Hyperspectral Data Analysis Research Articles

Recognition of Abnormal-Laying Hens Based on Fast Continuous Wavelet and Deep Learning Using Hyperspectral Images.

The egg production of laying hens is crucial to breeding enterprises in the laying hen breeding industry. However, there is currently no systematic or accurate method to identify low-egg-production-laying hens in commercial farms, and the majority of these hens are identified by breeders based on their experience. In order to address this issue, we propose a method that is widely applicable and highly precise. First, breeders themselves separate low-egg-production-laying hens and normal-laying hens. Then, under a halogen lamp, hyperspectral images of the two different types of hens are captured via hyperspectral imaging equipment. The vertex component analysis (VCA) algorithm is used to extract the cockscomb end member spectrum to obtain the cockscomb spectral feature curves of low-egg-production-laying hens and normal ones. Next, fast continuous wavelet transform (FCWT) is employed to analyze the data of the feature curves in order to obtain the two-dimensional spectral feature image dataset. Finally, referring to the two-dimensional spectral image dataset of the low-egg-production-laying hens and normal ones, we developed a deep learning model based on a convolutional neural network (CNN). When we tested the model's accuracy by using the prepared dataset, we found that it was 0.975 percent accurate. This outcome demonstrates our identification method, which combines hyperspectral imaging technology, an FCWT data analysis method, and a CNN deep learning model, and is highly effective and precise in laying-hen breeding plants. Furthermore, the attempt to use FCWT for the analysis and processing of hyperspectral data will have a significant impact on the research and application of hyperspectral technology in other fields due to its high efficiency and resolution characteristics for data signal analysis and processing.

Evaluation of Surface Damage of Strawberry Grasped by Manipulator Based on Vision and Hyperspectral Data Analysis

In the current study, an accurate strategy was developed to detect the surface damage of the strawberries, because the existing indicators for evaluating surface damage are single, resulting in low accuracy of judgment for damage detection. Strawberry surface damage can be divided into three specific categories: outline depression, surface oxidation, and surface damage. The methodologies employed in this study are experimental analysis and comparative validation. The outline, area of surface damage, and surface reflectance of strawberries were analyzed using visual and hyperspectral data. Principal component analysis was used to evaluate the damage characteristics comprehensively. A four-finger rigid manipulator having one degree of freedom was selected, and 280 samples were analyzed (the manipulator grasped 85% of them, and 15% were not treated). The accuracy of damage detection based on the outlines, area of surface damage, and surface reflectance of strawberries was 88.75%, 91.25%, and 75%, respectively. The evaluation method proposed in this paper improved detection accuracy by 10.79%, 7.76%, and 31.1%, respectively. Therefore, this method could contribute to the design of manipulators to further improve fruit production efficiency.

Sources of Clay‐Rich Sediment in Eberswalde Crater, Mars With Implications for Biopreservation Potential

AbstractThe Eberswalde crater paleolake system on Mars is home to an intact source‐to‐sink fluvial system where a catchment terminates in a river delta deposit. Through analysis of topographic and hyperspectral data from multiple geographic regions, we test the hypothesis that the smectites found within the Eberswalde deposit formed during the Noachian (i.e., >3.5 Ga) in the subsurface and are fluvial detritus transported during the Hesperian (i.e., 3.5–2.0 Ga). We find that the spectra from four pertinent regions (northwest Noachis Terra, Holden crater, and the Eberswalde watershed and delta deposit) are consistent with each other, and most consistent with Fe/Mg smectite clays, namely nontronite and saponite. This points to a detrital origin for the clays observed within Eberswalde crater. Analysis of the watershed also indicates that they likely formed within the subsurface, which previous studies suggest is the location most amenable to the preservation of evidence of past life on Mars. Furthermore, the presence of detrital clays, and not authigenic, indicates that the lake was ∼neutral‐pH and a potentially habitable environment. Taken together, these observations indicate that: (a) within Eberswalde crater is a delta deposit that formed in a habitable environment, (b) that the deposit is composed of minerals that formed during the most habitable period in Mars' history, and (c) those minerals formed in an environment with strong potential to preserve evidence of life. Lastly, the similarities between the Eberswalde fluvial system and the Jezero system make it a compelling site as a standard of comparison by NASA's Perseverance Rover.

Double Deep Q-Network for Hyperspectral Image Band Selection in Land Cover Classification Applications

Hyperspectral data usually consists of hundreds of narrow spectral bands and provides more detailed spectral characteristics compared to commonly used multispectral data in remote sensing applications. However, highly correlated spectral bands in hyperspectral data lead to computational complexity, which limits many applications or traditional methods when applied to hyperspectral data. The dimensionality reduction of hyperspectral data becomes one of the most important pre-processing steps in hyperspectral data analysis. Recently, deep reinforcement learning (DRL) has been introduced to hyperspectral data band selection (BS); however, the current DRL methods for hyperspectral data BS simply remove redundant bands, lack the significance analysis for the selected bands, and the reward mechanisms used in DRL only take basic forms in general. In this paper, a new reward mechanism strategy has been proposed, and Double Deep Q-Network (DDQN) is introduced during BS using DRL to improve the network stabilities and avoid local optimum. To verify the effect of the proposed BS method, land cover classification experiments were designed and carried out to analyze and compare the proposed method with other BS methods. In the land cover classification experiments, the overall accuracy (OA) of the proposed method can reach 98.37%, the average accuracy (AA) is 95.63%, the kappa coefficient (Kappa) is 97.87%. Overall, the proposed method is superior to other BS methods. Experiments have also shown that the proposed method works not only for airborne hyperspectral data (AVIRIS and HYDICE), but also for hyperspectral satellite data, such as PRISMA data. When hyperspectral data is applied to similar applications, the proposed BS method could be a candidate for the BS preprocessing options.

Independent component analysis of hyperspectral data measured from overlapping latent fingermarks: Forensic potential of independent component images

Overlapping fingermark images are sometimes discarded because fingermark collation for the individual fingermarks is difficult. Fluorescence hyperspectral data (HSD) measured using the models of double overlapping fingermarks obtained under the excitation of a high-power, continuous wave, green laser is suitable for obtaining individual fingermark images. However, there are limitations such as the problems on each spectrum of the individual fingermark and the forensic value of the obtained images. In this study, independent component analysis (ICA) was applied to the fluorescence HSD obtained from the models of doubly overlapping fingermarks, to obtain independent component (IC) spectra and the corresponding IC images. Forensic value of the obtained IC images was examined, considering the possibility of fingermark collation in masked fashion to the model sample information. The IC images obtained from the HSD had enough potential to enable extracting twelve minutiae required for fingermark collation if the image quality was good.

Hyperspectral Imaging for Mobile Robot Navigation.

The article presents the application of a hyperspectral camera in mobile robot navigation. Hyperspectral cameras are imaging systems that can capture a wide range of electromagnetic spectra. This feature allows them to detect a broader range of colors and features than traditional cameras and to perceive the environment more accurately. Several surface types, such as mud, can be challenging to detect using an RGB camera. In our system, the hyperspectral camera is used for ground recognition (e.g., grass, bumpy road, asphalt). Traditional global path planning methods take the shortest path length as the optimization objective. We propose an improved A* algorithm to generate the collision-free path. Semantic information makes it possible to plan a feasible and safe path in a complex off-road environment, taking traveling time as the optimization objective. We presented the results of the experiments for data collected in a natural environment. An important novelty of this paper is using a modified nearest neighbor method for hyperspectral data analysis and then using the data for path planning tasks in the same work. Using the nearest neighbor method allows us to adjust the robotic system much faster than using neural networks. As our system is continuously evolving, we intend to examine the performance of the vehicle on various road surfaces, which is why we sought to create a classification system that does not require a prolonged learning process. In our paper, we aimed to demonstrate that the incorporation of a hyperspectral camera can not only enhance route planning but also aid in the determination of parameters such as speed and acceleration.

Spectral mixture analysis of AVIRIS-NG hyperspectral data for material identification and classification for the part of Kolkata city

Identification and extraction of material in the urban scenario are always challenging due to the presence of different heterogeneous materials. Airborne hyperspectral sensors with the advantage of both high spectral and spatial resolution can be used in such areas for the identification and mapping of urban materials. Manual extraction of the endmembers in highly dense heterogeneous urban regions is a very difficult task to achieve, but with automated extraction techniques, it yields promising results. To counter the mixed pixels problem in heterogeneous urban areas automated endmember extraction techniques like Endmember Average RMS (EAR), Minimum Average Spectral Angle (MASA), and Count Based Endmember Selection (CoB) was used for pure endmember identification in the densely populated urban area of Kolkata region using AVIRIS NG imagery. Spectral Unmixing Multiple Endmember Spectral Mixture Analysis (MESMA), Linear Spectral Unmixing (LSU), and Mixture Tuned Matched Filtering (MTMF) was performed on AVIRIS NG data using the extracted endmembers of different urban materials. The rule-based classification was performed on abundance images obtained from MESMA, MTMF, and LSU techniques for comparative analysis. Classification using MESMA abundance images outperformed in mapping mixed urban environments with an overall accuracy (OA) of 88.6 % followed by the rule-based classified output of MTMF and LSU with OA of 75.27% and 72.45% respectively.

Identification of Defective Maize Seeds Using Hyperspectral Imaging Combined with Deep Learning

Seed quality affects crop yield and the quality of agricultural products, and traditional identification methods are time-consuming, complex, and irreversibly destructive. This study aims to establish a fast, non-destructive, and effective approach for defect detection in maize seeds based on hyperspectral imaging (HSI) technology combined with deep learning. Raw spectra collected from maize seeds (200 each healthy and worm-eaten) were pre-processed using detrending (DE) and multiple scattering correction (MSC) to highlight the spectral differences between samples. A convolutional neural network architecture (CNN-FES) based on a feature selection mechanism was proposed according to the importance of wavelength in the target classification task. The results show that the subset of 24 feature wavelengths selected by the proposed CNN-FES can capture important feature information in the spectral data more effectively than the conventional successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS) algorithms. In addition, a convolutional neural network architecture (CNN-ATM) based on an attentional classification mechanism was designed for one-dimensional spectral data classification and compared with three commonly used machine learning methods, linear discriminant analysis (LDA), random forest (RF), and support vector machine (SVM). The results show that the classification performance of the designed CNN-ATM on the full wavelength does not differ much from the above three methods, and the classification accuracy is above 90% on both the training and test sets. Meanwhile, the accuracy, sensitivity, and specificity of CNN-ATM based on feature wavelength modeling can reach up to 97.50%, 98.28%, and 96.77% at the highest, respectively. The study shows that hyperspectral imaging-based defect detection of maize seed is feasible and effective, and the proposed method has great potential for the processing and analysis of complex hyperspectral data.

Knowledge distillation: A novel approach for deep feature selection

High dimensional data in hyperspectral remote sensing leads to computational, analytical, and storage complexities. Dimensionality reduction serves as an efficient tool to remove redundant, irrelevant, and highly correlated features. Recently, deep learning approaches have received remarkable progress in hyperspectral data analysis. In this paper, a new end-to-end deep learning framework based on a teacher-student network inspired by knowledge distillation is proposed for deep feature selection. Initially, a complicated teacher deep neural network is employed on complex high dimensional data to learn its corresponding best low dimensional representation. Then, the knowledge from the network is transferred to a simple student network that performs feature selection. Hence, it eventually leads to deep neural network compression which is of prime concern in hyperspectral remote sensing. Limited studies have been carried out to explore the benefits of knowledge distillation on hyperspectral data. The proposed method could be employed to choose deep features for both supervised and unsupervised tasks. Experimental results reveal the performance of the proposed scheme using limited features. In comparison to 1D and simple autoencoder models, the 2D model based on convolutional autoencoder delivers greater classification accuracies, with a classification accuracy value of 96.15% for the Indian Pines dataset and 97.82% for the Pavia University dataset. A similar trend is reported with unsupervised learning as well. Furthermore, the proposed model has a low degree of sensitivity to parameter selection.

A novel strategy for classifying spectral-spatial shallow and deep hyperspectral image features using 1D-EWT and 3D-CNN

Hyperspectral images (HIs) are used in diverse disciplines, such as resource handling, land cover analysis, food science, anomaly detection, and precision agriculture. Researchers have been working on a number of visual processing and machine intelligence algorithms to handle this type of data as efficiently as feasible. Deep learning approaches have advanced significantly in the field of machine vision, which is also having a big impact on the analysis of hyperspectral data. To increase its discriminative potential for HI classification, this work suggests a powerful 3D-CNNs (Convolutional Neural Networks) architecture, where the shallow features extracted using 1D-EWT (Empirical Wavelet Transform) are served as input, and the ultimate output of the CNN are projected class-related outcomes. The framework is known as PEC, where P stands for PCA, E for 1D-EWT, and C for 3D-CNN model. Prior to features extraction, the HI undergoes spectral dimension reduction via Principal Component Anaalysis (PCA). To forecast segmentation for a volumetric area of a 3D HI sample, 3D CNNs use 3D convolutional kernels. The usage of more parameters by these CNNs means that the ability to employ interslice context that can boost performance. The CNN model’s parameters are optimised using a limited training set. The newly proposed PEC framework achieves a considerable overall accuracy of 99.58% and 99.94% (percentage increase of 0.08% and 0.54%) and Kappa score of 99.51% and 99.92 (percentage increase of 1.5% and 0.8%) compared to similar approaches for two real-world data sets IP and PU respectively with 30% training samples.

Two-Dimensional Clustering of Spectral Changes for the Interpretation of Raman Hyperspectra.

Two-dimensional correlation spectroscopy (2D-COS) is a technique that permits the examination of synchronous and asynchronous changes present in hyperspectral data. It produces two-dimensional correlation coefficient maps that represent the mutually correlated changes occurring at all Raman wavenumbers during an implemented perturbation. To focus our analysis on clusters of wavenumbers that tend to change together, we apply a k-means clustering to the wavenumber profiles in the perturbation domain decomposition of the two-dimensional correlation coefficient map. These profiles (or trends) reflect peak intensity changes as a function of the perturbation. We then plot the co-occurrences of cluster members two-dimensionally in a manner analogous to a two-dimensional correlation coefficient map. Because wavenumber profiles are clustered based on their similarity, two-dimensional cluster member spectra reveal which Raman peaks change in a similar manner, rather than how much they are correlated. Furthermore, clustering produces a discrete partitioning of the wavenumbers, thus a two-dimensional cluster member spectrum exhibits a discrete presentation of related Raman peaks as opposed to the more continuous representations in a two-dimensional correlation coefficient map. We demonstrate first the basic principles of the technique with the aid of synthetic data. We then apply it to Raman spectra obtained from a polystyrene perchlorate model system followed by Raman spectra from mammalian cells fixed with different percentages of methanol. Both data sets were designed to produce differential changes in sample components. In both cases, all the peaks pertaining to a given component should then change in a similar manner. We observed that component-based profile clustering did occur for polystyrene and perchlorate in the model system and lipids, nucleic acids, and proteins in the mammalian cell example. This confirmed that the method can translate to "real world" samples. We contrast these results with two-dimensional correlation spectroscopy results. To supplement interpretation, we present the cluster-segmented mean spectrum of the hyperspectral data. Overall, this technique is expected to be a valuable adjunct to two-dimensional correlation spectroscopy to further facilitate hyperspectral data interpretation and analysis.

Regional soil organic carbon prediction models based on a multivariate analysis of the Mid-infrared hyperspectral data in the middle Indo-Gangetic plains of India

Soil organic carbon (SOC) sequestration provides an opportunity to mitigate climate change impacts, since soils are the largest store of terrestrial carbon. Accurate estimates of SOC content across landscapes are therefore important to monitor and manage efficiently these SOC stocks. Mid-infrared (MIR) spectroscopy has been increasingly applied as a rapid, cost-effective, and accurate method for predictive soil analysis. This study assessed the performance of MIR spectroscopy for SOC prediction at a regional scale in the Indo-Gangetic plains, 280 soil samples were collected covering Inceptisols, Entisols and Alfisols and their spectra recorded. Five preprocessing techniques ((absorbance, multiplicative scatter correction (MSC), standard normal variate (SNV), Savitzky–Golay smoothing first derivative (SG-FD) and Savitzky–Golay smoothing second derivative (SG-SD)) and four multivariate methods (partial least-squares regression (PLSR), random forest (RF), support vector regression (SVR) and multivariate adaptive regression splines (MARS)) were evaluated to predict SOC from MIR spectra. The considerable prediction accuracy and robustness were achieved using the PLSR model (RV2 = 0.78, RMSEV = 0.04, and RPDV = 2.07), RF model (RV2 = 0.65, RMSEV = 0.09, and RPDV = 1.01), SVR model (RV2 = 0.65, RMSEV = 0.09, and RPDV = 1.12), and MARS model (RV2 = 0.67, RMSEV = 0.09, and RPDV = 1.20). Findings from this study identified the reliability of SOC determinations by examining how preprocessing techniques and multivariate methods affect spectral analyses.

Semi-Blind Source Separation with Learned Constraints

Blind source separation (BSS) algorithms are unsupervised methods, which are the cornerstone of hyperspectral data analysis by allowing for physically meaningful data decompositions. BSS problems being ill-posed, the resolution requires efficient regularization schemes to better distinguish between the sources and yield interpretable solutions. For that purpose, we investigate a semi-supervised source separation approach in which we combine a projected alternating least-square algorithm with a learning-based regularization scheme. In this article, we focus on constraining the mixing matrix to belong to a learned manifold by making use of generative models. Altogether, we show that this allows for an innovative BSS algorithm, with improved accuracy, which provides physically interpretable solutions. The proposed method, coined sGMCA, is tested on realistic hyperspectral astrophysical data in challenging scenarios involving strong noise, highly correlated spectra and unbalanced sources. The results highlight the significant benefit of the learned prior to reduce the leakages between the sources, which allows an overall better disentanglement.

Dual-Branch-AttentionNet: A Novel Deep-Learning-Based Spatial-Spectral Attention Methodology for Hyperspectral Data Analysis

Recently, deep learning-based classification approaches have made great progress and now dominate a wide range of applications, thanks to their Herculean discriminative feature learning ability. Despite their success, for hyperspectral data analysis, these deep learning based techniques tend to suffer computationally as the magnitude of the data soars. This is mainly because the hyperspectral imagery (HSI) data are multidimensional, as well as giving equal importance to the large amount of temporal and spatial information in the HSI data, despite the redundancy of information in the temporal and spatial domains. Consequently, in literature, this equal information emphasis has proven to affect the classification efficacy negatively in addition to increasing the computational time. As a result, this paper proposes a novel dual branch spatial-spectral attention based classification methodology that is computationally cheap and capable of selectively accentuating cardinal spatial and spectral features while suppressing less useful ones. The theory of feature extraction with 3D-convolutions alongside a gated mechanism for feature weighting using bi-directional long short-term memory is used as a spectral attention mechanism in this architecture. In addition, a union of 3D convolutional neural network (3D-CNN) and a residual network oriented spatial window-based attention mechanism is proposed in this work. To validate the efficacy of our proposed technique, the features collected from these spatial and spectral attention pipelines are transferred to a feed-forward neural network (FNN) for supervised pixel-wise classification of HSI data. The suggested spatial-spectral attention based hyperspectral data analysis and image classification methodology outperform other spatial-only, spectral-only, and spatial-spectral feature extraction based hyperspectral image classification methodologies when compared, according to experimental results.

Visible-NIR spectral characteristics and grade inversion model of skarn-type iron ore

Iron ore grade determination is a key step in mining production, and its efficiency has a significant impact on mining efficiency and profitability. At present, the most commonly used method for ore grade determination is chemical analysis, which not only is expensive but also has a long assay cycle and lag effect relative to the ore allocation process and cannot effectively reduce the loss depletion rate of ore mining. Therefore, in situ iron ore grade data determination technology based on visible and near-infrared hyperspectral data analysis is an effective way to solve the above problems, but the poor quality of raw hyperspectral data and low accuracy of the ore grade inversion model are still one of the bottlenecks in the application of this technology. In this paper, 225 ore samples were collected from Hongling skarn-type iron ore in Chifeng, Inner Mongolia Autonomous Region, China. The visible and near-infrared hyperspectral data of the sample set and the grade data obtained from chemical analysis were used as data sources. The raw spectral data were first smoothed, and then the smoothed spectral data were processed using two algorithms, continuum removal (CR) and standard normal variate transform (SNV). Then, two dimensionality reduction algorithms, principal component analysis (PCA) and genetic algorithm (GA), were used to reduce the dimensionality of the spectral data before and after preprocessing, and the data sources were obtained after the four different preprocessing combination algorithms were processed. Finally, several quantitative inversion models were established based on the random forest algorithm and the backpropagation neural network (BPNN) algorithm. The stability, accuracy and credibility of the models were evaluated by four metrics: the coefficient of determination (R2), root mean square error (RMSE), residual prediction deviation(RPD) and mean relative error (MRE). The results showed that the quantitative inversion model (SNV-PCA-BPNN) established based on the BP neural network algorithm using the data processed by SNV and PCA for dimensionality reduction was the best, with an R2 of 0.99, an RMSE of 0.0048, an MRE of 1.10% and a RPD of 37.70, providing an effective method for the high-precision inversion of iron ore grades.

A machine learning toolkit for CRISM image analysis

Hyperspectral images collected by remote sensing have played a significant role in the discovery of aqueous alteration minerals, which in turn have important implications for our understanding of the changing habitability on Mars. Traditional spectral analyzes based on summary parameters have been helpful in converting hyperspectral cubes into readily visualizable three channel maps highlighting high-level mineral composition of the Martian terrain. These maps have been used as a starting point in the search for specific mineral phases in images. Although the amount of labor needed to verify the presence of a mineral phase in an image is quite limited for phases that emerge with high abundance, manual processing becomes laborious when the task involves determining the spatial extent of detected phases or identifying small outcrops of secondary phases that appear in only a few pixels within an image. Thanks to extensive use of remote sensing data and rover expeditions, significant domain knowledge has accumulated over the years about mineral composition of several regions of interest on Mars, which allow us to collect reliable labeled data required to train machine learning algorithms. In this study we demonstrate the utility of machine learning in two essential tasks for hyperspectral data analysis: nonlinear noise removal and mineral classification. We develop a simple yet effective hierarchical Bayesian model for estimating distributions of spectral patterns and extensively validate this model for mineral classification on several test images. Our results demonstrate that machine learning can be highly effective in exposing tiny outcrops of specific phases in orbital data that are not uncovered by traditional spectral analysis. We package implemented scripts, documentation illustrating use cases, and pixel-scale training data collected from dozens of well-characterized images into a new toolkit. We hope that this new toolkit will provide advanced and effective processing tools and improve community’s ability to map compositional units in remote sensing data quickly, accurately, and at scale.

Editorial Paper for the Special Issue “Algorithms in Hyperspectral Data Analysis”

This Special Issue contains four papers focused on hyperspectral data analysis [...]

A review of hyperspectral image analysis techniques for plant disease detection and identif ication.

Plant diseases cause signif icant economic losses in agriculture around the world. Early detection, quantif ication and identif ication of plant diseases are crucial for targeted application of plant protection measures in crop production. Recently, intensive research has been conducted to develop innovative methods for diagnosing plant diseases based on hyperspectral technologies. The analysis of the ref lection spectrum of plant tissue makes it possible to classify healthy and diseased plants, assess the severity of the disease, differentiate the types of pathogens, and identify the symptoms of biotic stresses at early stages, including during the incubation period, when the symptoms are not visible to the human eye. This review describes the basic principles of hyperspectral measurements and different types of available hyperspectral sensors. Possible applications of hyperspectral sensors and platforms on different scales for diseases diagnosis are discussed and evaluated. Hyperspectral analysis is a new subject that combines optical spectroscopy and image analysis methods, which make it possible to simultaneously evaluate both physiological and morphological parameters. The review describes the main steps of the hyperspectral data analysis process: image acquisition and prepro cessing; data extraction and processing; modeling and analysis of data. The algorithms and methods applied at each step are mainly summarized. Further, the main areas of application of hyperspectral sensors in the diagnosis of plant diseases are considered, such as detection, differentiation and identif ication of diseases, estimation of disease severity, phenotyping of disease resistance of genotypes. A comprehensive review of scientif ic publications on the diagnosis of plant diseases highlights the benef its of hyperspectral technologies in investigating interactions between plants and pathogens at various measurement scales. Despite the encouraging progress made over the past few decades in monitoring plant diseases based on hyperspectral technologies, some technical problems that make these methods diff icult to apply in practice remain unresolved. The review is concluded with an overview of problems and prospects of using new technologies in agricultural production

Spectral and image analysis of hyperspectral data for internal and external quality assessment of peach fruit

Hyperspectral imaging was attempted to evaluate the internal and external quality of ‘Feicheng’ peach by providing the spectral and spatial data simultaneously. Mask-image was created from hyperspectral image at 810 nm and used to segment the fruit region where the average spectrum, after area normalization, was obtained for soluble solids content (SSC) and firmness evaluation. Pixel size and area were used for diameter and weight estimation. Then effective wavelengths were selected by competitive adaptive reweighted sampling (CARS) and random frog (RF), and employed to develop multiple linear regression (MLR) models. The more effective prediction performances emerged from CARS-MLR model withRV2 = 0.841, RMSEV = 0.546, RPD = 2.51 for SSC andRV2 = 0.826, RMSEV = 1.008, RPD = 2.401 for firmness, followed by creating pixel-wise and object-wise visualization maps for quantifying SSC and firmness. Furthermore, peach diameter was estimated by calculating the minimum bounding rectangle with an average percentage error of 1.01 %, and the MLR model forweightpredictionachieveda good performance ofRV2 = 0.957, RMSEV = 9.203, and RPD = 4.819. The overall results showed that hyperspectral imaging could be used as an effective and non-destructive tool for evaluating the internal and external quality attributes of ‘Feicheng’ peach, and provided a holistic approach to develop online grading systems for quality tiers identification.

Comparison of field survey and remote sensing techniques for detection of bark beetle-infested trees

Detection in the early phase of bark beetle infestation is a vital task for proactive management strategies, as practiced in most Central European forests, to minimize economic losses due to bark beetle infestation and to mitigate their further spreading. For this work, remote sensing methods are coming to be in great demand as an objective approach to enable monitoring bark beetle infestation even at individual tree level.This case study monitored bark beetle (Ips typographus) activity at local level in Norway spruce forest in the Czech Republic. The main aim of this study was to compare the remote sensing methods against classical field survey conducted by forest workers in detecting newly infested trees.To compare these two methods, an extensive field and aerial campaign was conducted in the southern part of the Czech Republic during 2020. Bark beetle infestation was monitored by traditional methods (i.e. field survey) on a weekly basis from mid-March to mid-September. During the same period, aerial scans were performed once per month (seven in total) using a CASI-1500 hyperspectral sensor (visible and near-infrared, 400–1000 nm) with spatial resolution of 0.5 m. This work mapped transition from healthy up to red attack of 75 Norway spruce trees that were infested during the same week. The same number of healthy trees were added to the data set for hyperspectral data analysis. Both groups were analysed by vegetation indices, with emphasis on effect caused in the canopy by bark beetles.The success rate for bark beetle detection is always associated with acquisition time. In order to define the optimal time for data acquisition, we employed a phenology model for I. typographus (RITY 2.0) to take into consideration bark beetle development.The results of the experiment showed that classic field survey detected infested trees earlier than did analysis using remote sensing data from the visible and near-infrared region. The difference was 23 days for the most successful indices (i.e. REIP, PRI, and ANCB650–720) in our test. Nevertheless, both methods detected the infested trees within 6 weeks after infestation, which is the recommended period for taking measures to prevent bark beetles from spreading further, and thus hyperspectral imagery can be used as a valid information source for bark beetle detection.