Analysis of the Temporal Structure Evolution of Physical Systems with the Self-Organising Tree Algorithm (SOTA): Application for Validating Neural Network Systems on Adaptive Optics Data before On-Sky Implementation

Gómez, Sergio Luis Suárez; Santos, Joana; Rodríguez, Francisco Javier Iglesias; Juez, Francisco Javier de Cos

doi:10.3390/e19030103

Cited by 19 publications

(11 citation statements)

References 30 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This structure or topology, shown in Figure 2, along with the training process for the correction of the weights, conforms with one of the most well-known models of ANN, the Multi-Layer Perceptron (MLP). Since the training process allows these techniques to learn directly from data measurements [17], ANNs are particularly useful for modeling, forecasting and prediction [15], being widely known for their capacity of representing both linear and non-linear models, and for extrapolating that knowledge to unknown data. Since the training process allows these techniques to learn directly from data measurements [17], ANNs are particularly useful for modeling, forecasting and prediction [15], being widely known for their capacity of representing both linear and non-linear models, and for extrapolating that knowledge to unknown data.…”

Section: Deep Learningmentioning

confidence: 99%

“…The topology of a CNN is illustrated in Figure 3. Since the training process allows these techniques to learn directly from data measurements [17], ANNs are particularly useful for modeling, forecasting and prediction [15], being widely known for their capacity of representing both linear and non-linear models, and for extrapolating that knowledge to unknown data.…”

Section: Deep Learningmentioning

confidence: 99%

“…This reconstruction algorithm uses data taken from the SH sensor, and provides a tomographic reconstruction of the atmospheric profile, allowing one to compensate for the aberrated wavefront by means of a deformable mirror [15]. The improvement of this deep learning approach has been tested for the adaptive optics reconstruction during night observations in real telescopes or even alternative validation procedures of the techniques previously to its on-sky implementation [16,17].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Defocused Image Deep Learning Designed for Wavefront Reconstruction in Tomographic Pupil Image Sensors

et al. 2020

View full text Add to dashboard Cite

Mathematical modelling methods have several limitations when addressing complex physics whose calculations require considerable amount of time. This is the case of adaptive optics, a series of techniques used to process and improve the resolution of astronomical images acquired from ground-based telescopes due to the aberrations introduced by the atmosphere. Usually, with adaptive optics the wavefront is measured with sensors and then reconstructed and corrected by means of a deformable mirror. An improvement in the reconstruction of the wavefront is presented in this work, using convolutional neural networks (CNN) for data obtained from the Tomographic Pupil Image Wavefront Sensor (TPI-WFS). The TPI-WFS is a modified curvature sensor, designed for measuring atmospheric turbulences with defocused wavefront images. CNNs are well-known techniques for its capacity to model and predict complex systems. The results obtained from the presented reconstructor, named Convolutional Neural Networks in Defocused Pupil Images (CRONOS), are compared with the results of Wave-Front Reconstruction (WFR) software, initially developed for the TPI-WFS measurements, based on the least-squares fit. The performance of both reconstruction techniques is tested for 153 Zernike modes and with simulated noise. In general, CRONOS showed better performance than the reconstruction from WFR in most of the turbulent profiles, with significant improvements found for the most turbulent profiles; overall, obtaining around 7% of improvements in wavefront restoration, and 18% of improvements in Strehl.

show abstract

Section: Deep Learningmentioning

confidence: 99%

Section: Deep Learningmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Defocused Image Deep Learning Designed for Wavefront Reconstruction in Tomographic Pupil Image Sensors

et al. 2020

View full text Add to dashboard Cite

show abstract

“…However, it is important to remark that the project is in a very early stage, and it is needed to test the different experiments in a much deeper way. But there are lots of different ideas to improve the performance of CNNs that could be applied to solar AO, like the use of recurrent neural networks [35], on-line training [36], [37] or even apply the notion of classification when computing the outputs. For last, it is interesting to keep in mind that in astronomy, not only AO could be benefited for the usage of neural networks since, for instance, the detection of exoplanets [38], [39] has already done.…”

Section: Conclusion and Future Linesmentioning

confidence: 99%

An approach using deep learning for tomographic reconstruction in solar observation

Gómez¹,

Gutiérrez²,

Lasheras³

et al. 2017

Proceedings of the Adaptive Optics for Extremely Large Telescopes 5

View full text Add to dashboard Cite

Solar observations with large telescopes have to face numerous challenges for its implementation. Adapting the current reconstructor systems of adaptive optics, which were developed for night observations to remove the atmospheric turbulence, is one of them. Neural networks were proved as a great solution in different fields, as image processing or in tomographic reconstruction. In this paper, possible uses of neural networks in solar adaptive optics will be explored, providing some early results of their application to different fields of adaptive optics.

show abstract

“…The use of Artificial Neural Networks (ANNs), a mathematical model from the field of Artificial Intelligence, is widely known for the uses in pattern recognition and prediction [17,18], image [19], speech and handwriting processing, and all sorts of complex modeling of physical system [20,21]. In particular, these techniques have been proven to be successful in several scenarios for nocturnal AO [22].…”

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Networks Approach for Solar Reconstruction in SCAO Configurations

Gómez

González-Gutiérrez

Riesgo

et al. 2019

Sensors

View full text Add to dashboard Cite

Correcting atmospheric turbulence effects in light with Adaptive Optics is necessary, since it produces aberrations in the wavefront of astronomical objects observed with telescopes from Earth. These corrections are performed classically with reconstruction algorithms; between them, neural networks showed good results. In the context of solar observation, the usage of Adaptive Optics on solar differs from nocturnal operations, bringing up a challenge to correct the image aberrations. In this work, a convolutional approach is given to address this issue, considering SCAO configurations. A reconstruction algorithm is presented, “Shack-Hartmann reconstruction with deep learning on solar–prototype” (proto-HELIOS), to correct on fixed solar images, achieving an average 85.39% of precision in the reconstruction. Additionally, results encourage to continue working with these techniques to achieve a reconstruction technique for all the regions of the sun.

show abstract

Analysis of the Temporal Structure Evolution of Physical Systems with the Self-Organising Tree Algorithm (SOTA): Application for Validating Neural Network Systems on Adaptive Optics Data before On-Sky Implementation

Cited by 19 publications

References 30 publications

Defocused Image Deep Learning Designed for Wavefront Reconstruction in Tomographic Pupil Image Sensors

Defocused Image Deep Learning Designed for Wavefront Reconstruction in Tomographic Pupil Image Sensors

An approach using deep learning for tomographic reconstruction in solar observation

Convolutional Neural Networks Approach for Solar Reconstruction in SCAO Configurations

Contact Info

Product

Resources

About