Single sensor systems and standard optical—usually RGB CCTV video cameras—fail to provide adequate observations, or the amount of spectral information required to build rich, expressive, discriminative features for object detection and tracking tasks in challenging outdoor and indoor scenes under various environmental/illumination conditions. Towards this direction, we have designed a multisensor system based on thermal, shortwave infrared, and hyperspectral video sensors and propose a processing pipeline able to perform in real-time object detection tasks despite the huge amount of the concurrently acquired video streams. In particular, in order to avoid the computationally intensive coregistration of the hyperspectral data with other imaging modalities, the initially detected targets are projected through a local coordinate system on the hypercube image plane. Regarding the object detection, a detector-agnostic procedure has been developed, integrating both unsupervised (background subtraction) and supervised (deep learning convolutional neural networks) techniques for validation purposes. The detected and verified targets are extracted through the fusion and data association steps based on temporal spectral signatures of both target and background. The quite promising experimental results in challenging indoor and outdoor scenes indicated the robust and efficient performance of the developed methodology under different conditions like fog, smoke, and illumination changes.
Recent advances in photonics and imaging technology allow the development of cutting-edge, lightweight hyperspectral sensors, both push-broom/line-scanning and snapshot/frame. At the same time, emerging applications in robotics, food inspection, medicine and earth observation are posing critical challenges on real-time processing and computational efficiency, both in terms of accuracy and power consumption. In this direction, in the current paper, we accelerate hyperspectral processing kernels by utilizing FPGAs, i.e., Zynq-7000 SoC, to perform similaritybased matching of spectral signatures. We propose a custom HW architecture based on multi-level parallelization, modularity, and parametric VHDL coding, which allows for in-depth design space exploration and trade-off analysis. Depending on configuration, our implementation processes 22−107 Megapixels per second providing an acceleration of 40−355x vs Intel-i3 CPU and 360−10 4 x vs the embedded ARM Cortex A9, whereas the overall detection quality ranges from 56% to 97% when evaluated with multiple objects and images of 285 spectral channels.
ABSTRACT:In order to exploit efficiently very high resolution satellite multispectral data for precision agriculture applications, validated methodologies should be established which link the observed reflectance spectra with certain crop/plant/fruit biophysical and biochemical quality parameters. To this end, based on concurrent satellite and field campaigns during the veraison period, satellite and in-situ data were collected, along with several grape samples, at specific locations during the harvesting period. These data were collected for a period of three years in two viticultural areas in Northern Greece. After the required data pre-processing, canopy reflectance observations, through the combination of several vegetation indices were correlated with the quantitative results from the grape/must analysis of grape sampling. Results appear quite promising, indicating that certain key quality parameters (like brix levels, total phenolic content, brix to total acidity, anthocyanin levels) which describe the oenological potential, phenolic composition and chromatic characteristics can be efficiently estimated from the satellite data.
Mapping water stress in vineyards, at the parcel level, is of significant importance for supporting crop management decisions and applying precision agriculture practices. In this paper, a novel methodology based on aerial Shortwave Infrared (SWIR) data is presented, towards the estimation of water stress in vineyards at canopy scale for entire parcels. In particular, aerial broadband spectral data were collected from an integrated SWIR and multispectral instrumentation, onboard an unmanned aerial vehicle (UAV). Concurrently, in-situ leaf stomatal conductance measurements and supplementary data for radiometric and geometric corrections were acquired. A processing pipeline has been designed, developed, and validated, able to execute the required analysis, including data pre-processing, data co-registration, reflectance calibration, canopy extraction and water stress estimation. Experiments were performed at two viticultural regions in Greece, for several vine parcels of four different vine varieties, Sauvignon Blanc, Merlot, Syrah and Xinomavro. The performed qualitative and quantitative assessment indicated that a single model for the estimation of water stress across all studied vine varieties was not able to be established (r2 < 0.30). Relatively high correlation rates (r2 > 0.80) were achieved per variety and per individual variety clone. The overall root mean square error (RMSE) for the estimated canopy water stress was less than 29 mmol m−2 s−1, spanning from no-stress to severe canopy stress levels. Overall, experimental results and validation indicated the quite high potentials of the proposed instrumentation and methodology.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.