Hyperspectral Reflectance Imaging Research Articles

Spectral Adaptation Transform for Multispectral Constancy

The spectral reflectance of an object surface provides valuable information of its characteristics. Reflectance reconstruction from multispectral image data is typically based on certain assumptions. One of the common assumptions is that the same illumination is used for system calibration and image acquisition. The authors propose the concept of multispectral constancy which transforms the captured sensor data into an illuminant-independent representation, analogously to the concept of computational color constancy. They propose to transform the multispectral image data to a canonical representation through spectral adaptation transform (SAT). The performance of such a transform is tested on measured reflectance spectra and hyperspectral reflectance images. The authors also investigate the robustness of the transform to the inaccuracy of illuminant estimation in natural scenes. Results of reflectance reconstruction show that the proposed SAT is efficient and is robust to error in illuminant estimation.

The application of near-infrared reflectance hyperspectral imaging for the detection and extraction of bloodstains

Blood is not only the traces of the murder scene but also the evidence, has a very important role in the acquisition of criminal evidence and forensic identification. However, the conventional testing methodology being currently deployed in bloodstain identification may destruct the human DNA. Some investigators used the method is undeserved, often regret in the exploration of the scene. From this point of view, efficient non-destructive technology for quickly identifying bloodstain is an emerging issue in this field. In this article, a method is proposed for nondestructive testing of identifying blood stains, and design a blood model named blood band inequality model to extract blood from eight kinds of suspected blood. The validity of the model was verified by the theory and experiment. Extraction efficiency of blood is as high as 94.71%, and the omission rate is only 5.29%. Experiments proved hyperspectral imaging can be used for the non-contact detection, identification of blood stains and the new model is effective in the extraction of blood.

Identification of Bruise and Fungi Contamination in Strawberries Using Hyperspectral Imaging Technology and Multivariate Analysis

Mechanical bruise and fungi contamination are two typical defective features for strawberries, resulting in quick quality deterioration of the strawberries during transportation and storage. In this work, the approach of combined image processing with spectra analysis was successfully developed to identify defective strawberries (bruised and fungal infected) using hyperspectral reflectance imaging system. Hyperspectral image data was exploited by minimum noise fraction (MNF) transformation for strawberry defects distinguished by combining thresholding and morphology procedures, and defective regions were located and separated for spectra extracting. The linkages between quality parameters and spectra features were established based on the target defective regions of the fruit. After spectra normalization, three different spectral regions (400 to 600 nm, 650 to 720 nm, and 900 to 1010 nm) were identified for healthy, bruised, or infected strawberries, and eight optimal wavelengths were selected by the successive projection algorithms (SPA) from the whole range of wavelengths. Both linear and non-linear algorithms were developed to identify defective types in strawberries. The results indicated that based on full wavelengths, SVM model performed the highest overall identification accuracy, with the accuracy of 96.91% for calibration and 92.59% for prediction of the fruit. This work shows that hyperspectral reflectance imaging technology has the potential for identifying defective strawberries and provides theoretical basis for the development of online classification of different defected fruits.

Non-destructive analysis of sucrose, caffeine and trigonelline on single green coffee beans by hyperspectral imaging

Hyperspectral imaging (HSI) is a novel technology for the food sector that enables rapid non-contact analysis of food materials. HSI was applied for the first time to whole green coffee beans, at a single seed level, for quantitative prediction of sucrose, caffeine and trigonelline content. In addition, the intra-bean distribution of coffee constituents was analysed in Arabica and Robusta coffees on a large sample set from 12 countries, using a total of 260 samples. Individual green coffee beans were scanned by reflectance HSI (980–2500nm) and then the concentration of sucrose, caffeine and trigonelline analysed with a reference method (HPLC-MS). Quantitative prediction models were subsequently built using Partial Least Squares (PLS) regression. Large variations in sucrose, caffeine and trigonelline were found between different species and origin, but also within beans from the same batch. It was shown that estimation of sucrose content is possible for screening purposes (R2=0.65; prediction error of ~0.7% w/w coffee, with observed range of ~6.5%), while the performance of the PLS model was better for caffeine and trigonelline prediction (R2=0.85 and R2=0.82, respectively; prediction errors of 0.2 and 0.1%, on a range of 2.3 and 1.1% w/w coffee, respectively). The prediction error is acceptable mainly for laboratory applications, with the potential application to breeding programmes and for screening purposes for the food industry. The spatial distribution of coffee constituents was also successfully visualised for single beans and this enabled mapping of the analytes across the bean structure at single pixel level.

Study the effects of varying interference upon the optical properties of turbid samples using NIR spatial light modulation

Optical properties of biological tissues are valuable diagnostic parameters which can provide necessary information regarding tissue state during disease pathogenesis and therapy. However, different sources of interference, such as temperature changes may modify these properties, introducing confounding factors and artifacts to data, consequently skewing their interpretation and misinforming clinical decision-making. In the current study, we apply spatial light modulation, a type of diffuse reflectance hyperspectral imaging technique, to monitor the variation in optical properties of highly scattering turbid media in the presence varying levels of the following sources of interference: scattering concentration, temperature, and pressure. Spatial near-infrared (NIR) light modulation is a wide-field, non-contact emerging optical imaging platform capable of separating the effects of tissue scattering from those of absorption, thereby accurately estimating both parameters. With this technique, periodic NIR illumination patterns at alternately low and high spatial frequencies, at six discrete wavelengths between 690 to 970 nm, were sequentially projected upon the medium while a CCD camera collects the diffusely reflected light. Data analysis based assumptions is then performed off-line to recover the medium’s optical properties. We conducted a series of experiments demonstrating the changes in absorption and reduced scattering coefficients of commercially available fresh milk and chicken breast tissue under different interference conditions. In addition, information on the refractive index was study under increased pressure. This work demonstrates the utility of NIR spatial light modulation to detect varying sources of interference upon the optical properties of biological samples.

Potential of VIS-NIR hyperspectral imaging and chemometric methods to identify similar cultivars of nectarine

Product inspection is essential to ensure good quality and to avoid fraud. New nectarine cultivars with similar external appearance but different physicochemical properties may be mixed in the market, causing confusion and rejection among consumers, and consequently affecting sales and prices. Hyperspectral reflectance imaging in the range of 450–1040 nm was studied as a non-destructive method to differentiate two cultivars of nectarines with a very similar appearance but different taste. Partial least squares discriminant analysis (PLS-DA) was used to develop a prediction model to distinguish intact fruits of the cultivars using pixel-wise and mean spectrum approaches, and then the model was projected onto the complete surface of fruits allowing visual inspection. The results indicated that mean spectrum of the fruit was the most accurate method, a correct discrimination rate of 94% being achieved. Wavelength selection reduced the dimensionality of the hyperspectral images using the regression coefficients of the PLS-DA model. An accuracy of 96% was obtained by using 14 optimal wavelengths, whereas colour imaging and a trained inspection panel achieved a rate of correct classification of only 57% of the fruits.

Identifying eighteenth century pigments at the Bodleian library using in situ Raman spectroscopy, XRF and hyperspectral imaging

There are multiple challenges in analysing pigments in historic watercolour paintings on paper, and typically non-invasive, in situ methods are required. Recent developments in portable analytical instrumentation have made this more accessible to heritage institutions, but many commercial systems are not optimised for the specific requirements of manuscripts and works on paper. This paper describes the successful use of Raman spectroscopy, X-ray fluorescence spectroscopy (XRF) and hyperspectral imaging to identify and map watercolour pigments used by the eighteenth century botanical illustrator, Ferdinand Bauer, and demystify the unusual colour code system found in his sketches. The value, delicate nature and large size of these paintings necessitated the use of using in situ, non-contact methods of analysis. A portable, bespoke Raman spectrometer specifically designed for analysing pigments from works on paper was used together with a bespoke portable Fibre optic reflectance spectrometer, portable X-Ray Fluorescence spectrometer and a hyperspectral imaging sensor. The results demonstrate that although there is a significant compromise between achieving good Raman spectroscopic results from artists’ pigments and using sufficiently low laser power densities so as not to cause damage to the pigments, good results could be obtained with this portable system, particularly when combined with XRF, fibre optic reflectance spectroscopy (FORS) and hyperspectral imaging. Eight pigments were identified unequivocally from 125 watercolour paintings analysed, suggesting that Bauer used a more traditional and more limited palette than previously considered, and that his palette changed significantly in his later paintings. Similar pigments identified by the authors on colour chart that was discovered in 1999 in Madrid and attributed to Bauer, add weight to the attribution of this chart to Bauer. The data provides a much deeper insight into Bauer’s colour annotations, and how he was able to achieve such an impressive degree of colour fidelity in his work.

Assessment of multispectral and hyperspectral imaging systems for digitisation of a Russian icon

In a study of multispectral and hyperspectral reflectance imaging, a Round Robin Test assessed the performance of different systems for the spectral digitisation of artworks. A Russian icon, mass-produced in Moscow in 1899, was digitised by ten institutions around Europe. The image quality was assessed by observers, and the reflectance spectra at selected points were reconstructed to characterise the icon’s colourants and to obtain a quantitative estimate of accuracy. The differing spatial resolutions of the systems affected their ability to resolve fine details in the printed pattern. There was a surprisingly wide variation in the quality of imagery, caused by unwanted reflections from both glossy painted and metallic gold areas of the icon’s surface. Specular reflection also degraded the accuracy of the reconstructed reflectance spectrum in some places, indicating the importance of control over the illumination geometry. Some devices that gave excellent results for matte colour charts proved to have poor performance for this demanding test object. There is a need for adoption of standards for digitising cultural heritage objects to achieve greater consistency of system performance and image quality.

Detection of blueberry internal bruising over time using NIR hyperspectral reflectance imaging with optimum wavelengths

Early detection of internal bruising is one of the major challenges in blueberry postharvest quality sorting processes. The potential of using near infrared (NIR) hyperspectral reflectance imaging (950–1650nm) with reduced spectral features was investigated for blueberry internal bruising detection 30min to 12h after mechanical impact. A least squares support vector machine (LS-SVM) was used to develop classification models to compute the spatial distribution of bruising based on the spectra extracted from regions of interest (ROIs) at four measurement times (30min, 2h, 6h, and 12h after mechanical impact). Three feature selection methods were used to select optimum wavelengths or band ratio images for bruising detection. The classification model, developed using optimum wavelengths selected by competitive adaptive reweighted sampling (CARS) (CARS-LS-SVM model) and full spectra (full spectra-LS-SVM), had similar performance in the identification of bruised blueberries. Band ratio images (1235nm/1035nm) achieved a comparable accuracy with the CARS-LS-SVM model at 6h, and higher accuracy than CARS-LS-SVM and full spectra-LS-SVM models at 12h. The overall classification accuracies of 77.5%, 83.8%, 92.5%, and 95.0% were obtained by band ratio images for blueberries 30min, 2h, 6h, and 12h after impact, respectively. In order to evaluate the performance of the proposed methods, additional validation samples were processed by the detection algorithm. The overall discrimination accuracies for healthy and bruised blueberries in the validation set were 93.3% and 98.0%, respectively, for CARS-LS-SVM model, and 93.3% and 95.9%, respectively, for band ratio images. The overall results indicated that NIR reflectance imaging can detect blueberry internal bruising as early as 30min after mechanical impact, and band ratio images computed from two wavelengths showed great potential to detect blueberry internal bruising on the packing line.

Label-free reflectance hyperspectral imaging for tumor margin assessment: a pilot study on surgical specimens of cancer patients.

.A label-free, hyperspectral imaging (HSI) approach has been proposed for tumor margin assessment. HSI data, i.e., hypercube , consist of a series of high-resolution images of the same field of view that are acquired at different wavelengths. Every pixel on an HSI image has an optical spectrum. In this pilot clinical study, a pipeline of a machine-learning-based quantification method for HSI data was implemented and evaluated in patient specimens. Spectral features from HSI data were used for the classification of cancer and normal tissue. Surgical tissue specimens were collected from 16 human patients who underwent head and neck (H&N) cancer surgery. HSI, autofluorescence images, and fluorescence images with 2-deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]-D-glucose (2-NBDG) and proflavine were acquired from each specimen. Digitized histologic slides were examined by an H&N pathologist. The HSI and classification method were able to distinguish between cancer and normal tissue from the oral cavity with an average accuracy of , sensitivity of , and specificity of . For tissue specimens from the thyroid, the method achieved an average accuracy of , sensitivity of , and specificity of . HSI outperformed autofluorescence imaging or fluorescence imaging with vital dye (2-NBDG or proflavine). This study demonstrated the feasibility of label-free, HSI for tumor margin assessment in surgical tissue specimens of H&N cancer patients. Further development of the HSI technology is warranted for its application in image-guided surgery.

Hyperspectral Imaging for Predicting the Internal Quality of Kiwifruits Based on Variable Selection Algorithms and Chemometric Models

We investigated the feasibility and potentiality of determining firmness, soluble solids content (SSC), and pH in kiwifruits using hyperspectral imaging, combined with variable selection methods and calibration models. The images were acquired by a push-broom hyperspectral reflectance imaging system covering two spectral ranges. Weighted regression coefficients (BW), successive projections algorithm (SPA) and genetic algorithm–partial least square (GAPLS) were compared and evaluated for the selection of effective wavelengths. Moreover, multiple linear regression (MLR), partial least squares regression and least squares support vector machine (LS-SVM) were developed to predict quality attributes quantitatively using effective wavelengths. The established models, particularly SPA-MLR, SPA-LS-SVM and GAPLS-LS-SVM, performed well. The SPA-MLR models for firmness (Rpre = 0.9812, RPD = 5.17) and SSC (Rpre = 0.9523, RPD = 3.26) at 380–1023 nm showed excellent performance, whereas GAPLS-LS-SVM was the optimal model at 874–1734 nm for predicting pH (Rpre = 0.9070, RPD = 2.60). Image processing algorithms were developed to transfer the predictive model in every pixel to generate prediction maps that visualize the spatial distribution of firmness and SSC. Hence, the results clearly demonstrated that hyperspectral imaging has the potential as a fast and non-invasive method to predict the quality attributes of kiwifruits.

Reflectance Spectroscopy and Hyperspectral Imaging of Sapphire-Bearing Marble From the Beluga Occurrence, Baffin Island, Nunavut

Abstract Baffin Island (Nunavut, Canada) is located in the remote arctic where geological exposure is high, terrain is expansive, and field seasons are short. Metacarbonates in the Lake Harbour Group, southern Baffin Island, are considered to be prospective for sapphire (corundum), lapis lazuli (lazurite), and spinel gem deposits. This research is an investigation of the hyperspectral signature of rocks from the Beluga sapphire occurrence near Kimmirut, Nunavut. Thirty-five rock and thin section offcut samples from the Beluga corundum occurrence were studied in a laboratory setting using high spatial resolution imaging spectroscopy in the shortwave infrared region (SWIR, ∼975 to 2500 nm). Analysis of hyperspectral imagery successfully reproduced mineralogical and textural information relevant to gem mineralization previously assessed through thin section petrography and scanning electron microscopy. Scapolite, phlogopite, and muscovite from these rocks are shown to have distinct spectral responses from the host marble that is dominated by calcite. Furthermore, spectrally distinct prehnite and zeolite, both locally associated with corundum mineralization, are readily distinguishable. From the perspective of regional exploration, the small footprint of colored gemstone deposits in combination with their unique mineralogical makeup, unusual settings, and unfamiliarity to most geoscientists has resulted in very few discoveries in this remote but otherwise fertile area. These factors, in combination with the unique spectral responses observed from hand samples and thin section offcuts, provide a positive case for mapping and exploration of gem corundum targets by high spatial resolution hyperspectral imaging as part of regional surveys.

Ripeness monitoring of two cultivars of nectarine using VIS-NIR hyperspectral reflectance imaging

Visible–near-infrared (450–1040 nm) hyperspectral reflectance imaging was studied in order to assess the internal physicochemical properties and sensory perception of ‘Big Top’ and ‘Magique’ nectarines (Prunus persica L. Batsch var. nucipersica) (yellow and white-flesh cultivar, respectively) during ripening using the Ripening Index (RPI) and the Internal Quality Index (IQI). Hyperspectral images of the intact fruits were acquired during the ripeness under controlled conditions, and their physicochemical properties (flesh firmness, total soluble solids, titratable acidity and flesh colour) were analysed. IQI and RPI were used to relate the spectral information obtained from nectarines with the physicochemical properties and the sensory perception of their maturity using Partial Least Square (PLS) regression with proper variable selection. Optimal results were obtained with R2 values higher than 0.87 for the two indices and the two cultivars. The ripeness of each fruit could be visualised by projecting the PLS models of the IQI on the pixels of the fruits in the images, showing great potential for further monitoring of the evolution of intact nectarine ripeness in industrial setups.

Hyperspectral Imaging for Presymptomatic Detection of Tobacco Disease with Successive Projections Algorithm and Machine-learning Classifiers

We investigated the feasibility and potentiality of presymptomatic detection of tobacco disease using hyperspectral imaging, combined with the variable selection method and machine-learning classifiers. Images from healthy and TMV-infected leaves with 2, 4, and 6 days post infection were acquired by a pushbroom hyperspectral reflectance imaging system covering the spectral range of 380–1023 nm. Successive projections algorithm was evaluated for effective wavelengths (EWs) selection. Four texture features, including contrast, correlation, entropy, and homogeneity were extracted according to grey-level co-occurrence matrix (GLCM). Additionally, different machine-learning algorithms were developed and compared to detect and classify disease stages with EWs, texture features and data fusion respectively. The performance of chemometric models with data fusion manifested better results with classification accuracies of calibration and prediction all above 80% than those only using EWs or texture features; the accuracies were up to 95% employing back propagation neural network (BPNN), extreme learning machine (ELM), and least squares support vector machine (LS-SVM) models. Hence, hyperspectral imaging has the potential as a fast and non-invasive method to identify infected leaves in a short period of time (i.e. 48 h) in comparison to the reference images (5 days for visible symptoms of infection, 11 days for typical symptoms).

Automatic Identification of Rock-Forming Minerals in Granite Using Laboratory Scale Hyperspectral Reflectance Imaging and Artificial Neural Networks

Laser cleaning of stones is a well-established technique in conservation of Cultural Heritage. In case of polymineralic granular rocks like granite, the tuning of the laser parameters for safe cleaning becomes a demanding task due to the differential response of the constituent minerals to laser irradiation. In this sense, a compromise solution must be reached between removing unwanted layers and safeguarding the original characteristics of the stone substrate. Consequently, the need of non-invasive, rapid and accurate methods for in situ identification of the minerals in the stone surface is highlighted. The aim of this paper is to demonstrate the ability of a laboratory scale hyperspectral reflectance imaging system, in combination with artificial neural networks for in situ, accurate identification of constituent minerals of a hercynian granite; with the ultimate goal of achieving automatic adjustment of laser irradiation parameters in cleaning processes. The effectiveness of a neural network with the structure of a three-layer perceptron was evaluated by comparing calculated results with modal analysis composition obtained by petrographic microscope and a high degree of accordance, above 95%, was achieved.

Detection of blood in fish muscle by constrained spectral unmixing of hyperspectral images

A new method for blood detection in whitefish fillets is presented. The method is based on diffuse reflectance hyperspectral imaging in the VIS/NIR range, and unmixing of measured absorbance spectra into known spectra for hemoglobin, water and muscle tissue. Scattering effects are modelled as a polynomial function of wavelength, and are included in the unmixing procedure. The unmixing method is based on a modified Beer-Lambert law and is performed using a non-negativity constrained least squares algorithm. Application of the method to images of samples with varying amounts of blood results in a very good fit between measured and modelled spectra. Reference measurements on homogenized cod muscle with controlled concentrations of added blood are used to model the relationship between the spectral unmixing results and hemoglobin concentration, enabling pixel-by-pixel estimation of hemoglobin concentration. The method is seen as a useful tool for automatic quality grading and processing of fillets in the whitefish industry.

Discrimination methods for biological contaminants in fresh-cut lettuce based on VNIR and NIR hyperspectral imaging

The rapid detection of biological contaminants such as worms in fresh-cut vegetables is necessary to improve the efficiency of visual inspections carried out by workers. Multispectral imaging algorithms were developed using visible-near-infrared (VNIR) and near-infrared (NIR) hyperspectral imaging (HSI) techniques to detect worms in fresh-cut lettuce. The optimal wavebands that can detect worms in fresh-cut lettuce were investigated for each type of HSI using one-way ANOVA. Worm-detection imaging algorithms for VNIR and NIR imaging exhibited prediction accuracies of 97.00% (RI547/945) and 100.0% (RI1064/1176, SI1064-1176, RSI-I(1064-1173)/1064, and RSI-II(1064-1176)/(1064+1176)), respectively. The two HSI techniques revealed that spectral images with a pixel size of 1×1mm or 2×2mm had the best classification accuracy for worms. The results demonstrate that hyperspectral reflectance imaging techniques have the potential to detect worms in fresh-cut lettuce. Future research relating to this work will focus on a real-time sorting system for lettuce that can simultaneously detect various defects such as browning, worms, and slugs.

On-line fresh-cut lettuce quality measurement system using hyperspectral imaging

In this study, an online quality measurement system for detecting foreign substances on fresh-cut lettuce was developed using hyperspectral reflectance imaging. The online detection system with a single hyperspectral camera in the range of 400–1000 nm was able to detect contaminants on both surfaces of fresh-cut lettuce. Algorithms were developed for this system to detect contaminants such as slugs and worms. The optimal wavebands for discriminating between contaminants and sound lettuce as well as between contaminants and the conveyor belt were investigated using the one-way analysis of variance (ANOVA) method. The subtraction imaging (SI) algorithm to classify slugs resulted in a classification accuracy of 97.5%, sensitivity of 98.0%, and specificity of 97.0%. The ratio imaging (RI) algorithm to discriminate worms achieved classification accuracy, sensitivity, and specificity rates of 99.5%, 100.0%, and 99.0%, respectively. The overall results suggest that the online quality measurement system using hyperspectral reflectance imaging can potentially be used to simultaneously discriminate foreign substances on fresh-cut lettuces.

Nede�trukcn� detekcia a biochemick� kvantifik�cia listov poh�nky s vyu�it�m hyperspektr�lneho zobrazovania s reflektanciou vo viditelnej (VIS) a bl�zkej infracervenej (NIR) oblasti

The present experimental study has been investigated way of use methodology of nondestructive detection with hyperspectral reflectance imaging together with wet chemistry quantitative analysis based on correlation analysis of receiving data. Higher correlation from reflectance of hyperspectral imaging analysis and total polyphenol and anthocyanin content has been observed for carotenoid reflectance index-2 compared to the carotenoid reflectance index-1. It was found that high total polyphenols content is related with high values of hyperspectral indices which characterize chlorophyll concentration and parameters of vegetation. In experimental buckwheat cultivars presence of vanillic, methoxycinnamic, cinnamic acids can be expressing by anthocyanin reflectance index, modified anthocyanin reflectance index and carotenoid reflectance index-2 because high correlation coefficient. Changes of chlorogenic acid during the vegetative period of plant growth can be studied with normalized difference vegetation index and normalized difference vegetation index-2 which found to have a positive correlation with this phenolic acid content.

Application of Hyperspectral Imaging and Acoustic Emission Techniques for Apple Quality Prediction

Abstract. There is a growing demand for developing effective non-destructive quality assessment methods with quick response, high accuracy, and low cost for fresh fruits. In this study, hyperspectral reflectance imaging (400 to 1000 nm) and acoustic emission (AE) tests were applied to ‘GoldRush’ apples (total number, n = 180) to predict fruit firmness, total soluble solids (TSS), and surface color parameters (L*, a*, b*) during an eight-week storage period. Partial least squares (PLS) regression, least squares support vector machine (LS-SVM), and multivariate linear regression (MLR) methods were used to establish models to predict the quality attributes of the apples. The results showed that hyperspectral imaging (HSI) could accurately predict all the attributes except TSS, while the AE method was capable of predicting fruit firmness, b* color index, and TSS. Overall, HSI regression using PLS had better comprehensive ability for predicting firmness, TSS, and color parameters (L*, a*, b*) than AE, with correlation coefficients of prediction (rp) of 0.92, 0.41, 0.83, 0.87, and 0.94 and root mean square errors of prediction (RMSEP) of 4.32 (N), 1.78 (°Brix), 3.41, 2.28, and 4.29, respectively, while AE regression using LS-SVM gave rp values of 0.88, 0.74, 0.34, 0.37, and 0.81 and RMSEP values of 4.26 (N), 0.64 (°Brix), 4.69, 1.8, and 5.17 for firmness, TSS, and color parameters (L*, a*, b*), respectively. The results show the potential of these two non-destructive methods for predicting some of the quality attributes of apples. Keywords: Apple, Acoustic emission, Fruit quality, Hyperspectral imaging, Regression model.