• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.3
• /
• pp.1434-1449
• /
• 2019

Video object segmentation is a significant task in computer vision, but its performance is not very satisfactory. A method of video object segmentation using weakly temporal information is presented in this paper. Motivated by the phenomenon in reality that the motion of the object is a continuous and smooth process and the appearance of the object does not change much between adjacent frames in the video sequences, we use a feed-forward architecture with motion estimation to predict the mask of the current frame. We extend an additional mask channel for the previous frame segmentation result. The mask of the previous frame is treated as the input of the expanded channel after processing, and then we extract the temporal feature of the object and fuse it with other feature maps to generate the final mask. In addition, we introduce multi-mask guidance to improve the stability of the model. Moreover, we enhance segmentation performance by further training with the masks already obtained. Experiments show that our method achieves competitive results on DAVIS-2016 on single object segmentation compared to some state-of-the-art algorithms.

• Korean Journal of Remote Sensing
• /
• v.33 no.4
• /
• pp.423-436
• /
• 2017

This paper presents a deep learning-based road segmentation framework from very high-resolution orthophotos. The proposed method uses Deep Convolutional Autoencoders for end-to-end mapping of orthophotos to road segmentations. In addition, a set of post-processing steps were applied to make the model outputs GIS-ready data that could be useful for various applications. The optimization of the model's parameters is explained which was conducted via grid search method. The model was trained and implemented in Keras, a high-level deep learning framework run on top of Tensorflow. The results show that the proposed model with the best-obtained hyperparameters could segment road objects from orthophotos at an average accuracy of 88.5%. The results of optimization revealed that the best optimization algorithm and activation function for the studied task are Stochastic Gradient Descent (SGD) and Exponential Linear Unit (ELU), respectively. In addition, the best numbers of convolutional filters were found to be 8 for the first and second layers and 128 for the third and fourth layers of the proposed network architecture. Moreover, the analysis on the time complexity of the model showed that the model could be trained in 4 hours and 50 minutes on 1024 high-resolution images of size $106{\times}106pixels$, and segment road objects from similar size and resolution images in around 14 minutes. The results show that the deep learning models such as Convolutional Autoencoders could be a best alternative to traditional machine learning models for road segmentation from aerial photographs.

• The Journal of the Korea Contents Association
• /
• v.9 no.6
• /
• pp.266-274
• /
• 2009

An abdominal aortic aneurysm occurs most commonly in older individuals (between 65 and 75), and more in men and smokers. The most important complication of an abdominal aortic aneurysm is rupture, which is most often a fatal event. An abdominal aortic aneurysm weakens the walls of the blood vessel, leaving it vulnerable to bursting open, or rupturing, and spilling large amounts of blood into the abdominal cavity. surface modeling is very useful to surgery for quantitative analysis of abdominal aortic aneurysm. the 3D representation and surface modeling an abdominal aortic aneurysm structure taken from Multi Detector Computed Tomography. The construction of the 3D model is generally carried out by staking the contours obtained from 2D segmentation of each CT slice, so the quality of the 3D model strongly defends on the precision of segmentation process. In this work we present deformable model algorithm. deformable model is an energy-minimizing spline guided by external constraint force. External force which we call Gradient Vector Flow, is computed as a diffusion of a gradient vectors of gray level or binary edge map derived from the image. Finally, we have used snakes successfully for abdominal aortic aneurysm segmentation the performance of snake was visually and quantitatively validated by experts.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.27 no.1
• /
• pp.30-36
• /
• 2023

In this study, an image segmentation model that can evaluate surface porosity based on concrete surface images was derived. Three types of concrete specimens with different water-cement ratios (w/c = 54, 35, and 30%) were prepared, and 2,729 surface images were obtained using an optical microscope. Benchmarking tests, parameter optimization, and final model derivation were performed using the surface images, and an image segmentation model with 97% verification accuracy was obtained. The model was verified by comparing the porosity obtained from the model and X-Ray Microscope (XRM). The model provided similar porosity to that of XRM for the specimens with a high water-cement ratio, but tended to give lower porosity for specimens with a low water-cement ratio.

• Korean Journal of Remote Sensing
• /
• v.40 no.1
• /
• pp.71-80
• /
• 2024

In this study, a proposed ensemble learning technique aims to enhance the semantic segmentation performance of images captured by Unmanned Aerial Vehicles (UAVs). With the increasing use of UAVs in fields such as urban planning, there has been active development of techniques utilizing deep learning segmentation methods for land cover segmentation. The study suggests a method that utilizes prominent segmentation models, namely U-Net, DeepLabV3, and Fully Convolutional Network (FCN), to improve segmentation prediction performance. The proposed approach integrates training loss, validation accuracy, and class score of the three segmentation models to enhance overall prediction performance. The method was applied and evaluated on a land cover segmentation problem involving seven classes: buildings,roads, parking lots, fields, trees, empty spaces, and areas with unspecified labels, using images captured by UAVs. The performance of the ensemble model was evaluated by mean Intersection over Union (mIoU), and the results of comparing the proposed ensemble model with the three existing segmentation methods showed that mIoU performance was improved. Consequently, the study confirms that the proposed technique can enhance the performance of semantic segmentation models.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2010.04a
• /
• pp.67-68
• /
• 2010

Segmentation of LiDAR data is an important procedure in building modeling. Therefore, in this study, aerial imagery is used to group LiDAR data for both improving segmentation accuracy and modeling detail surface patches of the roofs. The results show that the proposed method is efficient to analyze and to model various types of roof shape.

• Smart Structures and Systems
• /
• v.30 no.5
• /
• pp.501-511
• /
• 2022

Three-dimensional (3D) models have become crucial for improving civil infrastructure analysis, and they can be used for various purposes such as damage detection, risk estimation, resolving potential safety issues, alarm detection, and structural health monitoring. 3D point cloud data is used not only to make visual models but also to analyze the states of structures and to monitor them using semantic data. This study proposes automating the generation of high-quality 3D point cloud data and removing noise using deep learning algorithms. In this study, large-format aerial images of civilian infrastructure, such as cut slopes and dams, which were captured by drones, were used to develop a workflow for automatically generating a 3D point cloud model. Through image cropping, downscaling/upscaling, semantic segmentation, generation of segmentation masks, and implementation of region extraction algorithms, the generation of the point cloud was automated. Compared with the method wherein the point cloud model is generated from raw images, our method could effectively improve the quality of the model, remove noise, and reduce the processing time. The results showed that the size of the 3D point cloud model created using the proposed method was significantly reduced; the number of points was reduced by 20-50%, and distant points were recognized as noise. This method can be applied to the automatic generation of high-quality 3D point cloud models of civil infrastructures using aerial imagery.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.9 no.2
• /
• pp.196-204
• /
• 2016

World Health Organization reported that heart-related diseases such as coronary artery stenoses show the highest occurrence rate which may cause heart attack. Using Computed Tomography angiography images will allow radiologists to detect and have intervention by creating 3D roadmapping of the vessels. However, it is often complex and difficult do reconstruct 3D vessel which causes very large amount of time and previous researches were studied to segment vessels more accurate automatically. Therefore, in this paper, Region Competition, Geodesic Active Contour (GAC), Multi-atlas based segmentation and Active Shape Model algorithms were applied to segment aortic root from CTA images and the results were analyzed by using mean Hausdorff distance, volume to volume measure, computational time, user-interaction and coronary ostium detection rate. As a result, Extracted 3D aortic model using GAC showed the highest accuracy but also showed highest user-interaction results. Therefore, it is important to improve automatic segmentation algorithm in future

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.9
• /
• pp.37-47
• /
• 2021

In this paper, we propose a new deep learning model for tongue segmentation with improved accuracy compared to the existing model by diversifying the receptive field in the U-net. Methods such as parallel convolution, dilated convolution, and constant channel increase were used to diversify the receptive field. For the proposed deep learning model, a tongue region segmentation experiment was performed on two test datasets. The training image and the test image are similar in TestSet1 and they are not in TestSet2. Experimental results show that segmentation performance improved as the receptive field was diversified. The mIoU value of the proposed method was 98.14% for TestSet1 and 91.90% for TestSet2 which was higher than the result of existing models such as U-net, DeepTongue, and TongueNet.

• Journal of the Korean Society of Radiology
• /
• v.18 no.3
• /
• pp.231-237
• /
• 2024

Brain injury causes persistent disability in survivors, and epidural hematoma(EDH) and subdural hematoma (SDH) resulting from cerebral hemorrhage can be considered one of the major clinical diseases. In this study, we attempted to automatically segment and hematomas due to cerebral hemorrhage in three dimensions based on computed tomography(CT) images. An improved GVF(gradient vector flow) algorithm was implemented for automatic segmentation of hematoma. After calculating and repeating the gradient vector from the image, automatic segmentation was performed and a 3D model was created using the segmentation coordinates. As a result of the experiment, accurate segmentation of the boundaries of the hematoma was successful. The results were found to be good even in border areas and thin hematoma areas, and the intensity, direction of spread, and area of the hematoma could be known in various directions through the 3D model. It is believed that the planar information and 3D model of the cerebral hemorrhage area developed in this study can be used as auxiliary diagnostic data for medical staff.