Significance of Atrous Spatial Pyramid Pooling (ASPP) in Deeplabv3+ for Water Body Segmentation

Sunandini, Gosula; Sivanpillai, Ramesh; Sowmya, V; Sajith Variyar, V. V

doi:10.1109/spin57001.2023.10116882

Search citation statements

Order By: Relevance

Paper Sections

Select...

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2023

2024

Publication Types

Select...

Article3

Other1

Relationship

Self Cite0

Independent4

Authors

Journals

Cited by 4 publications

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Cafct-Net: A Cnn-Transformer Hybrid Network With Contextual And Attentional Feature Fusion For Liver Tumor Segmentation

Kang,

Ting,

Ting

et al. 2024

2024 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Cafct-Net: A Cnn-Transformer Hybrid Network With Contextual And Attentional Feature Fusion For Liver Tumor Segmentation

Kang,

Ting,

Ting

et al. 2024

2024 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

SCAFNet: Semantic-Guided Cascade Adaptive Fusion Network for Infrared Small Target Detection

Zhang,

Wang,

Xing

et al. 2024

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Water Bodies Extraction Using Mathematical Morphology

BENALI

2024

Kosm. nauka tehnol.

View full text Add to dashboard Cite

The management of water resources is vital for maintaining the world’s ecosystems. Conventional methods of extracting water bodies remain very limited due to the complexity of the implementation. This leads to a reduction in the extraction precision. Our main objec- tive is to improve the detection of water bodies. We tested the accuracy of our method on the Sentinel-2 Dataset that contains images with different complexity levels and heterogeneous structures like shadows, roads, buildings, etc. This article presents an original method that implements the idea of separating the three-component RGB image matrices and then processing only the green matrix because it contains all water bodies with high precision. Our method is based mainly on the mathematical morphology. Firstly, we propose a simple and fast binary algorithm to detect the maximum of water bodies existing in the images. This step was carried out using the Hit-or-Miss Transform. The second step exploits applying the Top-Hat Transform to refine the segmentation result. By comparing our method with several currently used methods, we notice that our method improves the quality of segmentation and gives excellent results, which exceed 95% for all the metrics used to calculate the classification quality in the purview of remote sensing. The error obtained with our method remains less than 1 %. We can affirm that our method is very suitable for detecting bodies of water compared to all current methods.

show abstract

Significance of Atrous Spatial Pyramid Pooling (ASPP) in Deeplabv3+ for Water Body Segmentation

Cited by 4 publications

References 8 publications

Cafct-Net: A Cnn-Transformer Hybrid Network With Contextual And Attentional Feature Fusion For Liver Tumor Segmentation

Cafct-Net: A Cnn-Transformer Hybrid Network With Contextual And Attentional Feature Fusion For Liver Tumor Segmentation

SCAFNet: Semantic-Guided Cascade Adaptive Fusion Network for Infrared Small Target Detection

Water Bodies Extraction Using Mathematical Morphology

Contact Info

Product

Resources

About