|
|
- Impact of water vapour on polymer classification using in situ short-wave
infrared hyperspectral imaging
Authors: Muhammad Saad Shaikh, Benny Thörnberg
Pages: 1 - 12
Abstract: Hyperspectral remote sensing is known to suffer from wavelength bands blocked by atmospheric gases. Short-wave infrared hyperspectral imaging at in situ installations is shown to be affected by water vapour even if the pathlength of light through air is only hundreds of centimetres. This impact is especially noticeable with large variations of relative humidity, the coefficient of variation reaching 5 % in our test case. Using repeated calibrations of imaging system at the same relative humidity as in the measurement, we were able to reduce the coefficient of variation to 1 %. The measurement variations are also shown to induce significant error in material classification. Polymer type identification was selected as the test case for material classification. The measurement variations due to the change in relative humidity are shown to result in 20 % classification error at its minimum. With repeated calibrations or by eliminating the most affected wavelength bands from measurements, we were able to reduce the classification error to less than 1 %. Such improvement of measurement and classification precision may be important for industrial applications such as waste sorting, polymer classification etc.
Citation: J. Spectral Imaging 11, a5 (2022)
PubDate: 2022-06-01
DOI: 10.1255/jsi.2022.a5
Issue No: Vol. 11 (2022)
- Hyperspectral image non-linear unmixing using joint extrinsic and
intrinsic priors with L1/2-norms to non-negative matrix factorisation
Authors: K. Priya, K. K. Rajkumar
Pages: 1 - 19
Abstract: Hyperspectral unmixing (HU) is one of the most active emerging areas in image processing that estimates the hyperspectral image’s endmember and abundance. HU enhances the quality of both spectral and spatial dimensions of the image by modifying the endmember and abundance parameters of the hyperspectral images. There are several HU algorithms available in the literature based on the linear mixing model (LMM) that deals with the microscopic contents of the pixels in the images. Non-negative matrix factorisation (NMF) is the prominent method widely used in LMMs that simultaneously estimates both the endmembers and abundances parameters along with some residual factors of the image to improve the quality of unmixing. In addition to this, the quality of the image is enhanced by incorporating some constraints to both endmember and abundance matrices with the NMF method. However, all the existing methods apply any of these constraints to the endmember and abundance matrices by considering the linearity features of the images. In this paper, we propose an unmixing model called joint extrinsic and intrinsic priors with L1/2 norms to non-negative matrix factorisation (JEIp L1/2-NMF) that applies multiple constraints simultaneously to both endmember and abundance matrices of the hyperspectral image to enhance its quality. Three main external and internal constraints such as minimum volume, sparsity and total variation are applied to both the endmembers and abundance parameters of the image. In addition, a L1/2-norms is imposed to extract good quality spectral data. Therefore, the proposed method enhances spatial as well as spectral data and considers the non-linearity of the pixels in the image by adding a residual term to the model. Performance of our proposed model is measured by using different quality measuring indexes on four benchmark public datasets and found that the proposed method shows outstanding performance compared to all the conventional baseline methods. Further, we also evaluated the performance of our method by varying the number of endmembers empirically and concluded that less than five endmembers provides high-quality spectral and spatial data during the unmixing process.
Citation: J. Spectral Imaging 11, a4 (2022)
PubDate: 2022-04-07
DOI: 10.1255/jsi.2022.a4
Issue No: Vol. 11 (2022)
- Data processing of three-dimensional vibrational spectroscopic chemical
images for pharmaceutical applications
Authors: Hannah Carruthers, Don Clark, Fiona C. Clarke, Karen Faulds, Duncan Graham
Pages: 1 - 8
Abstract: Vibrational spectroscopic chemical imaging is a powerful tool in the pharmaceutical industry to assess the spatial distribution of components within pharmaceutical samples. Recently, the combination of vibrational spectroscopic chemical mapping with serial sectioning has provided a means to visualise the three-dimensional (3D) structure of a tablet matrix. There are recognised knowledge gaps in current tablet manufacturing processes, particularly regarding the size, shape and distribution of components within the final drug product. The performance of pharmaceutical tablets is known to be primarily influenced by the physical and chemical properties of the formulation. Here, we describe the data processing methods required to extract quantitative domain size and spatial distribution statistics from 3D vibrational spectroscopic chemical images. This provides a means to quantitatively describe the microstructure of a tablet matrix and is a powerful tool to overcome knowledge gaps in current tablet manufacturing processes, optimising formulation development.
Citation: J. Spectral Imaging 11, a3 (2022)
PubDate: 2022-03-30
DOI: 10.1255/jsi.2022.a3
Issue No: Vol. 11 (2022)
- A semi-supervised cycle-GAN neural network for hyperspectral image
classification with minimum noise fraction
Authors: Tatireddy Subba Reddy, Jonnadula Harikiran
Pages: 1 - 14
Abstract: Hyperspectral imaging (HSI) is a popular mode of remote sensing imaging that collects data beyond the visible spectrum. Many classification techniques have been developed in recent years, since classification is the most crucial task in hyperspectral image processing. Furthermore, extracting features from hyperspectral images is challenging in many scenarios. The semi-supervised classification of HSI is motivated by the Cycle-GAN method that has been proposed in this research paper. Since the proposed HSI classification method is semi-supervised, it makes extensive use of the labelled samples, which are short and have numerous unlabelled images. The research is carried out in two phases. First, to extract the spectral–spatial features, the minimum noise fraction is adopted. And, second, the classification of the semi-supervised method is done by the cycle-GANs. Subsequently, the proposed architecture is implemented on three standard hyperspectral dataset methods. As a result, the performance comparison is carried out in the same field as state-of-the-art approaches. The obtained results successfully demonstrate the supremacy of the proposed technique in the classification of HSI.
Citation: J. Spectral Imaging 11, a2 (2022)
PubDate: 2022-03-29
DOI: 10.1255/jsi.2022.a2
Issue No: Vol. 11 (2022)
- An outlook: machine learning in hyperspectral image classification and
dimensionality reduction techniques
Authors: Jonnadula Harikiran, Tatireddy Subba Reddy
Pages: 1 - 17
Abstract: Hyperspectral imaging is used in a wide range of applications. When used in remote sensing, satellites and aircraft are employed to collect the images, which are used in agriculture, environmental monitoring, urban planning and defence. The exact classification of ground features in the images is a significant research issue and is currently receiving greater attention. Moreover, these images have a large spectral dimensionality, which adds computational complexity and affects classification precision. To handle these issues, dimensionality reduction is an essential step that improves the performance of classifiers. In the classification process, several strategies have produced good classification results. Of these, machine learning techniques are the most powerful approaches. As a result, this paper reviews three different types of hyperspectral image machine learning classification methods: cluster analysis, supervised and semi-supervised classification. Moreover, this paper shows the effectiveness of all these techniques for hyperspectral image classification and dimensionality reduction. Furthermore, this review will assist as a reference for future research to improve the classification and dimensionality reduction approaches.
Citation: J. Spectral Imaging 11, a1 (2022)
PubDate: 2022-01-07
DOI: 10.1255/jsi.2022.a1
Issue No: Vol. 11 (2022)
|
Open Access journal
