KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 5 of 5

Some of the metrics are blocked by your
consent settings
2D and 3D LiDAR with CNN Models for Detecting Sediment Accumulation Underground after Disasters
(2025-01-01)
Krungseanmuang, Woranidtha
;
Morita, Fuka
;
Chaowalittawin, Vasutorn
;
Sathaporn, Posathip
;
Kanamori, Chisato
Global climate change impacts all regions and leads to natural disasters such as typhoons, which cause destruction, debris, and flooding. Postdisaster restoration is a very important activity that is mostly done manually and can be time-consuming and challenging, especially in subterranean environments owing to accumulated objects such as pipes, pillars, and mud distributed in confined underground areas. Therefore, in this study, we aim to utilize emerging AI technologies by comparing deep learning algorithms and evaluating four models for 2D object detection and four for 3D point cloud segmentation for detecting sediment accumulation and navigating around obstacles in underground areas after a disaster. Additionally, a custom dataset was developed to simulate underground disaster scenarios. As a result, the You Only Look Once version 11 (YOLOv11) model achieved the highest mean average precision 50 (mAP50: 91.1%) for general detection within the pillar-pipe dataset, whereas the YOLOv12 model performed the best in detecting pipes (mAP50: 87.7%). In the mud dataset, the YOLOv8 segmentation (YOLOv8-seg) model demonstrated superior performance with mAP50 scores of 93.0% (detection) and 86.4% (segmentation). For 3D point cloud segmentation, PointNet achieved the highest accuracy (98.61%), whereas RandLA-Net was optimal for pipe segmentation, achieving an intersection over union score of 37.1%. These findings highlight AI’s potential to accelerate disaster recovery, reduce manual labor, and ensure faster cleanup. Integrating deep learning models into post-typhoon restoration efforts can enable communities to recover more quickly and efficiently after climate change impacts or disaster events.
Some of the metrics are blocked by your
consent settings
Comparative Analysis of Deep Learning Models for Building Extraction from High-resolution Satellite Imagery
(2025-01-01)
Chueprasert, Tachasit
;
Udomchaiporn, Akadej
;
Intagosum, Sarun
In this research, an approach to extract buildings from Google's satellite imagery was proposed. The performances of various deep learning models (U-Net, RIU-Net, U-Net++, Res-U-Net, and DeepLabV3+) on pre-processed datasets were compared. The models were trained using the similarity metrics of Intersection over Union (IoU) and Dice Similarity Coefficient (DSC). The best-performing models among the segmentation techniques were Res-U-Net and DeepLabV3+. Res-U-Net, an enhanced version of the traditional U-Net model that incorporates residual connections for improved feature propagation, achieved an F1 score of 85.43% when using the RGB dataset. Similarly, DeepLabV3+ also achieved high performance on the Enhanced RGB dataset, obtaining an F1 score of 85.18% after applying pre-processing techniques. This research highlights the significance of color as a dominant feature for accurate building extraction from satellite images. The findings contribute to improved methodologies for building identification, benefiting urban planning, and disaster management applications.
Some of the metrics are blocked by your
consent settings
PCL-PTD Net: Parallel Cross-Learning-Based Pixel Transferred Deconvolutional Network for Building Extraction in Dense Building Areas With Shadow
(2023-01-01)
Boonpook, Wuttichai
;
Tan, Yumin
;
Torsri, Kritanai
;
Kamsing, Patcharin
;
Torteeka, Peerapong
Urban building segmentation from remote sensed imageries is challenging because there usually exists a variety of building features. Furthermore, very high spatial resolution imagery can provide many details of the urban building, such as styles, small gaps among buildings, building shadows, etc. Hence, satisfactory accuracy in detecting and extracting urban features from highly detailed images still remains. Deep learning semantic segmentation using baseline networks works well on building extraction; however, their ability in building extraction in shadows area, unclear building feature, and narrow gaps among buildings in dense building zone is still limited. In this article, we propose parallel cross-learning-based pixel transferred deconvolutional network (PCL-PTD net), and then is used to segment urban buildings from aerial photographs. The proposed method is evaluated and intercompared with traditional baseline networks. In PCL-PTD net, it is composed of parallel network, cross-learning functions, residual unit in encoder part, and PTD in decoder part. The performance is applied to three datasets (Inria aerial dataset, international society for photogrammetry and remote sensing Potsdam dataset, and UAV building dataset), to evaluate its accuracy and robustness. As a result, we found that PCL-PTD net can improve learning capacities of the supervised learning model in differentiating buildings in dense area and extracting buildings covered by shadows. As compared to the baseline networks, we found that proposed network shows superior performance compared to all eight networks (SegNet, U-net, pyramid scene parsing network, PixelDCL, DeeplabV3+, U-Net++, context feature enhancement networ, and improved ResU-Net). The experiments on three datasets also demonstrate the ability of proposed framework and indicating its performance.
Some of the metrics are blocked by your
consent settings
The Study of Sementic Deep Learning Segmentation for Durian Orchard
(2023-01-01)
Boksuwan, Sungwan
The paper comparatively studies a deep learning based semantic segmentation for segmenting durian orchard environments using MATLAB platform. Experiments consist of four treatments that are the combinations of Deeplabv3+ with base networks including Resnet-18, Resnet-50, Xception and Interceptionresnetv2. IoU metric is utilized as the performance index. The environment is segmented into five classes. The experimental results tested by ANOVA reveal that base networks do not result in a different performance for the class of sky, tree, grass, and road but show different performance for background class.
Some of the metrics are blocked by your
consent settings
Semantic Bird'View Map Prediction Using Horizontally-Aware Pyramid Occupancy Network
(2023-01-01)
Teerarattanyu, Thanapat
;
Wongchai, Tunlaton
;
Pavarangkoon, Praphan
;
Dilokthanakul, Nat
Deep neural network has been used to predict the bird's-eye-view map from a frontal camera of an autonomous car. A state-of-the-art approach, namely pyramid occupancy network (PON), uses an encoder-decoder architecture to condense an image column into a context vector that describes the object occupancy along the radial direction. Our work, Horizontally-aware Pyramid Occupancy Network (H-PON), extends the PON model with a novel component that provides additional context information describing the relationships of the objects along the horizontal direction. This is done by also encoding the horizontal column of the image into an additional context vector using another encoder-decoder layer. This context vector is, then, expanded back providing improved features for semantic reasoning across the horizontal direction. We found that this simple extension significantly improves PON's semantic prediction performance in the nuScences dataset. Our experiment shows that the objects that are rarely seen and those that are further away from the center greatly benefit from this novel component.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results