HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK
In the past decade, significant progress has been made in hyperspectral image (HSI) classification due to advances in deep learning techniques. This study introduces a novel computational framework for classifying HSIs using a multiscale 3D convolutional neural network (3DCNN). The purpose of this r...
সংরক্ষণ করুন:
| প্রধান লেখক: | , , |
|---|---|
| বিন্যাস: | Journal article |
| ভাষা: | English |
| প্রকাশিত: |
2025
|
| বিষয়গুলি: | |
| অনলাইন ব্যবহার করুন: | https://scholar.dlu.edu.vn/handle/123456789/5801 https://tckh.dlu.edu.vn/index.php/tckhdhdl/article/view/1381 |
| ট্যাগগুলো: |
ট্যাগ যুক্ত করুন
কোনো ট্যাগ নেই, প্রথমজন হিসাবে ট্যাগ করুন!
|
| Thư viện lưu trữ: | Thư viện Trường Đại học Đà Lạt |
|---|
| id |
oai:scholar.dlu.edu.vn:123456789-5801 |
|---|---|
| record_format |
dspace |
| spelling |
oai:scholar.dlu.edu.vn:123456789-58012025-10-03T03:51:08Z HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK Thai Duy Quy Tran, Thi Ngoc Thao Avirmed, Enkhbat 2DCNN 3DCNN Deep learning HSI datasets Hyperspectral image Recurrent neural networks Temporal convolutional networks. In the past decade, significant progress has been made in hyperspectral image (HSI) classification due to advances in deep learning techniques. This study introduces a novel computational framework for classifying HSIs using a multiscale 3D convolutional neural network (3DCNN). The purpose of this research is to enhance the efficiency of HSI classification by leveraging the spatial-spectral characteristics of HSI data. The methodology involves preprocessing HSI datasets, extracting features using parallel 3DCNN layers with skip connections, and evaluating performance on five benchmark datasets. Key findings demonstrate that the proposed model significantly improves classification accuracy, achieving average accuracies of 85%, 84%, 94%, 98%, and 95% on the Kennedy Space Center (KSC), Indian Pines (IP), Pavia University (PU), Pavia Centre (PC), and Botswana (BW) datasets, respectively. The originality of this study lies in the multiscale parallel design of the 3DCNN architecture, which balances computational complexity with enhanced performance. Limitations include the high computational cost associated with training deep networks, which can be mitigated by graphics processing unit (GPU) acceleration and optimized implementations. 15 3 171-185 Khoa Công nghệ Thông tin 3 Thái Duy Quý Tạp chí Khoa học Đại học Đà Lạt 2025-10-02T07:37:22Z 2025-10-02T07:37:22Z 2025-09-29 Journal article Bài báo đăng trên tạp chí trong nước (có ISSN), bao gồm book chapter https://scholar.dlu.edu.vn/handle/123456789/5801 10.37569/DalatUniversity.15.3.1381(2025) https://tckh.dlu.edu.vn/index.php/tckhdhdl/article/view/1381 en Tạp chí Khoa học Đại học Đà Lạt |
| institution |
Thư viện Trường Đại học Đà Lạt |
| collection |
Thư viện số |
| language |
English |
| topic |
2DCNN 3DCNN Deep learning HSI datasets Hyperspectral image Recurrent neural networks Temporal convolutional networks. |
| spellingShingle |
2DCNN 3DCNN Deep learning HSI datasets Hyperspectral image Recurrent neural networks Temporal convolutional networks. Thai Duy Quy Tran, Thi Ngoc Thao Avirmed, Enkhbat HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| description |
In the past decade, significant progress has been made in hyperspectral image (HSI) classification due to advances in deep learning techniques. This study introduces a novel computational framework for classifying HSIs using a multiscale 3D convolutional neural network (3DCNN). The purpose of this research is to enhance the efficiency of HSI classification by leveraging the spatial-spectral characteristics of HSI data. The methodology involves preprocessing HSI datasets, extracting features using parallel 3DCNN layers with skip connections, and evaluating performance on five benchmark datasets. Key findings demonstrate that the proposed model significantly improves classification accuracy, achieving average accuracies of 85%, 84%, 94%, 98%, and 95% on the Kennedy Space Center (KSC), Indian Pines (IP), Pavia University (PU), Pavia Centre (PC), and Botswana (BW) datasets, respectively. The originality of this study lies in the multiscale parallel design of the 3DCNN architecture, which balances computational complexity with enhanced performance. Limitations include the high computational cost associated with training deep networks, which can be mitigated by graphics processing unit (GPU) acceleration and optimized implementations. |
| format |
Journal article |
| author |
Thai Duy Quy Tran, Thi Ngoc Thao Avirmed, Enkhbat |
| author_facet |
Thai Duy Quy Tran, Thi Ngoc Thao Avirmed, Enkhbat |
| author_sort |
Thai Duy Quy |
| title |
HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| title_short |
HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| title_full |
HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| title_fullStr |
HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| title_full_unstemmed |
HYPERSPECTRAL IMAGE CLASSIFICATION ON A MULTISCALE 3D CONVOLUTIONAL NEURAL NETWORK |
| title_sort |
hyperspectral image classification on a multiscale 3d convolutional neural network |
| publishDate |
2025 |
| url |
https://scholar.dlu.edu.vn/handle/123456789/5801 https://tckh.dlu.edu.vn/index.php/tckhdhdl/article/view/1381 |
| _version_ |
1845408772029153280 |