scholarly journals The Influence of Aerial Hyperspectral Image Processing Workflow on Nitrogen Uptake Prediction Accuracy in Maize

2021 ◽  
Vol 14 (1) ◽  
pp. 132
Author(s):  
Tyler Nigon ◽  
Gabriel Dias Paiao ◽  
David J. Mulla ◽  
Fabián G. Fernández ◽  
Ce Yang
Keyword(s):  
Image Processing ◽  
High Resolution ◽  
Nitrogen Uptake ◽  
Prediction Accuracy ◽  
Hyperspectral Image ◽  
Mean Squared Error ◽  
Hyperspectral Imagery ◽  
Unmanned Aerial Systems ◽  
Model Accuracy ◽  
Processing Step

A meticulous image processing workflow is oftentimes required to derive quality image data from high-resolution, unmanned aerial systems. There are many subjective decisions to be made during image processing, but the effects of those decisions on prediction model accuracy have never been reported. This study introduced a framework for quantifying the effects of image processing methods on model accuracy. A demonstration of this framework was performed using high-resolution hyperspectral imagery (<10 cm pixel size) for predicting maize nitrogen uptake in the early to mid-vegetative developmental stages (V6–V14). Two supervised regression learning estimators (Lasso and partial least squares) were trained to make predictions from hyperspectral imagery. Data for this use case were collected from three experiments over two years (2018–2019) in southern Minnesota, USA (four site-years). The image processing steps that were evaluated include (i) reflectance conversion, (ii) cropping, (iii) spectral clipping, (iv) spectral smoothing, (v) binning, and (vi) segmentation. In total, 648 image processing workflow scenarios were evaluated, and results were analyzed to understand the influence of each image processing step on the cross-validated root mean squared error (RMSE) of the estimators. A sensitivity analysis revealed that the segmentation step was the most influential image processing step on the final estimator error. Across all workflow scenarios, the RMSE of predicted nitrogen uptake ranged from 14.3 to 19.8 kg ha−1 (relative RMSE ranged from 26.5% to 36.5%), a 38.5% increase in error from the lowest to the highest error workflow scenario. The framework introduced demonstrates the sensitivity and extent to which image processing affects prediction accuracy. It allows remote sensing analysts to improve model performance while providing data-driven justification to improve the reproducibility and objectivity of their work, similar to the benefits of hyperparameter tuning in machine learning applications.

scholarly journals Prediction of Early Season Nitrogen Uptake in Maize Using High-Resolution Aerial Hyperspectral Imagery

2020 ◽  
Vol 12 (8) ◽  
pp. 1234 ◽  
Author(s):  
Tyler Nigon ◽  
Ce Yang ◽  
Gabriel Dias Paiao ◽  
David Mulla ◽  
Joseph Knight ◽  
...  
Keyword(s):  
High Resolution ◽  
Random Forest ◽  
Hyperspectral Imaging ◽  
Nitrogen Uptake ◽  
Nitrogen Fertilizer ◽  
Prediction Accuracy ◽  
Hyperspectral Imagery ◽  
Spectral Features ◽  
Early Season ◽  
Development Stages

The ability to predict spatially explicit nitrogen uptake (NUP) in maize (Zea mays L.) during the early development stages provides clear value for making in-season nitrogen fertilizer applications that can improve NUP efficiency and reduce the risk of nitrogen loss to the environment. Aerial hyperspectral imaging is an attractive agronomic research tool for its ability to capture spectral data over relatively large areas, enabling its use for predicting NUP at the field scale. The overarching goal of this work was to use supervised learning regression algorithms—Lasso, support vector regression (SVR), random forest, and partial least squares regression (PLSR)—to predict early season (i.e., V6–V14) maize NUP at three experimental sites in Minnesota using high-resolution hyperspectral imagery. In addition to the spectral features offered by hyperspectral imaging, the 10th percentile Modified Chlorophyll Absorption Ratio Index Improved (MCARI2) was made available to the learning models as an auxiliary feature to assess its ability to improve NUP prediction accuracy. The trained models demonstrated robustness by maintaining satisfactory prediction accuracy across locations, pixel sizes, development stages, and a broad range of NUP values (4.8 to 182 kg ha−1). Using the four most informative spectral features in addition to the auxiliary feature, the mean absolute error (MAE) of Lasso, SVR, and PLSR models (9.4, 9.7, and 9.5 kg ha−1, respectively) was lower than that of random forest (11.2 kg ha−1). The relative MAE for the Lasso, SVR, PLSR, and random forest models was 16.5%, 17.0%, 16.6%, and 19.6%, respectively. The inclusion of the auxiliary feature not only improved overall prediction accuracy by 1.6 kg ha−1 (14%) across all models, but it also reduced the number of input features required to reach optimal performance. The variance of predicted NUP increased as the measured NUP increased (MAE of the Lasso model increased from 4.0 to 12.1 kg ha−1 for measured NUP less than 25 kg ha−1 and greater than 100 kg ha−1, respectively). The most influential spectral features were oftentimes adjacent to each other (i.e., within approximately 6 nm), indicating the importance of both spectral precision and derivative spectra around key wavelengths for explaining NUP. Finally, several challenges and opportunities are discussed regarding the use of these results in the context of improving nitrogen fertilizer management.

scholarly journals Advancing Floating Macroplastic Detection from Space Using Experimental Hyperspectral Imagery

2021 ◽  
Vol 13 (12) ◽  
pp. 2335
Author(s):  
Paolo Tasseron ◽  
Tim van Emmerik ◽  
Joseph Peller ◽  
Louise Schreyers ◽  
Lauren Biermann
Keyword(s):  
High Resolution ◽  
Hyperspectral Image ◽  
Hyperspectral Imagery ◽  
Spectral Band ◽  
Plastic Debris ◽  
Linear Discriminant ◽  
Spectral Bands ◽  
High Resolution Images ◽  
Discriminant Analyses ◽  
Shortwave Infrared

Airborne and spaceborne remote sensing (RS) collecting hyperspectral imagery provides unprecedented opportunities for the detection and monitoring of floating riverine and marine plastic debris. However, a major challenge in the application of RS techniques is the lack of a fundamental understanding of spectral signatures of water-borne plastic debris. Recent work has emphasised the case for open-access hyperspectral reflectance reference libraries of commonly used polymer items. In this paper, we present and analyse a high-resolution hyperspectral image database of a unique mix of 40 virgin macroplastic items and vegetation. Our double camera setup covered the visible to shortwave infrared (VIS-SWIR) range from 400 to 1700 nm in a darkroom experiment with controlled illumination. The cameras scanned the samples floating in water and captured high-resolution images in 336 spectral bands. Using the resulting reflectance spectra of 1.89 million pixels in linear discriminant analyses (LDA), we determined the importance of each spectral band for discriminating between water and mixed floating debris, and vegetation and plastics. The absorption peaks of plastics (1215 nm, 1410 nm) and vegetation (710 nm, 1450 nm) are associated with high LDA weights. We then compared Sentinel-2 and Worldview-3 satellite bands with these outcomes and identified 12 satellite bands to overlap with important wavelengths for discrimination between the classes. Lastly, the Normalised Vegetation Difference Index (NDVI) and Floating Debris Index (FDI) were calculated to determine why they work, and how they could potentially be improved. These findings could be used to enhance existing efforts in monitoring macroplastic pollution, as well as form a baseline for the design of future multispectral RS systems.

scholarly journals ONBOARD/ON-GROUND IMAGE PROCESSING CHAIN FOR HIGH-RESOLUTION EARTH OBSERVATION SATELLITES

2021 ◽  
Vol XLIII-B3-2021 ◽  
pp. 755-762
Author(s):  
E. Cucchetti ◽  
C. Latry ◽  
G. Blanchet ◽  
J.-M. Delvit ◽  
M. Bruno
Keyword(s):  
Image Processing ◽  
High Resolution ◽  
Mean Squared Error ◽  
Low Cost ◽  
Image Size ◽  
Processing Chain ◽  
Space Agency ◽  
Non Local ◽  
Earth Observation Satellites ◽  
The Cost

Abstract. Over the last decade, the French space agency (CNES) has designed and successfully operated high-resolution satellites such as Pléiades. High-resolution satellites typically acquire panchromatic images with fine spatial resolutions and multispectral images with coarser samplings for downlink constraints. The multispectral image is reconstructed on the ground, using pan-sharpening techniques. Onboard compression and ground processing affect however the quality of the final product. In this paper, we describe our next-generation onboard/on-ground image processing chain for high-resolution satellites. This paper focuses on onboard compression, compression artefacts correction, denoising, deconvolution and pan-sharpening. In the first part, we detail our fixed-quality compression approach, which limits compression effects to a fraction of the noise, thus preserving the useful information in an image. This approach optimises the bitrate at the cost of image size, which depends on the scene complexity. This technique requires however pre- and post-processing steps. The noisy HR images obtained after decompression are suited for non-local denoising algorithms. We show in the second part of this paper that non-local denoising outperforms previous techniques by 15% in terms of root mean-squared error when tested on simulated noiseless references. Deconvolution is also detailed. In the final part of this paper, we put forward an adaptation of this chain to low-cost CMOS Bayer colour matrices. We demonstrate that the concept of our image chain remains valid, provided slight modifications (in particular dedicated transformations of the colour planes and demosaicing). A similar chain is under investigation for future missions.

scholarly journals High‐resolution hyperspectral imagery from pushbroom scanners on unmanned aerial systems

2021 ◽  
Author(s):  
Jae‐In Kim ◽  
Junhwa Chi ◽  
Ali Masjedi ◽  
John Evan Flatt ◽  
Melba M. Crawford ◽  
...  
Keyword(s):  
High Resolution ◽  
Hyperspectral Imagery ◽  
Unmanned Aerial Systems ◽  
Aerial Systems

Comparison of Conventional and Electron Spectroscopic Imaging in the Fixed Beam Electron Microscope of Ordered Protein Arrays

1985 ◽  
Vol 43 ◽  
pp. 74-75
Author(s):  
E. L. Buhle ◽  
U. Aebi
Keyword(s):  
Image Processing ◽  
Electron Microscope ◽  
High Resolution ◽  
Image Plane ◽  
Electron Spectroscopic Imaging ◽  
Protein Arrays ◽  
Beam Electron ◽  
Average Unit ◽  
Fixed Beam ◽  
Energy Components

CTEM brightfield images are formed by a combination of relatively high resolution elastically scattered electrons and unscattered and inelastically scattered electrons. In the case of electron spectroscopic images (ESI), the inelastically scattered electrons cause a loss of both contrast and spatial resolution in the image. In the case of ESI imaging on the Zeiss EM902, the transmited electrons are dispersed into their various energy components by passing them through a magnetic prism spectrometer; a slit is then placed in the image plane of the prism to select the electrons of a given energy loss for image formation. The purpose of this study was to compare CTEM with ESI images recorded on a Zeiss EM902 of ordered protein arrays. Digital image processing was employed to analyze the average unit cell morphologies of the two types of images.

The Convolution Neural Network Combined with the HT Person Fit Statistic to Develop an APP for Detecting Dengue Fever in Children: Development and Usability Study (Preprint)

2019 ◽  
Author(s):  
CHIEN WEI ◽  
Chi Chow Julie ◽  
Chou Willy
Keyword(s):  
Neural Network ◽  
Dengue Fever ◽  
Prediction Accuracy ◽  
Early Stage ◽  
Family Members ◽  
Convolution Neural Network ◽  
Public Health Issue ◽  
Person Fit ◽  
Model Accuracy ◽  
Comparison Results

UNSTRUCTURED Backgrounds: Dengue fever (DF) is an important public health issue in Asia. However, the disease is extremely hard to detect using traditional dichotomous (i.e., absent vs. present) evaluations of symptoms. Convolution neural network (CNN), a well-established deep learning method, can improve prediction accuracy on account of its usage of a large number of parameters for modeling. Whether the HT person fit statistic can be combined with CNN to increase the prediction accuracy of the model and develop an application (APP) to detect DF in children remains unknown. Objectives: The aim of this study is to build a model for the automatic detection and classification of DF with symptoms to help patients, family members, and clinicians identify the disease at an early stage. Methods: We extracted 19 feature variables of DF-related symptoms from 177 pediatric patients (69 diagnosed with DF) using CNN to predict DF risk. The accuracy of two sets of characteristics (19 symptoms and four other variables, including person mean, standard deviation, and two HT-related statistics matched to DF+ and DF−) for predicting DF, were then compared. Data were separated into training and testing sets, and the former was used to predict the latter. We calculated the sensitivity (Sens), specificity (Spec), and area under the receiver operating characteristic curve (AUC) across studies for comparison. Results: We observed that (1) the 23-item model yields a higher accuracy rate (0.95) and AUC (0.94) than the 19-item model (accuracy = 0.92, AUC = 0.90) based on the 177-case training set; (2) the Sens values are almost higher than the corresponding Spec values (90% in 10 scenarios) for predicting DF; (3) the Sens and Spec values of the 23-item model are consistently higher than those of the 19-item model. An APP was subsequently designed to detect DF in children. Conclusion: The 23-item model yielded higher accuracy rates (0.95) and AUC (0.94) than the 19-item model (accuracy = 0.92, AUC = 0.90). An APP could be developed to help patients, family members, and clinicians discriminate DF from other febrile illnesses at an early stage.

Sign in / Sign up

Export Citation Format

Share Document