1,720,965 research outputs found
Real time Optimal Allocation for I-AUV with Interacting Thrusters
No full textEnergy saving is a relevant issue for battery powered Intervention Autonomous Underwater Vehicles which are designed for both short and long-range mission. The energy consumption of an I-AUV is affected by several effects like hydrodynamics, onboard electronics and thrusters cross-couplings. I-AUVs are usually over-actuated system, thus the actuator's interaction takes relevance and should take part within the energy saving process. In this paper the authors present a study on optimal control allocation that aims at considering the interactions between the propellers of an over actuated vehicle where usually two or more thrusters can interfere each other resulting in reduction of the allocation efficiency. The paper presents the mathematical formulation for interacting propeller considering the wake effect and proposes to dynamically adjust the control allocation matrix in order to obtain a cost effective control allocation without modifying the control layer. The existence of a minimum in the energy consumption during the cruising task in function of the parametric control allocation matrix is proved numerically. Thus a perturbation-based extremum seeking approach is used in order to dynamically adapt the parametric allocation matrix and seek the optimal allocation setpoint without explicit knowledge of the real coupling
A Fuzzy Guidance System for Rendezvous and Pursuit of Moving Targets
Full text linkThis article presents the development of a fuzzy guidance system (FGS) for unmanned aerial
vehicles capable of pursuing and performing rendezvous with static and mobile targets. The system is
designed to allow the vehicle to approach a maneuvering target from a desired direction of arrival and
to terminate the rendezvous at a constant distance from the target. In order to perform a rendezvous
with a maneuvering target, the desired direction of arrival is adjusted over time to always approach
the target from behind, so that the aircraft and target velocity vectors become aligned. The proposed
guidance system assumes the presence of an autopilot and uses a set of Takagi–Sugeno fuzzy controllers
to generate the orientation and speed references for the velocity and heading control loops, given the
relative position and velocity between the aircraft and the target. The FGS treats the target as a mobile
waypoint in a 4-D space (position in 2-dimensions, desired crossing heading and speed) and guides
the aircraft on suitable trajectories towards the target. Only when the vehicle is close enough to the
rendezvous point, the guidance law is complemented with an additional linear controller to manage
the terminal formation keeping phase. The capabilities of the proposed rendezvous-FGS are verified in
simulation on both maneuvering and non-maneuvering targets. Finally, experimental results using a
multi-rotor aerial system are presented for both fixed and accelerating targets
Sentinel-2 Image Generation via Stylegan3 Model
No full textSatellite imagery is crucial for remote sensing, allowing observation of Earth from various platforms with diverse payloads. These images have been extensively collected and utilized in civil and military applications. While relied upon for analysis and decision-making, satellite images are susceptible to manipulation. Advancements in image generation have led to sophisticated techniques, such as Generative Adversarial Neural Networks (GANs), capable of creating realistic synthetic images. However, there is limited research on applying GANs to satellite imagery. This study focuses on training the state of the art StyleGAN3 model on multispectral Sentinel-2 data to generate synthetic data
Multispectral satellite image generation using StyleGAN3
No full textSatellite-based remote sensing images are essential for Earth surface analysis, serving diverse purposes in both civilian and military domains. Satellite images are used for analysis and decision making and are considered a reliable source of information. Recently, the field of image generation has developed increasingly sophisticated techniques, such as generative neural models, usually known as generative adversarial networks (GANs), to create synthetic images from scratch that appear almost real. Generative models have traditionally been applied to RGB or grayscale images and have been used for generating fake images of faces, animals, or objects. Currently, there are few studies regarding the application of GAN to multispectral satellite images. This work aims to test GAN models against the generation of multispectral satellite images, and in particular, the work explores the ability of the state-of-the-art StyleGAN3 model to produce high-quality synthetic Sentinel-2 images. The work delves into the configuration, training process, and evaluation of StyleGAN3 using custom Sentinel-2 datasets. StyleGAN3 results are compared with those provided by the proGAN model, the only GAN model tested so far on multispectral satellite data. Evaluation methods include visual inspection, spectral signature analysis, and a modified Fréchet inception distance for quantitative assessment. Results show that StyleGAN3 outperforms proGAN model exhibiting visually pleasing images. The quantitative comparison shows that StyleGAN3 provides the best results in terms of FID scores, in particular the improvement compared to proGAN increases as the spatial extent and spectral dimension of the generated images increases
Matched Filter Based on the Radiative Transfer Model for CO2Estimation from PRISMA Hyperspectral Data
Full text linkThe rapid growth of hyperspectral satellite missions and the subsequent availability of hyperspectral images have encouraged the remote sensing community to investigate their potential in estimating the concentration of gases, such as methane (CH4) and carbon dioxide (CO2), which are related to the greenhouse effect. Though satellite hyperspectral sensors are not specifically designed for this purpose, they are expected to complement more specific satellite missions, such as NASA's Orbiting Carbon Observatory -2 and -3 (OCO-2 and OCO-3), both in terms of enriched temporal sampling and improved spatial resolution. In this work, we present a new method to estimate the column-averaged dry-air mole fraction of CO2 from hyperspectral data on a per-pixel basis. The method, which is here tailored to PRISMA images, leverages the spectral radiance samples collected in the short-wave InfraRed (SWIR) spectral region around the CO2 absorption band at 2000 nm. By assuming a linear model to describe the dependence of the observed radiance on the CO2 concentration, the estimation problem is reduced to matched filtering and can be effectively implemented in compliance with the low computational burden required to perform a pixel-by-pixel analysis. The performance of the presented method is investigated by means of a rigorous, physically based simulator that accurately reproduces the at-sensor radiance allowing one to check the validity of the assumptions and to assess the algorithm accuracy. The results show that the presented algorithm outperforms a benchmark continuum interpolated band ratio (CIBR)-based approach, which has been proposed in the literature to get fast per-pixel estimates of CO2 concentration
Automatic Detection and Correction of Defective Pixels in PRISMA Hyperspectral Data
No full textIn satellite hyperspectral sensors, a standard procedure based on homogeneous reference sources is used to regularly update the map of defective pixels (DPs). Unfortunately, this procedure often fails to detect some DPs both because they may arise in the interval between two standard calibration steps and because they may be characterized by subtle or unexpected signal values. The resulting hyperspectral image is affected by residual nonuniformity noise, which correlates in the along-track direction. This noise source reduces the quality of hyperspectral products, such as classification, unmixing, and material detection. In this article, we present a new procedure to find the location of the residual DPs in the detection matrix and we also propose an effective method to estimate the missing radiance values inferring them from the image pixels by leveraging both spatial and spectral correlation. The procedure, here tailored to PRecursore IperSpettrale della Missione Applicativa (PRISMA) hyperspectral images, is quite general and can be easily adapted to process images recorded by any satellite pushbroom hyperspectral sensor. The improved image quality yielded by the proposed procedure is first demonstrated qualitatively, by comparing the global RX (GRX) maps on the original image with those obtained after detection of the DPs and correction of their radiance values. The image quality improvement is then quantified on a set of nine PRISMA images, recorded in an interval of about two years, using two ad hoc defined indexes. The analysis is carried out separately for the visible and near-infrared (VNIR) and shortwave infrared (SWIR) spectrometers of the PRISMA mission
Defective Pixel Detection and Correction in Prisma Hyperspectral Data
No full textWe present a new strategy to identify bad pixels in hyperspectral pushbroom sensors and to replace the inaccurate radiance values with estimates derived from spectral and spatial analysis. The proposed method is quite effective to correct spaceborne hyperspectral data where the regular calibration of the instrument is more complex than in airborne applications. In this paper we discuss the results obtained on images acquired by the PRISMA hyperspectral instrument operated by the Italian Space Agency (ASI). These preliminary results show the effectiveness of the proposed strategy both in detecting even subtle sources of fixed pattern noise, otherwise undetectable using visual inspection of a single spectral band, and in accurately reconstructing the missing radiance values
Simultaneous tracking of multiple drones using convolutional neural networks and optical flow
No full textThe aim of this paper is to implement a multi-object tracker for detecting, discriminating and following multiple drones on the 2D image plane. The main system consists in a neural network trained for detecting drones on the image plane. Once the drones are identified, their Region Of Interest information are sent to the Kalman filter based tracking system in order to compute position estimates. Each detected drone is given an ID using the Global Nearest Neighbor and Auction Algorithms, then the tracking system employs a set of Kalman filters and each measure is assigned to the respective estimate/filter using the unique ID. In order to reduce the effects of the neural network computation (e.g detection) time, every position estimate is corrected using Optical Flow. Results are evaluated through MOT (Multi-Object Tracking) Metrics, aiming to demonstrate the benefits of implementing Optical Flow algorithms for building a robust multi-tracking system
A hybrid approach to detection and tracking of unmanned aerial vehicles
No full textThis paper proposes a novel, hybrid vision-based system to autonomously detect and track a specific moving target, in our case an evader UAV (Unmanned Aerial Vehicles), with a moving camera. The framework is based on a detection stage which exploits a Faster Region-based Convolutional Neural Network (Faster R-CNN) designed to detect the Region Of Interest (ROI) associated to the UAV’s position in the image plane. The moving target is tracked by using an Optical Flow-based tracking system and a Kalman Filter is used to give temporal consistency between consecutive measurements. The tracking system is designed to be able to achieve real-time image processing on embedded systems, for this reason a lag compensation algorithm for the delay due to the Faster R-CNN computation time is implemented. Algorithm’s performance is evaluated by computing the error between the true UAV position in the image plane and the estimated position resulting from the tracking system
- …
