Multispectral Camera Research Articles

Lucy L′Ralph In-flight Calibration and Results at (152830) Dinkinesh

Abstract The L’Ralph instrument is a key component of NASA’s Lucy mission, intended to provide spectral image data of multiple Jupiter Trojans. The instrument operates from ∼0.35 to 4 μm using two focal plane assemblies: a 350–950 nm multispectral imager, Multi-spectral Visible Imaging Camera (MVIC), and a 0.97–4 μm imaging spectrometer, Linear Etalon Imaging Spectral Array (LEISA). Instrument calibration was established through ground testing before launch and has been monitored during cruise utilizing internal calibration sources and stellar targets. In-flight data have shown that the instrument thermal performance is exceeding expectations, allowing for early updates to LEISA radiometric and pointing calibrations. MVIC radiometric performance remains stable more than 3 yr since launch. The serendipitous identification of a new flyby target, (152830) Dinkinesh, allowed testing of instrument performance and interleaved LEISA and MVIC acquisitions on an asteroid target. Both MVIC and LEISA obtained data of Dinkinesh and its moon, Selam, demonstrating that they show good spectral agreement with an S- or Sq-type asteroid, along with evidence of a 3 μm absorption feature.

Non-destructive Detection of Growth and Quality in Basil Seedlings Grown in a Plant Factory

Background and objective: Basil is one of the high-value crops cultivated in plant factories. The production of uniform and healthy seedlings has a direct impact on the yield and quality of the final harvest. The objectives of this study were; (1) to ascertain whether non-destructive detecting parameters [projected canopy area (PCA) and vegetation indices (VIs)] could predict changes in growth and quality of basil seedlings, and (2) to determine the feasibility of grading basil seedlings based on PCA to establish a baseline for stable yields after transplanting.Methods: Basil seedlings were grown in 2- and 3-day irrigation cycles for 25 days after sowing, and the growth parameters and image analysis parameters using a multispectral camera were examined at regular intervals. The correlations between growth parameters and PCA/VIs were investigated to detect growth and quality of basil seedlings. At the time of transplanting, the basil seedlings were classified into grades A-D based on their PCA values, and the growth and yield after transplanting of basil seedlings in each grade were evaluated.Results: The basil seedlings in the 3-day irrigation treatment showed higher growth, and the correlation between the PCA values and the leaf area and fresh weight resulted in a coefficient of determination greater than 0.93. Among the VIs, the VARI, GI, and NGRDI were correlated with the growth and quality with coefficients of determination greater than 0.6. And, the growth and yield after transplanting were dependent on the seedling grade based on PCA values at the time of transplanting.Conclusion: This study confirmed that it is possible to predict the growth and quality of basil seedlings using non-destructive image analysis, and that the grading criteria for basil seedlings that can be expected to produce stable yields after transplanting can be determined using image analysis.

Comparative Analysis of Row Gradient and BRDF Corrections in UAV’s Multispectral Camera under Varied Cloud Cover

Comparative Analysis of Row Gradient and BRDF Corrections in UAV’s Multispectral Camera under Varied Cloud Cover

A Rice Leaf Area Index Monitoring Method Based on the Fusion of Data from RGB Camera and Multi-Spectral Camera on an Inspection Robot

Automated monitoring of the rice leaf area index (LAI) using near-ground sensing platforms, such as inspection robots, is essential for modern rice precision management. These robots are equipped with various complementary sensors, where specific sensor capabilities partially overlap to provide redundancy and enhanced reliability. Thus, leveraging multi-sensor fusion technology to improve the accuracy of LAI monitoring has become a crucial research focus. This study presents a rice LAI monitoring model based on the fused data from RGB and multi-spectral cameras with an ensemble learning algorithm. The results indicate that the estimation accuracy of the rice LAI monitoring model is effectively improved by fusing the vegetation index and textures from RGB and multi-spectral sensors. The model based on the LightGBM regression algorithm has the most improvement in accuracy, with a coefficient of determination (R2) of 0.892, a root mean square error (RMSE) of 0.270, and a mean absolute error (MAE) of 0.160. Furthermore, the accuracy of LAI estimation in the jointing stage is higher than in the heading stage. At the jointing stage, both LightGBM based on optimal RGB image features and Random Forest based on fused features achieved an R2 of 0.95. This study provides a technical reference for automatically monitoring rice growth parameters in the field using inspection robots.

A 3D Surface Reconstruction Pipeline for Plant Phenotyping

Plant phenotyping plays a crucial role in crop science and plant breeding. However, traditional methods often involve time-consuming and manual observations. Therefore, it is essential to develop automated, sensor-driven techniques that can provide objective and rapid information. Various methods rely on camera systems, including RGB, multi-spectral, and hyper-spectral cameras, which offer valuable insights into plant physiology. In recent years, 3D sensing systems such as laser scanners have gained popularity due to their ability to capture structural plant parameters that are difficult to obtain using spectral sensors. Unlike images, point clouds are not structured and require pre-processing steps to extract precise information and handle noise or missing points. One approach is to generate mesh-based surface representations using triangulation. A key challenge in the 3D surface reconstruction of plants is the pre-processing of point clouds, which involves removing non-plant noise from the scene, segmenting point clouds from populations to individual plants, and further dividing individual plants into their respective organs. In this study, we will not focus on the segmentation aspect but rather on the other pre-processing steps, like denoising parameters, which depend on the data type. We present an automated pipeline for converting high-resolution point clouds into surface models of plants. The pipeline incorporates additional pre-processing steps such as outlier removal, denoising, and subsampling to ensure the accuracy and quality of the reconstructed surfaces. Data were collected using three different sensors: a handheld scanner, a terrestrial laser scanner (TLS), and a mobile mapping platform, under varying conditions from controlled laboratory environments to complex field settings. The investigation includes five different plant species, each with distinct characteristics, to demonstrate the potential of the pipeline. In a next step, phenotypic traits such as leaf area, leaf area index (LAI), and leaf angle distribution (LAD) were calculated to further illustrate the pipeline’s potential and effectiveness. The pipeline is based on the Open3D framework and is available open source.

Normalized Difference Vegetation Index Prediction for Blueberry Plant Health from RGB Images: A Clustering and Deep Learning Approach

In precision agriculture (PA), monitoring individual plant health is crucial for optimizing yields and minimizing resources. The normalized difference vegetation index (NDVI), a widely used health indicator, typically relies on expensive multispectral cameras. This study introduces a method for predicting the NDVI of blueberry plants using RGB images and deep learning, offering a cost-effective alternative. To identify individual plant bushes, K-means and Gaussian Mixture Model (GMM) clustering were applied. RGB images were transformed into the HSL (hue, saturation, lightness) color space, and the hue channel was constrained using percentiles to exclude extreme values while preserving relevant plant hues. Further refinement was achieved through adaptive pixel-to-pixel distance filtering combined with the Davies–Bouldin Index (DBI) to eliminate pixels deviating from the compact cluster structure. This enhanced clustering accuracy and enabled precise NDVI calculations. A convolutional neural network (CNN) was trained and tested to predict NDVI-based health indices. The model achieved strong performance with mean squared losses of 0.0074, 0.0044, and 0.0021 for training, validation, and test datasets, respectively. The test dataset also yielded a mean absolute error of 0.0369 and a mean percentage error of 4.5851. These results demonstrate the NDVI prediction method’s potential for cost-effective, real-time plant health assessment, particularly in agrobotics.

Estimation of Water Interception of Winter Wheat Canopy Under Sprinkler Irrigation Using UAV Image Data

Canopy water interception is a key parameter to study the hydrological cycle, water utilization efficiency, and energy balance in terrestrial ecosystems. Especially in sprinkler-irrigated farmlands, the canopy interception further influences field energy distribution and microclimate, then plant transpiration and photosynthesis, and finally crop yield and water productivity. To reduce the field damage and increase measurement accuracy under traditional canopy water interception measurement, UAVs equipped with multispectral cameras were used to extract in situ crop canopy information. Based on the correlation coefficient (r), vegetative indices that are sensitive to canopy interception were screened out and then used to develop canopy interception models using linear regression (LR), random forest (RF), and back propagation neural network (BPNN) methods, and lastly these models were evaluated by root mean square error (RMSE) and mean relative error (MRE). Results show the canopy water interception is first closely related to relative normalized difference vegetation index (R△NDVI) with r of 0.76. The first seven indices with r from high to low are R△NDVI, reflectance values of the blue band (Blue), reflectance values of the near-infrared band (Nir), three-band gradient difference vegetation index (TGDVI), difference vegetation index (DVI), normalized difference red edge index (NDRE), and soil-adjusted vegetation index (SAVI) were chosen to develop canopy interception models. All the developed linear regression models based on three indices (R△NDVI, Blue, and NDRE), the RF model, and the BPNN model performed well in canopy water interception estimation (r: 0.53–0.76, RMSE: 0.18–0.27 mm, MRE: 21–27%) when the interception is less than 1.4 mm. The three methods underestimate the canopy interception by 18–32% when interception is higher than 1.4 mm, which could be due to the saturation of NDVI when leaf area index is higher than 4.0. Because linear regression is easy to perform, then the linear regression method with NDVI is recommended for canopy interception estimation of sprinkler-irrigated winter wheat. The proposed linear regression method and the R△NDVI index can further be used to estimate the canopy water interception of other plants as well as forest canopy.

Assessing the Feasibility of Using Remote Sensing Data and Vegetation Indices in the Estimation of Land Subject to Consolidation.

The values of vegetation indices can provide a new source of data for use in the estimation of land to be consolidated. The results of research work carried out so far indicate a significant advantage of low-volume imaging over satellite methods when it comes to calculating vegetation index values. This paper analyses multispectral images for the areas of selected croplands acquired via the Sentinel-2 satellite and an unmanned aerial vehicle (UAV) equipped with a multispectral camera. The research work consisted of evaluating NDVI (Normalised Difference Vegetation Index) and SAVI (Soil Adjusted Vegetation Index) values depending on the type of crop grown, the size of the cultivated area and the method of data acquisition. The data obtained were used to assess their potential use in the estimation of land to be consolidated. The effect of land consolidation is primarily to create more favourable living conditions and increase agricultural productivity. The results of the study showed that it would be preferable to use multispectral images acquired using UAVs rather than those from Sentinel satellites. This is due to the insufficient resolution of the satellite data, the correlation of NDVI and SAVI values at only a satisfactory level and the low accuracy of the data obtained for small registered plots of land.

Comparative Analysis of RGB and Multispectral UAV Image Data for Leaf Area Index Estimation of Sweet Potato

This study aims to evaluate whether the utilization of a broader array of image types and the extraction of plant canopy using RGB and multispectral cameras mounted on unmanned aerial vehicles (UAVs) can improve leaf area index (LAI) estimation accuracy. The accuracy of LAI estimation for sweet potato was compared across different model types—mono-regression models based on a single image, multivariate regression models based on a single type of image, and multivariate regression models based on multiple image types—by employing 10 types of images based on color, morphological indices, vegetation indices (VIs) from RGB cameras, reflectance, VIs from multispectral cameras, and each of their canopy part image. For each regression model, we compared the estimation accuracy based on the whole image with that based on the canopy part image. For the mono-regression model, EVI2 from the whole image exhibited the lowest test root mean squared error (RMSE) of 0.403, which is attributed to EVI2’s capacity to mitigate the effects of spectral saturation. Contrary to the models with multiple image types that demonstrated improved accuracy, the multivariate regression models based on a single image type did not enhance estimation accuracy; this shows that the use of multiple image types improves the LAI estimation accuracy owing to the synergy of different spectral information from various image types. Plant canopy extraction did not enhance the estimation accuracy in mono-regression models or multivariate regression models based on a single image type; however, it improved the accuracy of multivariate regression models based on multiple image types, provided the image types were appropriately combined. The highest estimation accuracy was achieved by a partial least squares regression model based on VI (from the RGB camera) and reflectance (from the multispectral camera) of the canopy part (test R2 = 0.887, test RMSE = 0.351), suggesting that this is an effective approach for LAI estimation in sweet potato (cultivar Beniharuka). Overall, this study demonstrates that an optimal combination of image type and plant canopy extraction can enhance LAI estimation accuracy in UAV-based monitoring.

Evaluation of Vegetation Health in the PT X Reclamation Area Using the NDVI Method Based on Unmanned Aerial Vehicle (UAV) Multispectral Orthophoto Data

Abstract According to the International Energy Agency (IEA) in 2023, Indonesia became the third-largest coal-producing country globally after China and India. Indonesia’s coal production amounted to 775.2 million tons in 2023, and around 27% was used for domestic needs. Coal mining in Indonesia generally uses an open-pit mining system, which impacts changes in land cover, landscape form, and the loss of flora and fauna. To dismiss this stigma, good mining practices must be applied in every process to ensure the balance of post-mining land’s natural and social functions. One effort to restore the condition of post-mining land is revegetation. Revegetation activities on ex-mining land require much time, money, and energy, so the planted vegetation must be monitored and evaluated. The problem often encountered when using Landsat and Sentinel satellite image data is that clouds cover many areas because the island of Kalimantan has a tropical climate with high rainfall, humidity, and temperature throughout the year. Then, using multispectral aerial photography data is one solution for monitoring plant health. With the advancement of photogrammetric technology using multispectral cameras, it is possible to analyze the health of reclaimed plants using the NDVI method. The results of the NDVI analysis on reclaimed plants produced the highest index value of 0.877 and the lowest value of 0.013. With very healthy plant criteria covering 12.47 ha (71.5%), healthy plant criteria covering 4.95 ha (28.4%), and unhealthy criteria covering 0.02 ha (0.1%).

Observations of Uranus at High Phase Angle as Seen by New Horizons

We present flux measurements of Uranus observed at phase angles of 43.°9, 44.°0, and 52.°4 by the Multispectral Visible Imaging Camera on the New Horizons spacecraft during 2023, 2010, and 2019, respectively. New Horizons imaged Uranus at a distance of about 24–70 au (2023) in four color filters, with bandpasses of 400–550 nm, 540–700 nm, 780–975 nm, and 860–910 nm. High-phase-angle observations are of interest for studying the energy balance of Uranus, constraining the atmospheric scattering behavior, and understanding the planet as an analog for ice giant exoplanets. The new observations from New Horizons provide access to a wider wavelength range and different season compared with previous observations from the Voyager spacecraft. We performed aperture photometry on the New Horizons observations of Uranus to obtain its brightness in each photometric band. The photometry suggests that Uranus may be darker than predicted by a Lambertian phase curve in the blue and red filters. Comparison to simultaneous low-phase Hubble WFC3 and ground-based community-led observations indicates a lack of large-scale features at full phase that would introduce variation in the rotational light curve. The New Horizons reflectance in the blue (492 nm) and red (624 nm) filters does not exhibit statistically significant variation and is consistent with the expected error bars. These results place new constraints on the atmospheric model of Uranus and its reflectivity. The observations are analogous to those from future exoplanet direct-imaging missions, which will capture unresolved images of exoplanets at partial phases. These results will serve as a “ground truth” with which to interpret exo-ice-giant data.

Automatic Tissue Differentiation in Parotidectomy using Hyperspectral Imaging

Abstract In head and neck surgery, continuous intraoperative tissue differentiation is of great importance to avoid injury to sensitive structures such as nerves and vessels. Hyperspectral imaging (HSI) with neural network analysis could support the surgeon in tissue differentiation. A 3D Convolutional Neural Network with hyperspectral data in the range of 400 − 1000 nm is used in this work. The acquisition system consisted of two multispectral snapshot cameras creating a stereo-HSI-system. For the analysis, 27 images with annotations of glandular tissue, nerve, muscle, skin and vein in 18 patients undergoing parotidectomy are included. Three patients are removed for evaluation following the leave-one-subjectout principle. The remaining images are used for training, with the data randomly divided into a training group and a validation group. In the validation, an overall accuracy of 98.7% is achieved, indicating robust training. In the evaluation on the excluded patients, an overall accuracy of 83.4% has been achieved showing good detection and identification abilities. The results clearly show that it is possible to achieve robust intraoperative tissue differentiation using hyperspectral imaging. Especially the high sensitivity in parotid or nerve tissue is of clinical importance. It is interesting to note that vein was often confused with muscle. This requires further analysis and shows that a very good and comprehensive data basis is essential. This is a major challenge, especially in surgery.

Mapping Paleolacustrine Deposits with a UAV-borne Multispectral Camera: Implications for Future Drone Mapping on Mars

NASA’s Ingenuity Mars Helicopter has marked a new era in planetary exploration by employing unmanned aerial vehicles (UAVs) to enhance our understanding of planetary surfaces. This study evaluates the potential of UAVs for mapping Martian environments, with Lake Natron, Tanzania, serving as an analog for Martian paleolakes. During two field seasons (2023 January and July), we used a Phantom 4 Pro drone equipped with a MicaSense RedEdge-M multispectral camera, supplemented by in situ analysis using a TerraSpec Halo VNIR-SWIR spectrometer, to capture high-resolution imagery and spectral data. Almost all image processing and analysis, except for image mosaic and digital elevation model (DEM) generation, was performed using Python scripting. We benchmarked the onboard image processing capabilities using a Raspberry Pi 5 single-board computer. Processing steps include digital number (DN)-to-radiance conversion, assessment of the best radiance-to-reflectance conversion method, image mosaic creation, DEM generation, calculation of optimal band indices, and selection of the best classification technique. The research underscores Lake Natron’s diverse lithologies as a suitable analog site and demonstrates significant improvements in classification when normalized elevation data are incorporated with spectral index maps through unsupervised classification methods. The study also addresses challenges related to high-resolution image transmission and processing, advocating for advanced techniques such as image compression and low-power computational models. Additionally, it highlights computational and power limitations as key obstacles, suggesting that emerging technologies such as photonic computing and hybrid controllers could provide viable solutions. These findings emphasize the transformative potential of UAVs in planetary exploration while outlining key areas for future research and technological development.

Bentayga-I: Development of a Low-Cost and Open-Source Multispectral CubeSat for Marine Environment Monitoring and Prevention.

CubeSats have emerged as a promising alternative to satellite missions for studying remote areas where satellite data are scarce and insufficient, such as coastal and marine environments. However, their standard size and weight limitations make integrating remote sensing optical instruments challenging. This work presents the development of Bentayga-I, a CubeSat designed to validate PANDORA, a self-made, lightweight, cost-effective multispectral camera with interchangeable spectral optical filters, in near-space conditions. Its four selected spectral bands are relevant for ocean studies. Alongside the camera, Bentayga-I integrates a power system for short-time operation capacity; a thermal subsystem to maintain battery function; environmental sensors to monitor the CubeSat's internal and external conditions; and a communication subsystem to transmit acquired data to a ground station. The first helium balloon launch with B2Space proved that Bentayga-I electronics worked correctly in near-space environments. During this launch, the spectral capabilities of PANDORA alongside the spectrum were validated using a hyperspectral camera. Its scientific applicability was also tested by capturing images of coastal areas. A second launch is planned to further validate the multispectral camera in a real-world scenario. The integration of Bentayga-I and PANDORA presents promising results for future low-cost CubeSats missions.

In-vivo non-contact multispectral oral disease image dataset with segmentation

In imaging spectroscopy, gathering oral tissue spectral data from resected samples may not accurately represent tissue signatures due to time-dependent changes, blood loss, protein degeneration, and preservation chemicals. In-vivo spectral imaging is employed to address these limitations, but it poses challenges like device dimensions, tissue accessibility, and motion artifacts, impacting data quality and reliability. Our study publishes a dataset of spectral images focusing on oral diseases, addressing these challenges. We used a state-of-the-art multispectral camera, capturing images at 270*510 pixels resolution in 16 spectral bands (460 nm to 600 nm). The dataset includes 91 participants (15 healthy and 76 diseased), with multiple images per patient, totalling 243 spectral images. The dataset encompasses three oral health conditions: Oral Submucous Fibrosis (OSMF), Leukoplakia, and Oral Squamous Cell Carcinoma (OSCC). Detailed patient history records accompany each case. This publicly available oral health multispectral dataset has the potential to advance spectroscopy diagnosis. Integrating artificial intelligence with a comprehensive spectral signature repository holds promise for accurate disease analysis.

Vegetation Density Mapping of Urban Solid Waste Landfill Coverage Using Vegetation Indexes Obtained with UAV

Objective: The objective of this study is to map the vegetation cover of a municipal solid waste landfill using high-resolution vegetation index images obtained with UAV, as a tool for geotechnical monitoring. Theoretical Framework: The theoretical framework describes the state of the art of vegetation indexes, starting from the first proposal of NVDI by Rouse et al. (1973), through to the NDWI by McFeeters (1996). Additionally, it presents the method of representing radiometric variables, considering that these indexes operate based on the ratio between spectral bands, which are displayed within intervals ranging from -1 to +1. Method: The study used a UAV equipped with MAPIR multispectral camera to capture images of the landfill. NDVI and NDWI indexes were applied to assess vegetation health and coverage. Field control points determined with GNSS-RTK, along with protogrametry software, ensured mapping accuracy, creating high-resolution orhomosaics for monitoring and determining vegetation indexes. Results and Discussion: The results indicate variation in vegetation cover, revealing areas of exposed soil ang vegetation at different stages of vigor, from senescence to full growth. The NDVI and NDWI indexes highlighted regions with water stress and soil exposure, demonstrating the effectiveness of the method in environmental and geotechnical monitoring. Conclusions: It is concluded that the use of UAVs and vegetation indexes are effective for monitoring vegetation health and landfill stability, enanbling the identification of erosion areas and the need for vegetation cover restoration.

Prediction of Oil Palm Plantation Block Productivity Based On Canopy Area And Vegetation Index Using Multispectral Aerial Photographs

The demands of productivity and efficiency of oil palm plantations can be met by implementing Precision Farming (PF), where production analysis is carried out on a narrower area. Currently, yield evaluation or harvest results are carried out based on block area (25-100 Ha). The diverse information in one block is standardized, potentially eliminating detailed information that is useful for management. This study aims to develop an empirical estimation model of oil palm plantation productivity per harvest day through the canopy area approach and oil palm vegetation index with multispectral camera technology. The oil palm productivity estimation method is carried out by comparing the processing results using the Normalized Difference Vegetation Index and Atmospherically Resistant Vegetation Index transformation algorithms. The basis for estimation is per Harvesting Collection Point, namely an area of ± 0.3 ha. The results showed: 1) NDVI, NDRE are strongly and very strongly correlated with FFB production; 2) canopy area has a strong relationship to FFB production; 3) NDVI vegetation index values are strongly correlated with FFB production. The productivity estimation equation from NDRE value: y = 3621.9x – 1153.8, the productivity estimation equation based on canopy area: y = 0.5479x – 717.08, and the productivity estimation equation based on NDVI: y = 11.639x – 9709.2

Remote estimation of leaf nitrogen content, leaf area, and berry yield in wild blueberries

Nitrogen (N) fertilization is a major management requirement for wild blueberry fields. Its presence and estimation can be difficult given the perennial and heterogeneous nature of the plant, low N requirement, and residual N effects, resulting in the frequent over-application of N, excessive canopy growth, and resulting reduction in berry yields. Therefore, this study aimed to estimate nitrogen content and growth parameters using remote sensing approaches. Three trials were established in three commercial fields in Nova Scotia, Canada. An RCBD with 5 replicates and a plot size of 6 × 8 m with a 2 m buffer was used. Treatments consisted of 0, 20, 40, 60, and 100 kg N ha-1 of fertilizer. Using a DJI Matrice 300 UAV mounted with an RGB and a multispectral camera, aerial measurements were collected at 30 m altitude. Several field measurements including leaf nitrogen content (LNC), leaf area, floral bud numbers, stem height, and yield were conducted. Several vegetation indices (VIs) were computed for each plot, and correlation and regression analyses were conducted. Results indicated that treatments with high nitrogen rates had correspondingly high LAI measurements with the 60 kg ha-1 rate achieving the best growth parameters compared to the other treatments. LNC, LAI, and berry yield estimations using VIs [green leaf index (GLI), green red vegetation index (GRVI), and visible atmospheric red index (VARI)] produced significantly positive R2 values of 0.43, 0.48, and 0.30 respectively. Results from this study illustrated the potential of using VIs to estimate LNC, LAI, and berry yield parameters. It was established that the near-infrared VIs are the most effective in estimating differences in nitrogen rates, making them suitable for use in prescription maps for N fertilization applications.

Early detection of verticillium wilt in eggplant leaves by fusing five image channels: a deep learning approach

BackgroundAs one of the world’s most important vegetable crops, eggplant production is often severely affected by verticillium wilt, leading to significant declines in yield and quality. Traditional multispectral disease-imaging equipment is expensive and complicated to operate. Low-cost multispectral devices cannot capture images and cover less information. The traditional approach to early disease diagnosis involves using multispectral disease-imaging equipment in conjunction with machine learning technology. However, this approach has significant limitations in early disease diagnosis, including challenges such as high costs, complex operation, and low model performance.ResultsThe aim of this study was to combine low-cost multispectral cameras with deep learning technology to detect early stage Verticillium wilt in eggplant effectively. Using the Manual FS-3200T-10GE-NNC multispectral camera to perform multispectral imaging of the leaves of eggplant seedlings at the early infection stage, information fusion was performed on the collected multispectral images, and a five-channel image information fusion model was established. Image information fusion technology was combined with deep learning technology, among which the VGG16-triplet attention model performed the best, achieving a precision of 86.73% on the test set. Model validation on 48- and 72-hour data reached a precision of 75% and 82%, respectively, achieving an early diagnosis of Verticillium wilt. This highlighted the potential of multispectral cameras for early disease detection.ConclusionsIn this study, we successfully developed a method for the non-destructive detection of the early stages of eggplant wilt disease by combining multispectral imaging technology with deep learning algorithms. While ensuring high accuracy, this method significantly reduces the cost of experimental equipment. The application of this method can reduce the cost of agricultural equipment and provide a scientific basis for agricultural production practices, helping to reduce losses caused by diseases.

UAV multispectral remote sensing for agriculture: A comparative study of radiometric correction methods under varying illumination conditions

Unmanned aerial vehicles (UAVs) equipped with multispectral cameras have been widely used in precision agriculture. However, a notable challenge is the variation in ambient illumination, which affects the accuracy and reliability of UAV-based spectral-data acquisition. In this study, the aim is to evaluate and enhance the performance of existing radiometric correction techniques under varying illumination conditions, primarily concerning radiometric accuracy and homogeneity. Seven methods including three conventional methods and four new methods were employed for correcting the MicaSense Altum multispectral system which equips with a downwelling light sensor (DLS). Two specific strategies were adopted: (1) capturing reference panels at UAV flying altitudes, and (2) strategically placing multiple sets of reference panels throughout the study area. The result shows that calibrating images one time, for instance, the empirical line method (ELM), is seriously affected by the variable illumination. The commercial solution that using the DLS helps improve the uniformity of orthomosaics but lower its radiometric accuracy. Optimising the use of the DLS by capturing panels at the UAV's flight altitude can greatly improve accuracy. Additionally, when the DLS is unavailable, strategically placing multiple reference panels across the field and correcting calibration parameters for each image can effectively help mitigate the impact of varying illumination on generated reflectance orthomosaics. In conclusion, selecting suitable radiometric correction methods is crucial for UAV multi-spectral data collection when facing variable illumination conditions.