Cespe UnB

Editorial Office:
R. S. Oyarzabal

Technical Support:
D. H. Diaz
M. A. Gomez
W. Abrahão
G. Oliveira

Publisher by
Knobook Pub



Evaluaiting the Deep Learning Accuracy in Data Extraction from Synthetic Image Sequence

doi: 10.6062/jcis.2019.10.03.0168

(Free PDF)


André Franceschi de Angelis and Thaı́s Rocha


We have investigating the use of Deep Learning (DL) to process sequences of images captured by satellites aiming to improve the quality of river flow forecasting methods, because current ones are still not accurate enough to support efficient management of large national electrical systems. Towards this goal, we are assessing the accuracy of DL networks in extracting information from image sequences by means of classification processes. We have set a test environment composed by an image sequence generator, some generating models, the Nvidia DIGITS tool, two DL preset networks, and the needed hardware. Each model produced one image sequence and one data series corresponding to a selected measure in the images. We have trained the DL networks and evaluated its accuracy in extracting the measured data. In this paper, we show that the performance of DL is extremely sensible to the image type, the measure taken into account, and the DL network applied. Our process presented better performance recognizing coverage area rates in images that resemble clouds and linear distances, but had poor accuracy with angles.


Deep learning, synthetic images, image classification, river flow forecasting.


Combining wavelets and linear spectral mixture model for MODIS satellite sensor time-series analysis
doi: 10.6062/jcis.2008.01.01.0005
Freitas and Shimabukuro(Free PDF)

Riddled basins in complex physical and biological systems
doi: 10.6062/jcis.2009.01.02.0009
Viana et al.(Free PDF)

Use of ordinary Kriging algorithm and wavelet analysis to understanding the turbidity behavior in an Amazon floodplain
doi: 10.6062/jcis.2008.01.01.0006
Alcantara.(Free PDF)

A new multi-particle collision algorithm for optimization in a high performance environment
doi: 10.6062/jcis.2008.01.01.0001
Luz et al.((Free PDF)

Reviewer Guidelines
(Under Construction)
Advertises Media Information