• Korean Journal of Agricultural and Forest Meteorology
• /
• v.23 no.3
• /
• pp.156-162
• /
• 2021

The number of smart farms has increased to save labor in agricultural production as the subsidy become available from central and local governments. The number of illegal greenhouses has also increased, which causes serious issues for the local governments. In the present study, we developed Mask-RCNN model to detect greenhouses based on satellite images. Greenhouses in the satellite images were labeled for training and validation of the model. The Mask-RC NN model had the average precision (AP) of 75.6%. The average precision values for 50% and 75% of overlapping area were 91.1% and 81.8%, respectively. This results indicated that the Mask-RC NN model would be useful to detect the greenhouses recently built without proper permission using a periodical screening procedure based on satellite images. Furthermore, the model can be connected with GIS to establish unified management system for greenhouses. It can also be applied to the statistical analysis of the number and total area of greenhouses.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.2
• /
• pp.33-42
• /
• 2022

In this paper, we proposed a deep learning model based on multi-scale dilated convolution feature fusion for the segmentation of black ice region in road image to send black ice warning to drivers in real time. In the proposed multi-scale dilated convolution feature fusion network, different dilated ratio convolutions are connected in parallel in the encoder blocks, and different dilated ratios are used in different resolution feature maps, and multi-layer feature information are fused together. The multi-scale dilated convolution feature fusion improves the performance by diversifying and expending the receptive field of the network and by preserving detailed space information and enhancing the effectiveness of diated convolutions. The performance of the proposed network model was gradually improved with the increase of the number of dilated convolution branch. The mIoU value of the proposed method is 96.46%, which was higher than the existing networks such as U-Net, FCN, PSPNet, ENet, LinkNet. The parameter was 1,858K, which was 6 times smaller than the existing LinkNet model. From the experimental results of Jetson Nano, the FPS of the proposed method was 3.63, which can realize segmentation of black ice field in real time.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.25 no.4
• /
• pp.63-80
• /
• 2022

As a technology to analyze and predict an issue has been developed by constructing real space into virtual space, it is becoming more important to acquire precise spatial information in complex cities. In this study, images were acquired using an unmanned aerial vehicle for urban area with complex landscapes, and land cover classification was performed object-based image analysis and semantic segmentation techniques, which were image classification technique suitable for high-resolution imagery. In addition, based on the imagery collected at the same time, the replicability of land cover classification of each artificial intelligence (AI) model was examined for areas that AI model did not learn. When the AI models are trained on the training site, the land cover classification accuracy is analyzed to be 89.3% for OBIA-RF, 85.0% for OBIA-DNN, and 95.3% for U-Net. When the AI models are applied to the replicability assessment site to evaluate replicability, the accuracy of OBIA-RF decreased by 7%, OBIA-DNN by 2.1% and U-Net by 2.3%. It is found that U-Net, which considers both morphological and spectroscopic characteristics, performs well in land cover classification accuracy and replicability evaluation. As precise spatial information becomes important, the results of this study are expected to contribute to urban environment research as a basic data generation method.

• Journal of the Korean Society of Radiology
• /
• v.16 no.1
• /
• pp.25-34
• /
• 2022

Brain computed tomography (CT) is useful for brain lesion diagnosis, such as brain hemorrhage, due to non-invasive methodology, 3-dimensional image provision, low radiation dose. However, there has been numerous misdiagnosis owing to a lack of radiologist and heavy workload. Recently, object detection technologies based on artificial intelligence have been developed in order to overcome the limitations of traditional diagnosis. In this study, the applicability of a deep learning-based YOLOv5s model was evaluated for brain hemorrhage detection using brain CT images. Also, the effect of hyperparameters in the trained YOLOv5s model was analyzed. The YOLOv5s model consisted of backbone, neck and output modules. The trained model was able to detect a region of brain hemorrhage and provide the information of the region. The YOLOv5s model was trained with various activation functions, optimizer functions, loss functions and epochs, and the performance of the trained model was evaluated in terms of brain hemorrhage detection accuracy and training time. The results showed that the trained YOLOv5s model is able to provide a bounding box for a region of brain hemorrhage and the accuracy of the corresponding box. The performance of the YOLOv5s model was improved by using the mish activation function, the stochastic gradient descent (SGD) optimizer function and the completed intersection over union (CIoU) loss function. Also, the accuracy and training time of the YOLOv5s model increased with the number of epochs. Therefore, the YOLOv5s model is suitable for brain hemorrhage detection using brain CT images, and the performance of the model can be maximized by using appropriate hyperparameters.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.27 no.6
• /
• pp.30-38
• /
• 2023

As the number of old buildings subject to safety inspection increases, the burden on designated institutions and management entities that are responsible for safety management is increasing. Accordingly, when selecting buildings subject to safety inspection, appropriate safety inspection standards and appropriate technology are essential. The current safety inspection standards for old buildings give low scores when it is difficult to confirm damage such as cracks in structural members due to finishing materials. This causes the evaluation results to be underestimated regardless of the actual safety status of the structure, resulting in an increase in the number of aging buildings subject to safety inspection. Accordingly, this study proposed a thermal imaging technique, a non-destructive and non-contact inspection, to detect cracks inside finishing materials. A concrete specimen was produced to observe cracks inside the finishing material using a thermal imaging camera, and thermal image data was measured by exciting a heat source on the concrete surface and cracked area. As a result of the measurement, it was confirmed that it was possible to observe cracks inside the finishing material with a width of 0.3mm, 0.5mm, and 0.7mm, but it was difficult to determine the cracks due to uneven temperature distribution due to surface peeling and peeling of the wallpaper. Accordingly, as a result of performing data analysis by deriving the amplitude and phase difference of the thermal image data, clear crack measurement was possible for 0.5mm and 0.7mm cracks. Based on this study, we hope to increase the efficiency of field application and analysis through the development of technology using big data-based deep learning in the diagnosis of internal crack damage in finishing materials.

• Journal of Biomedical Engineering Research
• /
• v.38 no.6
• /
• pp.321-329
• /
• 2017

This paper presents a bone metastasis Detection algorithm on abdominal computed tomography images for early detection using fully convolutional neural networks. The images were taken from patients with various cancers (such as lung cancer, breast cancer, colorectal cancer, etc), and thus the locations of those lesions were varied. To overcome the lack of data, we augmented the data by adjusting the brightness of the images or flipping the images. Before the augmentation, when 70% of the whole data were used in the pre-test, we could obtain the pixel-wise sensitivity of 18.75%, the specificity of 99.97% on the average of test dataset. With the augmentation, we could obtain the sensitivity of 30.65%, the specificity of 99.96%. The increase in sensitivity shows that the augmentation was effective. In the result obtained by using the whole data, the sensitivity of 38.62%, the specificity of 99.94% and the accuracy of 99.81% in the pixel-wise. lesion-wise sensitivity is 88.89% while the false alarm per case is 0.5. The results of this study did not reach the level that could substitute for the clinician. However, it may be helpful for radiologists when it can be used as a screening tool.

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.449-460
• /
• 2021

ASVspoof 2017 deals with detection of replay attacks and aims to classify real human voices and fake voices. The spoofed voice refers to the voice that reproduces the original voice by different types of microphones and speakers. data augmentation research on image data has been actively conducted, and several studies have been conducted to attempt data augmentation on voice. However, there are not many attempts to augment data for voice replay attacks, so this paper explores how audio modification through data augmentation techniques affects the detection of replay attacks. A total of 7 data augmentation techniques were applied, and among them, dynamic value change (DVC) and pitch techniques helped improve performance. DVC and pitch showed an improvement of about 8% of the base model EER, and DVC in particular showed noticeable improvement in accuracy in some environments among 57 replay configurations. The greatest increase was achieved in RC53, and DVC led to an approximately 45% improvement in base model accuracy. The high-end recording and playback devices that were previously difficult to detect were well identified. Based on this study, we found that the DVC and pitch data augmentation techniques are helpful in improving performance in the voice spoofing detection problem.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.2
• /
• pp.63-70
• /
• 2022

Although 2D Object detection has been largely improved in the past years with the advance of deep learning methods and the use of large labeled image datasets, 3D object detection from 2D imagery is a challenging problem in a variety of applications such as robotics, due to the lack of data and diversity of appearances and shapes of objects within a category. Google has just announced the launch of Objectron that has a novel data pipeline using mobile augmented reality session data. However, it also is corresponding to 2D-driven 3D object detection technique. This study explores more mature 2D object detection method, and applies its 2D projection to Objectron 3D lifting system. Most object detection methods use bounding boxes to encode and represent the object shape and location. In this work, we explore a stochastic representation of object regions using Gaussian distributions. We also present a similarity measure for the Gaussian distributions based on the Hellinger Distance, which can be viewed as a stochastic Intersection-over-Union. Our experimental results show that the proposed Gaussian representations are closer to annotated segmentation masks in available datasets. Thus, less accuracy problem that is one of several limitations of Objectron can be relaxed.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1643-1652
• /
• 2022

Deposited Marine Debris(DMD) can negatively affect marine ecosystems, fishery resources, and maritime safety and is mainly detected by sonar sensors, lifting frames, and divers. Considering the limitation of cost and time, recent efforts are being made by integrating underwater images and artificial intelligence (AI). We conducted a comparative study of You Only Look Once Version 5 (YOLOv5) and You Only Look Once Version 7 (YOLOv7) models to detect DMD from underwater images for more accurate and efficient management of DMD. For the detection of the DMD objects such as glass, metal, fish traps, tires, wood, and plastic, the two models showed a performance of over 0.85 in terms of Mean Average Precision (mAP@0.5). A more objective evaluation and an improvement of the models are expected with the construction of an extensive image database.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1663-1676
• /
• 2022

Recently, a number of deep-learning based land cover segmentation studies have been introduced. Some studies denoted that the performance of land cover segmentation deteriorated due to insufficient training data. In this study, we verified the improvement of land cover segmentation performance through data augmentation. U-Net was implemented for the segmentation model. And 2020 satellite-derived landcover dataset was utilized for the study data. The pixel accuracies were 0.905 and 0.923 for U-Net trained by original and augmented data respectively. And the mean F1 scores of those models were 0.720 and 0.775 respectively, indicating the better performance of data augmentation. In addition, F1 scores for building, road, paddy field, upland field, forest, and unclassified area class were 0.770, 0.568, 0.433, 0.455, 0.964, and 0.830 for the U-Net trained by original data. It is verified that data augmentation is effective in that the F1 scores of every class were improved to 0.838, 0.660, 0.791, 0.530, 0.969, and 0.860 respectively. Although, we applied data augmentation without considering class balances, we find that data augmentation can mitigate biased segmentation performance caused by data imbalance problems from the comparisons between the performances of two models. It is expected that this study would help to prove the importance and effectiveness of data augmentation in various image processing fields.