MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising
Remotely sensed images degraded by additive white Gaussian noise (AWGN) have low-level vision, resulting in a poor analysis of their contents. To reduce AWGN, two types of denoising strategies, sparse-coding-model-based and deep-neural-network-based (DNN), are commonly utilized, which have their res...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/2/445 |
| _version_ | 1797437438136680448 |
|---|---|
| author | Zhenghua Huang Zifan Zhu Yaozong Zhang Zhicheng Wang Biyun Xu Jun Liu Shaoyi Li Hao Fang |
| author_facet | Zhenghua Huang Zifan Zhu Yaozong Zhang Zhicheng Wang Biyun Xu Jun Liu Shaoyi Li Hao Fang |
| author_sort | Zhenghua Huang |
| collection | DOAJ |
| description | Remotely sensed images degraded by additive white Gaussian noise (AWGN) have low-level vision, resulting in a poor analysis of their contents. To reduce AWGN, two types of denoising strategies, sparse-coding-model-based and deep-neural-network-based (DNN), are commonly utilized, which have their respective merits and drawbacks. For example, the former pursue enjoyable performance with a high computational burden, while the latter have powerful capacity in completing a specified task efficiently, but this limits their application range. To combine their merits for improving performance efficiently, this paper proposes a model-driven deep denoising (MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula>) scheme. To solve the MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> model, we first decomposed it into several subproblems by the alternating direction method of multipliers (ADMM). Then, the denoising subproblems are replaced by different learnable denoisers, which are plugged into the unfolded MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> model to efficiently produce a stable solution. Both quantitative and qualitative results validate that the proposed MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> approach is effective and efficient, while it has a more powerful ability in generating enjoyable denoising performance and preserving rich textures than other advanced methods. |
| first_indexed | 2024-03-09T11:20:19Z |
| format | Article |
| id | doaj.art-8282daa9c3c24d7da02aad1acd458759 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T11:20:19Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-8282daa9c3c24d7da02aad1acd4587592023-12-01T00:20:56ZengMDPI AGRemote Sensing2072-42922023-01-0115244510.3390/rs15020445MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image DenoisingZhenghua Huang0Zifan Zhu1Yaozong Zhang2Zhicheng Wang3Biyun Xu4Jun Liu5Shaoyi Li6Hao Fang7Artificial Intelligence School, Wuchang University of Technology, Wuhan 430223, ChinaSchool of Electrical and Information Engineering, Wuhan Institute of Technology, Wuhan 430205, ChinaSchool of Electrical and Information Engineering, Wuhan Institute of Technology, Wuhan 430205, ChinaSchool of Electrical and Information Engineering, Wuhan Institute of Technology, Wuhan 430205, ChinaSchool of Electrical and Information Engineering, Wuhan Institute of Technology, Wuhan 430205, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, ChinaCollege of Astronautics, Northwestern Polytechnical University (NWPU), Xi’an 710072, ChinaSchool of Electronic and Information Engineering, Wuhan Donghu University, Wuhan 430212, ChinaRemotely sensed images degraded by additive white Gaussian noise (AWGN) have low-level vision, resulting in a poor analysis of their contents. To reduce AWGN, two types of denoising strategies, sparse-coding-model-based and deep-neural-network-based (DNN), are commonly utilized, which have their respective merits and drawbacks. For example, the former pursue enjoyable performance with a high computational burden, while the latter have powerful capacity in completing a specified task efficiently, but this limits their application range. To combine their merits for improving performance efficiently, this paper proposes a model-driven deep denoising (MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula>) scheme. To solve the MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> model, we first decomposed it into several subproblems by the alternating direction method of multipliers (ADMM). Then, the denoising subproblems are replaced by different learnable denoisers, which are plugged into the unfolded MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> model to efficiently produce a stable solution. Both quantitative and qualitative results validate that the proposed MD<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mrow></mrow><mn>3</mn></msup></semantics></math></inline-formula> approach is effective and efficient, while it has a more powerful ability in generating enjoyable denoising performance and preserving rich textures than other advanced methods.https://www.mdpi.com/2072-4292/15/2/445remotely sensed imagesadditive white Gaussian noise (AWGN)model-driven deep denoising (MD<sup>3</sup>)deep neural network (DNN)alternating direction method of multipliers (ADMM) |
| spellingShingle | Zhenghua Huang Zifan Zhu Yaozong Zhang Zhicheng Wang Biyun Xu Jun Liu Shaoyi Li Hao Fang MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising Remote Sensing remotely sensed images additive white Gaussian noise (AWGN) model-driven deep denoising (MD<sup>3</sup>) deep neural network (DNN) alternating direction method of multipliers (ADMM) |
| title | MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising |
| title_full | MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising |
| title_fullStr | MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising |
| title_full_unstemmed | MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising |
| title_short | MD<sup>3</sup>: Model-Driven Deep Remotely Sensed Image Denoising |
| title_sort | md sup 3 sup model driven deep remotely sensed image denoising |
| topic | remotely sensed images additive white Gaussian noise (AWGN) model-driven deep denoising (MD<sup>3</sup>) deep neural network (DNN) alternating direction method of multipliers (ADMM) |
| url | https://www.mdpi.com/2072-4292/15/2/445 |
| work_keys_str_mv | AT zhenghuahuang mdsup3supmodeldrivendeepremotelysensedimagedenoising AT zifanzhu mdsup3supmodeldrivendeepremotelysensedimagedenoising AT yaozongzhang mdsup3supmodeldrivendeepremotelysensedimagedenoising AT zhichengwang mdsup3supmodeldrivendeepremotelysensedimagedenoising AT biyunxu mdsup3supmodeldrivendeepremotelysensedimagedenoising AT junliu mdsup3supmodeldrivendeepremotelysensedimagedenoising AT shaoyili mdsup3supmodeldrivendeepremotelysensedimagedenoising AT haofang mdsup3supmodeldrivendeepremotelysensedimagedenoising |