Research on the deep learning-based exposure invariant spectral reconstruction method
The surface spectral reflectance of an object is the key factor for high-fidelity color reproduction and material analysis, and spectral acquisition is the basis of its applications. Based on the theoretical imaging model of a digital camera, the spectral reflectance of any pixels in the image can b...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-10-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2022.1031546/full |
| _version_ | 1798029800194965504 |
|---|---|
| author | Jinxing Liang Jinxing Liang Jinxing Liang Lei Xin Zhuan Zuo Jing Zhou Anping Liu Hang Luo Xinrong Hu |
| author_facet | Jinxing Liang Jinxing Liang Jinxing Liang Lei Xin Zhuan Zuo Jing Zhou Anping Liu Hang Luo Xinrong Hu |
| author_sort | Jinxing Liang |
| collection | DOAJ |
| description | The surface spectral reflectance of an object is the key factor for high-fidelity color reproduction and material analysis, and spectral acquisition is the basis of its applications. Based on the theoretical imaging model of a digital camera, the spectral reflectance of any pixels in the image can be obtained through spectral reconstruction technology. This technology can avoid the application limitations of spectral cameras in open scenarios and obtain high spatial resolution multispectral images. However, the current spectral reconstruction algorithms are sensitive to the exposure variant of the test images. That is, when the exposure of the test image is different from that of the training image, the reconstructed spectral curve of the test object will deviate from the real spectral to varying degrees, which will lead to the spectral data of the target object being accurately reconstructed. This article proposes an optimized method for spectral reconstruction based on data augmentation and attention mechanisms using the current deep learning-based spectral reconstruction framework. The proposed method is exposure invariant and will adapt to the open environment in which the light is easily changed and the illumination is non-uniform. Thus, the robustness and reconstruction accuracy of the spectral reconstruction model in practical applications are improved. The experiments show that the proposed method can accurately reconstruct the shape of the spectral reflectance curve of the test object under different test exposure levels. And the spectral reconstruction error of our method at different exposure levels is significantly lower than that of the existing methods, which verifies the proposed method’s effectiveness and superiority. |
| first_indexed | 2024-04-11T19:30:11Z |
| format | Article |
| id | doaj.art-cb791f52148e4a46b1539ee705e19b12 |
| institution | Directory Open Access Journal |
| issn | 1662-453X |
| language | English |
| last_indexed | 2024-04-11T19:30:11Z |
| publishDate | 2022-10-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroscience |
| spelling | doaj.art-cb791f52148e4a46b1539ee705e19b122022-12-22T04:07:01ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2022-10-011610.3389/fnins.2022.10315461031546Research on the deep learning-based exposure invariant spectral reconstruction methodJinxing Liang0Jinxing Liang1Jinxing Liang2Lei Xin3Zhuan Zuo4Jing Zhou5Anping Liu6Hang Luo7Xinrong Hu8School of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaHubei Provincial Engineering Research Center for Intelligent Textile and Fashion, Wuhan, Hubei, ChinaEngineering Research Center of Hubei Province for Clothing Information, Wuhan, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan Textile University, Wuhan, Hubei, ChinaThe surface spectral reflectance of an object is the key factor for high-fidelity color reproduction and material analysis, and spectral acquisition is the basis of its applications. Based on the theoretical imaging model of a digital camera, the spectral reflectance of any pixels in the image can be obtained through spectral reconstruction technology. This technology can avoid the application limitations of spectral cameras in open scenarios and obtain high spatial resolution multispectral images. However, the current spectral reconstruction algorithms are sensitive to the exposure variant of the test images. That is, when the exposure of the test image is different from that of the training image, the reconstructed spectral curve of the test object will deviate from the real spectral to varying degrees, which will lead to the spectral data of the target object being accurately reconstructed. This article proposes an optimized method for spectral reconstruction based on data augmentation and attention mechanisms using the current deep learning-based spectral reconstruction framework. The proposed method is exposure invariant and will adapt to the open environment in which the light is easily changed and the illumination is non-uniform. Thus, the robustness and reconstruction accuracy of the spectral reconstruction model in practical applications are improved. The experiments show that the proposed method can accurately reconstruct the shape of the spectral reflectance curve of the test object under different test exposure levels. And the spectral reconstruction error of our method at different exposure levels is significantly lower than that of the existing methods, which verifies the proposed method’s effectiveness and superiority.https://www.frontiersin.org/articles/10.3389/fnins.2022.1031546/fullspectral reconstructionmultispectral imagecolor scienceconvolutional neural networkexposure invariantdense connections |
| spellingShingle | Jinxing Liang Jinxing Liang Jinxing Liang Lei Xin Zhuan Zuo Jing Zhou Anping Liu Hang Luo Xinrong Hu Research on the deep learning-based exposure invariant spectral reconstruction method Frontiers in Neuroscience spectral reconstruction multispectral image color science convolutional neural network exposure invariant dense connections |
| title | Research on the deep learning-based exposure invariant spectral reconstruction method |
| title_full | Research on the deep learning-based exposure invariant spectral reconstruction method |
| title_fullStr | Research on the deep learning-based exposure invariant spectral reconstruction method |
| title_full_unstemmed | Research on the deep learning-based exposure invariant spectral reconstruction method |
| title_short | Research on the deep learning-based exposure invariant spectral reconstruction method |
| title_sort | research on the deep learning based exposure invariant spectral reconstruction method |
| topic | spectral reconstruction multispectral image color science convolutional neural network exposure invariant dense connections |
| url | https://www.frontiersin.org/articles/10.3389/fnins.2022.1031546/full |
| work_keys_str_mv | AT jinxingliang researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT jinxingliang researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT jinxingliang researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT leixin researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT zhuanzuo researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT jingzhou researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT anpingliu researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT hangluo researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod AT xinronghu researchonthedeeplearningbasedexposureinvariantspectralreconstructionmethod |