Hyperspectral Analysis Research Articles

Corrosion Risk Prediction of Carbon Steels by Hyperspectral Analysis

Corrosion Risk Prediction of Carbon Steels by Hyperspectral Analysis

Deep learning-driven super-resolution in Raman hyperspectral imaging: Efficient high-resolution reconstruction from low-resolution data

Deep learning (DL) has become an indispensable tool in hyperspectral data analysis, automatically extracting valuable features from complex, high-dimensional datasets. Super-resolution reconstruction, an essential aspect of hyperspectral data, involves enhancing spatial resolution, particularly relevant to low-resolution hyperspectral data. Yet, the pursuit of super-resolution in hyperspectral analysis is fraught with challenges, including acquiring ground truth high-resolution data for training, generalization, and scalability. The pressing issue of extended spectral acquisition times, notably for high-resolution scans, is a significant roadblock in hyperspectral imaging. Super-resolution methods offer a promising solution by providing higher spatial resolution data to expedite data collection and yield more efficient outcomes. This paper delves into a practical application of these concepts using Raman imaging, where spectral acquisition times can be prohibitively long. In this context, DL-based super-resolution models demonstrate their efficacy by predicting and reconstructing high-resolution Raman data from low-resolution input, eliminating the need for resource-intensive high-resolution scans. While previous work often relied on substantial high-resolution datasets, this study showcases the ability to achieve similar outcomes even with limited data, presenting a more practical and cost-effective approach. The results offer a glimpse into the transformative potential of this technology to streamline hyperspectral imaging applications by saving valuable time and resources through the successful generation of high-resolution data from low-resolution inputs.

Using a Two-Stage Hybrid Dimensionality Reduction Method on Hyperspectral Data to Predict Chlorophyll Content of Camellia oleifera

Camellia oleifera is an oilseed crop that holds significant economic, ecological, and social value. In the realm of Camellia oleifera cultivation, utilizing hyperspectral analysis techniques to estimate chlorophyll content can enhance our understanding of its physiological parameters and response characteristics. However, hyperspectral datasets contain information from many wavelengths, resulting in high-dimensional data. Therefore, selecting effective wavelengths is crucial for processing hyperspectral data and modeling in retrieval studies. In this study, by using hyperspectral data and chlorophyll content from Camellia oleifera samples, three different dimensionality reduction methods (Taylor-CC, NCC, and PCC) are used in the first round of dimensionality reduction. Combined with these methods, various thresholds and dimensionality reduction methods (with/without further dimensionality reduction) are used in the second round of dimensionality reduction; different sets of core wavelengths with equal size are identified respectively. Using hyperspectral reflectance data at different sets of core wavelengths, multiple machine learning models (Lasso, ANN, and RF) are constructed to predict the chlorophyll content of Camellia oleifera. The purpose of this study is to compare the performance of various dimensionality reduction methods in conjunction with machine learning models for predicting the chlorophyll content of Camellia oleifera. Results show that (1) the Taylor-CC method can effectively select core wavelengths with high sensitivity to chlorophyll variation; (2) the two-stage hybrid dimensionality reduction methods demonstrate superiority in three models; (3) the Taylor-CC + NCC method combined with an ANN achieves the best predictive performance of chlorophyll content. The new two-stage dimensionality reduction method proposed in this study not only improves both the efficiency of hyperspectral data processing and the predictive accuracy of models, but can serve as a complement to the study of Camellia oleifera properties using the Taylor-CC method.

Emerging Trends and Technologies Used for the Identification, Detection, and Characterisation of Plant-Parasitic Nematode Infestation in Crops.

Accurate identification and estimation of the population densities of microscopic, soil-dwelling plant-parasitic nematodes (PPNs) are essential, as PPNs cause significant economic losses in agricultural production systems worldwide. This study presents a comprehensive review of emerging techniques used for the identification of PPNs, including morphological identification, molecular diagnostics such as polymerase chain reaction (PCR), high-throughput sequencing, meta barcoding, remote sensing, hyperspectral analysis, and image processing. Classical morphological methods require a microscope and nematode taxonomist to identify species, which is laborious and time-consuming. Alternatively, quantitative polymerase chain reaction (qPCR) has emerged as a reliable and efficient approach for PPN identification and quantification; however, the cost associated with the reagents, instrumentation, and careful optimisation of reaction conditions can be prohibitive. High-throughput sequencing and meta-barcoding are used to study the biodiversity of all tropical groups of nematodes, not just PPNs, and are useful for describing changes in soil ecology. Convolutional neural network (CNN) methods are necessary to automate the detection and counting of PPNs from microscopic images, including complex cases like tangled nematodes. Remote sensing and hyperspectral methods offer non-invasive approaches to estimate nematode infestations and facilitate early diagnosis of plant stress caused by nematodes and rapid management of PPNs. This review provides a valuable resource for researchers, practitioners, and policymakers involved in nematology and plant protection. It highlights the importance of fast, efficient, and robust identification protocols and decision-support tools in mitigating the impact of PPNs on global agriculture and food security.

Scanless Spectral Imaging of Terahertz Vortex Beams Generated by High‐Resolution 3D‐Printed Spiral Phase Plates

Terahertz technology has experienced significant advances in the past years, leading to new applications in the fields of spectroscopy, imaging, and communications. This progress requires the development of dedicated optics to effectively direct, control and manipulate terahertz radiation. In this regard, 3D printing technologies have shown great potential, offering fast prototyping, high design flexibility, and good reproducibility. While traditional 3D printing techniques allow for the preparation of terahertz optical components operating at relatively low frequencies (<0.4 THz) due to their limited resolution, two‐photon polymerization lithography (TPL) exhibits high detail resolution and low surface roughness and can thus potentially enable the fabrication of high‐frequency terahertz devices. Here, as a proof of principle, spiral phase plates operating at 1 THz are designed and fabricated by means of TPL. Moreover, these samples are characterized via a rapid and scanless terahertz imaging technique customized to obtain a coherent hyperspectral analysis of the generated vortex beams at varying distances along propagation. Numerical simulations are also conducted for comparison with experiments, revealing a good agreement. Current limitations of the technique are found to be mainly related with terahertz loss in TPL polymers, and possible solutions are discussed.

Hyperspectral Analysis of Silicon Nanowires Manufactured Through Metal-Assisted Chemical Etching

Abstract This study used hyperspectral imaging to analyze localized near-field interactions between incident electromagnetic waves and silicon nanowire (SiNW) arrays manufactured through catalytic etching of Si wafers for different durations. The results revealed that the unetched upper surface area on Si wafers and reflection of incident light decreased with increasing etching time. A light reflection band peaking at approximately 880 nm was generated from arrays etched for more than 1 h. We used six separate hyperspectral images to analyze the wavelength-dependent spatial optical responses of the fabricated SiNW arrays. The images revealed hot spots of light reflection from unetched Si surfaces in the wavelength range of 470–750 nm and a resonant peak at 880 nm for a photonic crystal derived from a random SiNW array. Accordingly, hyperspectral imaging enables the assessment of localized optical responses of SiNW arrays, which can then be optimized to cater to various applications.

Detection Methods for Pine Wilt Disease: A Comprehensive Review.

Pine wilt disease (PWD), caused by the nematode Bursaphelenchus xylophilus, is a highly destructive forest disease that necessitates rapid and precise identification for effective management and control. This study evaluates various detection methods for PWD, including morphological diagnosis, molecular techniques, and remote sensing. While traditional methods are economical, they are limited by their inability to detect subtle or early changes and require considerable time and expertise. To overcome these challenges, this study emphasizes advanced molecular approaches such as real-time polymerase chain reaction (RT-PCR), droplet digital PCR (ddPCR), and loop-mediated isothermal amplification (LAMP) coupled with CRISPR/Cas12a, which offer fast and accurate pathogen detection. Additionally, DNA barcoding and microarrays facilitate species identification, and proteomics can provide insights into infection-specific protein signatures. The study also highlights remote sensing technologies, including satellite imagery and unmanned aerial vehicle (UAV)-based hyperspectral analysis, for their capability to monitor PWD by detecting asymptomatic diseases through changes in the spectral signatures of trees. Future research should focus on combining traditional and innovative techniques, refining visual inspection processes, developing rapid and portable diagnostic tools for field application, and exploring the potential of volatile organic compound analysis and machine learning algorithms for early disease detection. Integrating diverse methods and adopting innovative technologies are crucial to effectively control this lethal forest disease.

Magnesite hosted by the Neoarchean ultramafic rocks in Attappadi, southern India: Insights from spectral and stable isotope investigation

Magnesite is an economically important mineral commonly found in ultramafic complexes worldwide, primarily in Archean to Proterozoic ultramafic complexes. This study focuses on the chemical and spectral characterization of magnesite found in the Neoarchean ultramafic rocks in the Attappadi region in the Southern Granulite Terrane of southern India. The research utilizes x‐ray diffraction analysis, hyperspectral, laser Raman, Fourier Transform Infrared, and Isotope Ratio Mass Spectrometry. The studied ultramafic rocks are part of a well‐exposed ophiolitic suite known as the Agali ophiolitic complex. Magnesite primarily occurs as veins, veinlets, and lenses within weathered ultramafic rocks. The hyperspectral analysis of the magnesite samples shows absorption bands in the shortwave infrared region, particularly around 2.3 and 2.5 μm, which correspond to the stretching and bending of the CO bond in the (CO3)2− ion in MgCO3. The laser Raman spectra show intensity peaks at 1095, 738, and 330 cm−1, which may be attributed to the translational and librational vibrations. The Fourier transform infrared data reveal transmittance at 1434, 880, and 747 cm−1, corresponding to MgO bond stretching and asymmetrical CO stretching. The x‐ray powder diffraction spectra exhibit diffraction peaks at 32°, 35°, 42°, 46° and 53°, characteristic of pure magnesite. The spectroscopic parameters derived from various analyses indicate that the magnesite is high quality and free from gangue minerals. Stable isotope analysis of the magnesite samples yielded δ13C values ranging from −5‰ to −9‰ and δ18O values in the range of 21‰–25‰. The estimated water temperature from which the magnesite has been precipitated is ~59 ± 3.9°C. Based on the field relations, mode of occurrence and isotopic signatures, the mineralization is considered to have been formed by the low‐temperature alteration of ultramafic rocks facilitated by CO2‐rich fluids in the near‐surface environment. This study compares the characteristics of magnesite from the study area with a few Neoproterozoic serpentinite‐hosted magnesite veins in the ophiolitic sequence of the Egyptian Eastern Desert, which is part of the Arabian Nubian shield. The research aims to contribute to understanding magnesite formation in Archaean to Proterozoic mafic–ultramafic rocks on the Earth's crust. It also provides insights into the geological processes that govern the genesis of ultramafic‐hosted magnesite globally, particularly in East Gondwana fragments. This information can enhance mineral exploration and resource evaluation in these regions, helping to identify economic prospects and assess the feasibility of magnesite resource extraction and utilization in East Gondwana fragments.

APiS (AI powered intelligence spectrum analyser): hyperspectral imaging machine learning analyser app for mineral processing

ABSTRACT Hyperspectral data, encompassing hundreds of spectral wavelengths, is a powerful tool for mineral processing, as minerals display unique reflection spectra. Machine learning can identify mineral species from hyperspectral data, but its high dimensionality poses challenges for ordinary users. To address this, we developed APiS (AI Powered intelligence Spectrum Analyser), an interactive hyperspectral analysis application requiring no coding. APiS offers a user-friendly interface for data visualisation, learning data generation, advanced machine learning analysis, and model evaluation. This App surpasses previous script-based methods, enabling diverse users in geometallurgy to conduct sophisticated hyperspectral data analysis and fostering creativity in their work.

Oxidative weathering of pyrite in acidic environments: Data-driven experimental evaluation coupled with Raman hyperspectral imaging

This study deals with the identification of the factors that affect pyrite oxidation in acid mine drainage conditions. For this scope, weathering experiments have been carried at laboratory scale based on the design of experiments methodology to evaluate the effect of factors such as major ion concentrations, crystal size, and humic acids presence over the amount of elemental sulfur produced due to the involved weathering reactions. In particular, metal and anionic concentrations in solution were quantified by inductively coupled plasma-atomic emission spectroscopy and ion-chromatography techniques, respectively, whereas the amount of elemental sulfur was quantified with a high-performance liquid chromatography with diode-array detection technique after proper extraction procedure. A partial least squares regression was calculated to establish a quantitative relationship between the considered factors and the amount of elemental sulfur. After evaluation of the model, ferric iron, crystal size and the presence of humic acids were identified as the relevant factors for pyrite oxidation under acidic conditions. In addition, the surface of the samples was characterized by Raman imaging spectroscopy and subsequently analyzed by explorative hyperspectral analysis methods to assess the spatial distribution of the elemental sulfur as the main weathering product, resulting in a homogenous distribution.

Effects of Oxiris® Therapy on Cytokine Elimination after a LPS Infusion-An Experimental Animal Study.

The clinical effectiveness of Oxiris®, particularly in reducing cytokines, remains uncertain due to the limited data provided. This study explored and analyzed the application value of Oxiris® endotoxin adsorption technology in a large animal model. Pigs received an intravenous LPS infusion. Six animals were treated 2 h after the infusion with an Oxiris® hemadsorption using a pumpless extracorporeal technique for 6 h. Five animals served as controls. Cardiocirculatory parameters, hyperspectral analysis, and a panel of cytokines were measured. The lipopolysaccharide infusion induced sepsis-like inflammation with tachycardia, elevated pulmonary pressure, elevated lactate level, as well as elevated pro-inflammatory cytokines like interferon (IFN)-γ, interleukin (IL)-1β, IL-2, IL-6, IL-8, IL-12 and tumor necrosis factor alpha (TNF-α). In addition, increases of anti-inflammatory cytokines like IL-1ra and IL-10 were found. After 3 and 6 h in both groups, pro-inflammatory cytokines were significantly reduced. No differences between the intervention and the control group could be detected after 3 and 6 h for IL-1β, IL-2, IL-6, IL-8, IL-12 and TNF-α, suggesting no effect of the Oxiris® filter on the elimination of elevated cytokines with a pumpless extracorporeal hemadsorption technique. The presented large animal model may be a promising option for studying the effects of hemadsorption techniques.

Enhancing soil particle content prediction accuracy: advanced hyperspectral analysis and machine learning models

Enhancing soil particle content prediction accuracy: advanced hyperspectral analysis and machine learning models

Prediction of secondary metabolites in maize under different nitrogen inputs by hyperspectral sensing and machine learning

Flavonoids are compounds resulting from secondary plant metabolism that provide benefits to human health by food. This study aimed to accuracy of predicting flavonoids in maize plants subjected to different nitrogen rates using hyperspectral reflectance and machine learning (ML) algorithms. The experiment was carried out in randomized blocks in a 4 × 5 factorial design (four N inputs: 0; 30; 60 and 120 % of the recommended N input; and five readings of the reflectance spectra in maize leaves from different vegetative stages: V6, V8, V10, V12 and V14, in four replications, totaling 80 treatments. N rates were applied in the V4 and V8 phenological stages, using urea as the N source. For hyperspectral analysis, four leaves from each treatment were collected and analyzed using a spectroradiometer (FieldSpec 4 HRes, Analytical Spectral Devices), capturing the spectrum in the 350 to 2500 nm range. Subsequently, the leaf samples used in the reflectance readings were dried, ground and subjected to isoflavone quantification, analyzed by ultra-performance liquid chromatography in three repetitions, quantifying daidzein 1 (D1), daidzein 2 (D2), genistein 1 (G1), genistein 2 (G2), and total isoflavones. Data obtained was subjected to machine learning analysis, testing two data set input configurations: wavelengths (WL) and calculated spectral bands (B), and D1, D2, G1, G2 and total isoflavones as output variables. The ML algorithms tested were artificial neural networks (ANN), REPTree (DT), M5P decision tree (M5P), ZeroR (R), Random Forest (RF) and support vector machine (SVM), evaluated according to their performance by the correlation coefficient (r) and mean absolute error (MAE). The results show that the SVM algorithm had the highest accuracy in predicting the variables D1, D2, G1, G2 and total isoflavones, outperforming the other algorithms when WL was used as input in dataset.

Hyperspectral Tensor Completion Using Low-Rank Modeling and Convex Functional Analysis.

Hyperspectral tensor completion (HTC) for remote sensing, critical for advancing space exploration and other satellite imaging technologies, has drawn considerable attention from recent machine learning community. Hyperspectral image (HSI) contains a wide range of narrowly spaced spectral bands hence forming unique electrical magnetic signatures for distinct materials, and thus plays an irreplaceable role in remote material identification. Nevertheless, remotely acquired HSIs are of low data purity and quite often incompletely observed or corrupted during transmission. Therefore, completing the 3-D hyperspectral tensor, involving two spatial dimensions and one spectral dimension, is a crucial signal processing task for facilitating the subsequent applications. Benchmark HTC methods rely on either supervised learning or nonconvex optimization. As reported in recent machine learning literature, John ellipsoid (JE) in functional analysis is a fundamental topology for effective hyperspectral analysis. We therefore attempt to adopt this key topology in this work, but this induces a dilemma that the computation of JE requires the complete information of the entire HSI tensor that is, however, unavailable under the HTC problem setting. We resolve the dilemma, decouple HTC into convex subproblems ensuring computational efficiency, and show state-of-the-art HTC performances of our algorithm. We also demonstrate that our method has improved the subsequent land cover classification accuracy on the recovered hyperspectral tensor.

Metric-based meta-learning combined with hyperspectral imaging for rapid detection of adulteration in domain-shifted camel milk powder

Camel milk powder possesses high nutritional and economic value. The adulteration of camel milk powder with other varieties seriously compromises consumer rights. In the practical application of hyperspectral analysis for camel milk powder detection, the sample categories used for testing often differ from those used to construct the model. As a learning approach adept at domain-shifted and few shot scenarios, meta-learning is employed to tackle this issue. In this study, we used camel milk powder adulterated with cow milk powder as training samples and adulterated with goat milk powder as test samples. In the detection of eleven adulteration levels, the detection accuracy for pure camel milk powder reached 98.92%. Notably, the detection accuracy for the less conspicuous 70% adulteration level achieved 77.69%. The comprehensive detection accuracy of meta-learning reached 84.4%, showcasing notable improvements compared to SVM, BP, and CNN, which saw increases of 24.67%, 28.16%, and 18.4%, respectively. The detailed analysis of feature vectors and contributions substantiates the reliability and stability of the meta-learning-based qualitative analysis. The introduction of meta-learning methods is poised to make significant contributions to rapid detection by relevant testing agencies and the protection of consumer rights.

Improving the Estimation Accuracy of Soil Organic Matter Content Based on the Spectral Reflectance from Soils with Different Grain Sizes

Accurate and rapid estimation of soil organic matter (SOM) content is of great significance for advancing precision agriculture. Compared with traditional chemical methods, the hyperspectral estimation is superior in rapidly estimating SOM content. Soil grain size affects soil spectral reflectance, thereby affecting the accuracy of hyperspectral estimation. However, the appropriate soil grain size for the hyperspectral analysis is nearly unknown. This study propose a best hyperspectral estimation method for determining SOM content of farmland soil in the Ibinur Lake Irrigation Area (ILIA) of the northwest arid zones of China. The original spectral reflectance of the 20-mesh (0.85 mm) and 60-mesh (0.25 mm) sieved soil were obtained, and the feature wavebands were selected using five types of spectral transformations. Then, hyperspectral estimation models were constructed based on the partial least squares regression (PLSR), support vector machine (SVM), random forest (RF), and extreme gradient boosting (XGBoost) models. Results show that the SOM content had relatively higher correlation coefficient with spectral reflectance of the 0.85 mm sieved soil than that of the 0.25 mm sieved soil. The transformation of original spectral reflectance of soil effectively enhanced the spectral characteristics related to SOM content. Soil grain size obviously affected spectral reflectance and the accuracy of hyperspectral estimation models. The overall stability and estimation accuracy of RF model was significantly higher compared with the PLSR, SVM, and XGBoost. Finally, the RF model combined with the root mean first-order differentiation (RMSFD) of spectral reflectance of the 0.85 mm sieved soil (R2 = 0.82, RMSE = 2.37, RPD = 2.27) was identified as the best method for estimating SOM content of farmland soil in the ILIA.

O-249 METAPHOR: METabolic imaging through AI-powered Phasor-based Hyperspectral analysis and Organelle recognition for the classification of human blastocysts

Abstract Study question Can we classify human blastocysts non-invasively using hyperspectral imaging? Summary answer HS imaging can significantly enhance our classification of human embryos and reveal numerous cellular features present in the embryo invisible to brightfield imaging. What is known already Non-invasive measurement of metabolism is a safe and powerful tool to reliably predict oocyte and embryo quality. We have previously shown that our METAPHOR technology, based on HS imaging coupled to artificial intelligence, was able to separate mouse oocyte categories according to maternal age with an average AUC of 96.2% and predict blastocyst formation (AUC 82.2%). Moreover, it was able to classify mouse embryos deprived from different nutrients with an average AUC of 93.7% showing clear superiority to grading provided by experienced embryologists (51%). Study design, size, duration A total of 74 human embryos donated for research at the blastocyst stage have been imaged to date. Embryos were warmed and imaged according to our established protocol and survival was statistically indistinguishable from control embryos. Participants/materials, setting, methods Embryos are imaged using multiphoton illumination near the infra-red. The HS detection covered the simultaneous measurement of the whole visible spectrum. Embryos are then allowed to implant in our proprietary 3D ex vivo platform. Implantation is monitored up until day 9 before immunostaining is performed to quantify presence and prevalence of lineage and implantation markers. Main results and the role of chance After imaging, embryos were placed on the 3D ex vivo implantation platform and let implant. Successful implantation (75.41%) was scored by observation of the characteristic deformation of collagen fibers within the matrix and staining for relevant cell lineages (OCT4, pMLC2). Interestingly, we observed that embryos with a higher trophectoderm (TE) grading tended to implant with a higher efficiency and penetrate deeper into the matrix than embryos with lower grading. HS images have unveiled a profound cellular diversity within embryos, elucidating distinctions between the Inner Cell Mass (ICM) and TE, as well as revealing different metabolic states defined by NADH dominance, FAD dominance, protoporphyrin levels, and apoptosis. HS raw images are processed using our HS-phasor approach which allows quick computing by converting the multidimensional HS data into comprehensive visual representation of bidimensional histograms and phasor plots. In summary, our analysis pipeline encompasses the algorithms to detect and segment the embryos, the computation of histogram and phasor-plot, and the classifier. AI allows us to classify embryos in distinct spectral spaces based on their metabolic features, paving the way to the establishment of a new grading system based on HS imaging. Limitations, reasons for caution The cohort size is still limited, and ongoing experiments will increase the dataset needed for more accurate prediction of embryo implantation. The implantation platform is an in vitro test that will require future clinical validation. Wider implications of the findings The safety and speed of METAPHOR warrants its incorporation at the IVF clinic, adding an additional layer of information to the current embryo selection methods. Moreover, the non-invasive measurement of the metabolic function opens new alleys of research such as the effect of media and culture supplements on embryo culture. Trial registration number Not applicable

Advanced mineralogical classification and concentration estimation in mining with MATLAB-powered hyper-spectral imaging and machine learning

ABSTRACT The study presents a new technique that combines hyperspectral imaging and machine learning to identify minerals. Overcoming the challenges of processing hyperspectral data, the solution offers a user-friendly hyperspectral analysis tool tailored for constructing datasets and enhancing mineral identification and concentration estimation. The tool integrates hyperspectral data processing with segmentation, simplifying complex operations and making mineral identification accessible to non-experts. The tool’s capabilities extend to handling multi-spectral data efficiently, potentially leading to energy-efficient analysis when combined with dimensionality reduction techniques. Presented as a novel approach, it improves geological surveys in mining areas, enabling industrial applications and mineral research.

Soybean Saponin Content Detection Based on Spectral and Image Information Combination

Soybean saponin is a natural antioxidant and is anti-inflammatory. Hyperspectral analysis technology was applied to detect soybean saponin content rapidly and nondestructively in this paper. Firstly, spectral preprocessing methods were studied, and standard normal variable (SNV) was used to remove noise information. Secondly, a two-step hybrid variable selection approach based on synergy interval partial least squares (SiPLS) and iteratively retains informative variables (IRIV) was proposed to extract characteristic variables. Then, the ensemble learning model was constructed by back propagation neural network (BPNN), deep forest (DF), partial least squares regression (PLSR), and extreme gradient boosting (EXG). Finally, image information was combined into spectral data to improve model accuracy. The prediction coefficient of determination (R2) of the final model reached 0.9216. It can provide rapid, nondestructive, and accurate detection technology of soybean saponin content. A combination of spectral and image information will provide a new idea for application of hyperspectral.

Estimating the chromium concentration of farmland soils in an arid zone from hyperspectral reflectance by using partial least squares regression methods

Heavy metal contamination of farmland soils is a severe concern worldwide.Chromium (Cr), a major environmental carcinogen, is a crucial indicator of farmland soil pollution. Compared with conventional chemical analysis, the hyperspectral analysis is superior in accurately and rapidly surveilling heavy metals in soils. However, various heavy metals, and types or geochemical traits of soils often affect the accuracy of hyperspectral estimation models, which makes guaranteeing the concentration estimation precision for a designated heavy metal in different regions considerably challenging. We here propose an optimal hyperspectral model for determining the Cr concentration of farmland soils first time in the Yanqi Basin of China's northwest arid zone. In total, 171 soil samples were gathered, and their Cr concentrations and relevant spectral reflectance data were determined. The best spectral transformation for measuring the Cr concentration was determined by using 12 transformation types of soil spectral curves, and characteristic wavebands were obtained. Then, by using random forest (RF), support vector machine (SVM), and partial least squares regression (PLSR) models, we constructed hyperspectral estimation models for Cr concentration in farmland soils. The estimated results of these models were compared to identify the best estimation model. Results show that the spectral transformations allow an improvement in the association between the soil Cr concentration and spectra. The accuracy of PLSR was significantly higher than that of SVM and RF. Finally, the second-order differentiation (SD) combined with the PLSR model (R2 = 0.903, RMSE = 40.363, MAE = 30.631) was identified as the optimal model for estimating the Cr concentration of farmland soils in the investigated area. The study findings offer a technical reference for the hyperspectral assessment of the Cr concentration in arid zone farmland soils.