• Journal of the Korea Convergence Society
• /
• v.12 no.11
• /
• pp.45-51
• /
• 2021

With the development of deep learning, semantic segmentation methods are being studied in various fields. There is a problem that segmenation accuracy drops in fields that require accuracy such as medical image analysis. In this paper, we improved PSPNet, which is a deep learning based segmentation method to minimized the loss of features during semantic segmentation. Conventional deep learning based segmentation methods result in lower resolution and loss of object features during feature extraction and compression. Due to these losses, the edge and the internal information of the object are lost, and there is a problem that the accuracy at the time of object segmentation is lowered. To solve these problems, we improved PSPNet, which is a semantic segmentation model. The multi-scale attention proposed to the conventional PSPNet was added to prevent feature loss of objects. The feature purification process was performed by applying the attention method to the conventional PPM module. By suppressing unnecessary feature information, eadg and texture information was improved. The proposed method trained on the Cityscapes dataset and use the segmentation index MIoU for quantitative evaluation. As a result of the experiment, the segmentation accuracy was improved by about 1.5% compared to the conventional PSPNet.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.40 no.3
• /
• pp.187-193
• /
• 2022

In the Yeongdong region of Gangwon-do, coastal areas are important resources in terms of cultural, social and economic aspects. However, the coast of Gangwon-do is experiencing severe erosion, and it is concerned that its adverse effects will gradually increase. In this study, coastline changes of Yangyang and Gangneung in Gangwon-do were tracked and analyzed over a long period of time. In order to build time series image data, aerial photos from the 1940s to the present were mainly used, and data from CORONA satellite, which operated from the 1960s to the early 1970s, were collected and used together. Using 51cm resolution ortho image and 2m resolution Digital Elevation Model(DEM) as reference, ground control points were selected to perform geometric correction on the aerial photos and CORONA images. Subsequently, Canny edge detector applied to these images to extract the coastlines. As a result of analyzing the extracted and vectorized coastlines by overlaying them in chronological order, erosion and deposition occurring around the artificial structures and on the nearby beaches were observed. In this study, the effect of seasonal variation, tide, and various coastal management including the beach filling were not considered. Because coastal erosion is greatly affected by geographic factors, each local government must find its own solution. Continuous research and local data accumulation are required.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.10
• /
• pp.323-332
• /
• 2022

As various services using AI technology are being developed, much attention is being paid to AI service production. Recently, AI technology is acknowledged as one of ICT services, a lot of research is being conducted for general-purpose AI service production. In this paper, I describe the research results in terms of systems for AI service production, focusing on the distribution and production of machine learning models, which are the final steps of general machine learning development procedures. Three different Ubuntu systems were built, and experiments were conducted on the system, using data from 2017 validation COCO dataset in combination of different AI models (RFCN, SSD-Mobilenet) and different communication methods (gRPC, REST) to request and perform AI services through Tensorflow serving. Through various experiments, it was found that the type of AI model has a greater influence on AI service inference time than AI machine communication method, and in the case of object detection AI service, the number and complexity of objects in the image are more affected than the file size of the image to be detected. In addition, it was confirmed that if the AI service is performed remotely rather than locally, even if it is a machine with good performance, it takes more time to infer the AI service than if it is performed locally. Through the results of this study, it is expected that system design suitable for service goals, AI model development, and efficient AI service production will be possible.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.29 no.6
• /
• pp.536-542
• /
• 2023

Satellites represent cutting-edge technology, of ering significant advantages in spatial and temporal observations. National agencies worldwide harness satellite data to respond to marine accidents and analyze ocean fluctuations effectively. However, challenges arise with high-resolution satellite-based sea surface temperature data (Operational Sea Surface Temperature and Sea Ice Analysis, OSTIA), where gaps or empty areas may occur due to satellite instrumentation, geographical errors, and cloud cover. These issues can take several hours to rectify. This study addressed the issue of missing OSTIA data by employing LaMa, the latest deep learning-based algorithm. We evaluated its performance by comparing it to three existing image processing techniques. The results of this evaluation, using the coefficient of determination (R²) and mean absolute error (MAE) values, demonstrated the superior performance of the LaMa algorithm. It consistently achieved R² values of 0.9 or higher and kept MAE values under 0.5 ℃ or less. This outperformed the traditional methods, including bilinear interpolation, bicubic interpolation, and DeepFill v1 techniques. We plan to evaluate the feasibility of integrating the LaMa technique into an operational satellite data provision system.

• Journal of Korea Multimedia Society
• /
• v.14 no.8
• /
• pp.981-991
• /
• 2011

MIP(Maximum Intensity Projection) is a volume rendering technique which is essential for the medical imaging system. MIP rendering based on the ray casting method produces high quality images but takes a long time. Our aim is improvement of the rendering speed using GPGPU(General-purpose computing on Graphic Process Unit) technique. In this paper, we present the ray casting algorithm based on CUDA(an acronym for Compute Unified Device Architecture) which is a programming language for GPGPU and we suggest new acceleration methods for CUDA. In detail, we propose the block based space leaping which skips unnecessary regions of volume data for CUDA, the bisection method which is a fast method to find a block edge, and the initial value estimation method which improves the probability of space leaping. Due to the proposed methods, we noticeably improve the rendering speed without image quality degradation.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.11
• /
• pp.449-456
• /
• 2019

This paper is a study on the construction of a wall-climbing mobile robot using vacuum suction and wheel-type movement, and a comparison of the performance of an automatic wall crack detection algorithm based on machine learning that is suitable for such an embedded environment. In the embedded system environment, we compared performance by applying recently developed learning methods such as YOLO for object learning, and compared performance with existing edge detection algorithms. Finally, in this study, we selected the optimal machine learning method suitable for the embedded environment and good for extracting the crack features, and compared performance with the existing methods and presented its superiority. In addition, intelligent problem - solving function that transmits the image and location information of the detected crack to the manager device is constructed.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3D
• /
• pp.535-540
• /
• 2006

This paper presented an effective strategy to extract the buildings and to reconstruct 3-D buildings using high-resolution multispectral stereo satellite images. Proposed scheme contained three major steps: building enhancement and segmentation using both BDT (Background Discriminant Transformation) and ISODATA algorithm, conjugate building identification using the object matching with Hausdorff distance and color indexing, and 3-D building reconstruction using photogrammetric techniques. IKONOS multispectral stereo images were used to evaluate the scheme. As a result, the BDT technique was verified as an effective tool for enhancing building areas since BDT suppressed the dominance of background to enhance the building as a non-background. In building recognition, color information itself was not enough to identify the conjugate building pairs since most buildings are composed of similar materials such as concrete. When both Hausdorff distance for edge information and color indexing for color information were combined, most segmented buildings in the stereo images were correctly identified. Finally, 3-D building models were successfully generated using the space intersection by the forward RFM (Rational Function Model).

• Journal of Korea Water Resources Association
• /
• v.38 no.10 s.159
• /
• pp.841-849
• /
• 2005

Two-phase flows, e.g. sediment-laden flow and bubbly flow, have two different flow profiles; flow velocity and sediment velocity. To measure velocity distributions of two-phase flows, it is necessary to use sophisticated instruments which can separate velocity profiles of two-phases. For bubbly flows, PIV (Particle Image Velocimetry) or PTV (Particle Tracking Velocimetry) has given fairly good velocity profiles of two-phases. However, for sediment-laden flows, the applications of PIV or PTV has not been so successful, because the sediment particles introduced to the flow kept the images from being analyzed. A new algorithm, which consists of several image analysis methods, is proposed to analyze sediment-laden flows. For detection algorithm, threshold method, edge detection method, and thinning method are adapted, and for finding matching pair PIV and PTV routines are combined. The proposed method can (1) detect sediment particles with irregular boundaries, (2) remove reflected images and scattered images, and (3) discriminate tracer particles from reflected images of sediment particles.

• Investigative Magnetic Resonance Imaging
• /
• v.13 no.1
• /
• pp.15-21
• /
• 2009

Purpose : A method to estimate a real k-space trajectory based on a circuit model of the gradient system is proposed for spiral imaging. The estimated k-space trajectory instead of the ideal trajectory is used in the reconstruction to improve the image quality in the spiral imaging. Materials and Methods : Since the gradient system has self resistance, capacitance, and inductance, as well as the mutual inductance between the magnet and the gradient coils, the generated gradient fields have delays and transient responses compared to the input waveform to the gradient system. The real gradient fields and their trajectory in k-space play an important role in the reconstruction. In this paper, the gradient system is modeled with R-L-C circuits, and real gradient fields are estimated from the input to the model. An experimental method to determine the model parameters (R, L, C values) is also suggested from the quality of the reconstructed image. Results : The gradient fields are estimated from the circuit model of the gradient system at 1.5 Tesla MRI system. The spiral trajectory obtained by the integration of the estimated gradient fields is used for the reconstruction. From experiments, the reconstructed images using the estimated trajectory show improved uniformity, reduced overshoots near the edges, and enhanced resolutions compared to those using the ideal trajectory without model. Conclusion : The gradient system was successfully modeled by the R-L-C circuits. Much improved reconstruction was achieved in the spiral imaging using the trajectory estimated by the proposed model.

• Geophysics and Geophysical Exploration
• /
• v.24 no.3
• /
• pp.113-130
• /
• 2021

Recently, ground-penetrating radar (GPR) surveys have been actively employed to obtain a large amount of data on occurrences such as ground subsidence and road safety. However, considering the cost and time efficiency, more intuitive and accurate interpretation methods are required, as interpreting a whole survey data set is a cost-intensive process. For this purpose, GPR data can be subjected to attribute analysis, which allows quantitative interpretation. Among the seismic attributes that have been widely used in the field of exploration, complex trace analysis and similarity are the most suitable methods for analyzing GPR data. Further, recently proposed attributes such as edge detecting and texture attributes are also effective for GPR data analysis because of the advances in image processing. In this paper, as a reference for research on the attribute analysis of GPR data, we introduce the useful attributes for GPR data and describe their concepts. Further, we present an analysis of the interpretation methods based on the attribute analysis and past cases.