR)-IQA method based on multi-task training and quality awareness is proposed. First, single or multiple distortion types and levels are imposed on the original image, and different strategies are used to augment different types of distortion datasets. With the idea of weak supervision, we use the Full Reference (FR)-IQA methods to obtain the pseudo-score label of the generated image. Then, we combine the classification information of the distortion type, level, and the information of the image quality score. The ResNet50 network is trained in the pre-train stage on the augmented dataset to obtain more quality-aware pre-training weights. Finally, the fine-tuning stage training is performed on the target IQA dataset using the quality-aware weights to predicate the final prediction score. Various experiments designed on the synthetic distortions and authentic distortions datasets (LIVE, CSIQ, TID2013, LIVEC, KonIQ-10K) prove that the proposed method can utilize the image quality-related features better than the method using only single-task training. The extracted quality-aware features improve the accuracy of the model.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.5
• /
• pp.1466-1488
• /
• 2022

Recently, the importance and necessity of artificial intelligence (AI), especially machine learning, has been emphasized. In fact, studies are actively underway to solve complex and challenging problems through the use of AI systems, such as intelligent CCTVs, intelligent AI security systems, and AI surgical robots. Information security that involves analysis and response to security vulnerabilities of software is no exception to this and is recognized as one of the fields wherein significant results are expected when AI is applied. This is because the frequency of malware incidents is gradually increasing, and the available security technologies are limited with regard to the use of software security experts or source code analysis tools. We conducted a study on MalDC, a technique that converts malware into images using machine learning, MalDC showed good performance and was able to analyze and classify different types of malware. MalDC applies a preprocessing step to minimize the noise generated in the image conversion process and employs an image augmentation technique to reinforce the insufficient dataset, thus improving the accuracy of the malware classification. To verify the feasibility of our method, we tested the malware classification technique used by MalDC on a dataset provided by Microsoft and malware data collected by the Korea Internet & Security Agency (KISA). Consequently, an accuracy of 97% was achieved.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.5
• /
• pp.233-240
• /
• 2023

Nutrition information about food is written on the label paper, which is very inconvenient for visually impaired people to recognize. In order to solve the inconvenience of visually impaired people with nutritional information recognition, this paper proposes a product nutrition information guide system for visually impaired people. In the proposed system, user's image data input through UI, and object recognition is carried out through YOLO v5. The proposed system is a system that provides voice guidance on the names and nutrition information of recognized products. This paper constructs a new dataset that augments the 319 classes of canned/late-night snack product image data using rotate matrix techniques, pepper noise, and salt noise techniques. The proposed system compared and analyzed the performance of YOLO v5n, YOLO v5m, and YOLO v5l models through hyperparameter tuning and learned the dataset built with YOLO v5n models. This paper compares and analyzes the performance of the proposed system with that of previous studies.