Linear unmixing protocol for hyperspectral image fusion analysis applied to a case study of vegetal tissues

Abstract

Hyperspectral imaging (HSI) is a useful non-invasive technique that offers spatial and chemical information of samples. Often, different HSI techniques are used to obtain complementary information from the sample by combining different image modalities (Image Fusion). However, issues related to the different spatial resolution, sample orientation or area scanned among platforms need to be properly addressed. Unmixing methods are helpful to analyze and interpret the information of HSI related to each of the components contributing to the signal. Among those, Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) offers very suitable features for image fusion, since it can easily cope with multiset structures formed by blocks of images coming from different samples and platforms and allows the use of optional and diverse constraints to adapt to the specific features of each HSI employed. In this work, a case study based on the investigation of cross-sections from rice leaves by Raman, synchrotron infrared and fluorescence imaging techniques is presented. HSI of these three different techniques are fused for the first time in a single data structure and analyzed by MCR-ALS. This example is challenging in nature and is particularly suitable to describe clearly the necessary steps required to perform unmixing in an image fusion context. Although this protocol is presented and applied to a study of vegetal tissues, it can be generally used in many other samples and combinations of imaging platforms.

Multimodal Imaging of Silicified Sorghum Leaves

“…In the future, the combination of the vibrational and elemental information in one type of hyper‐spectral image is desirable and being developed, similar to a recently presented approach that demonstrated the hyperspectral images of FTIR and Raman information [47] . The role of lignin and other cell wall components in the biomineralization in other plant tissues should be further studied, as substantial variation in cell wall autofluorescence is found between silica deposition sites and the rest of cell walls, such as sorghum roots, where autofluorescence depends on silicon supplementation/deprivation [29b] .…”

“…In the future, the combination of the vibrational and elemental information in one type of hyper-spectral image is desirable and being developed, similar to a recently presented approach that demonstrated the hyperspectral images of FTIR and Raman information. [47] The role of lignin and other cell wall components in the biomineralization in other plant tissues should be further studied, as substantial variation in cell wall autofluorescence is found between silica deposition sites and the rest of cell walls, such as sorghum roots, where autofluorescence depends on silicon supplementation/deprivation. [29b] The future in situ structural characterization of the lignin in such tissue regions will help to further delineate the relationship between the lignin structure and molecular composition and silica deposition without the need to isolate the lignin and to disrupt its interaction with the other cell wall components.…”

Advancing recycling efficiency through hyperspectral identification of fabric blends

“…The method optimizes C and S ′ matrices from equation 2 to provide concentrations of the components of the mixture under the action of certain constraints. 3 The main constraints that we provide to the algorithm are that the sum of all the concentrations is equal to one and that no concentration can be negative. We also provide the number of components in the textile blend.…”

“…They are supported by the Beer-Lambert law, which states that the absorption or transmission of light in a material is proportional to its concentration. 3 Unmixing methods provide pure spectra, or endmembers, which serve as the hyperspectral fingerprint of the components. MCR-ALS, unlike MR-PLS, doesn't require prior knowledge of fabric blends, making it advantageous for recycling purposes.…”

Comparison of 2D and 3D semantic segmentation in urban areas using fused hyperspectral and lidar data

“…Variousscientificfieldshaveappliedspectralunmixingin HSdataanalysisfordecades. [56][57][58][59][60] Here,abundancemapsforeachendmemberaregeneratedwithunsupervisedendmemberextractionandspectral unmixing, simplifying the analysis and effectively reducingdimensions.Hence,weavoidacurseofdimensionality, 61 whichoccurseasilywhenworkingwithlimited trainingdata.Insuchscenarios,thedeeplearningmodel hastodealwithmanyfeaturestoachievereliableand accurate results.…”