Browsing by Author "Ferahtia, J."

Now showing 1 - 8 of 8

Acoustic impedance inversion by feedback artificial neural network
(Elsevier, 2010) Baddari, K.; Djarfour, Noureddine; Aïfa, Tahar; Ferahtia, J.
The determination of acoustic impedance distribution from the seismic data field measurement can be expressed as an ill-posed inverse problem. This work deals with the use of the Elman artificial neural network (ANN) (feedback connection) for the seismic data inversion. In the proposed structure the hidden neuron outputs from the previous time step are fed back to their inputs through time delay units; this enables them to process temporal behaviour and provide multi-step-ahead predictions. The ANN architectures and learning rules are presented to allow the best estimate of acoustic impedance from seismic data. The effects of network architectures using 5 to 60 neurons and 10 to 90 neurons in the hidden layer respectively for synthetic and real data on the rate of convergence and prediction accuracy of ANN models are discussed. The behaviour of networks observed on training data is very similar to the one observed on test data. The results obtained clearly prove the feasibility of the proposed method for seismic data inversion by feedback neural networks. Different tests indicate that the back-propagation conjugate gradient algorithm can easily train the proposed Elman ANN structure without getting stuck in local minima
Application of feedback connection artificial neural network to seismic data filtering
(Elsevier, 2008) Djarfour, Noureddine; Aïfa, Tahar; Baddari, K.; Mihoubi, A.; Ferahtia, J.
The Elman artificial neural network (ANN) (feedback connection) was used for seismic data filtering. The recurrent connection that characterizes this network offers the advantage of storing values from the previous time step to be used in the current time step. The proposed structure has the advantage of training simplicity by a back-propagation algorithm (steepest descent). Several trials were addressed on synthetic (with 10% and 50% of random and Gaussian noise) and real seismic data using respectively 10 to 30 neurons and a minimum of 60 neurons in the hidden layer. Both an iteration number up to 4000 and arrest criteria were used to obtain satisfactory performances. Application of such networks on real data shows that the filtered seismic section was efficient. Adequate cross-validation test is done to ensure the performance of network on new data sets
Application of signal dependent rank-order mean filter to the removal of noise spikes from 2D electrical resistivity imaging data
(2009) Ferahtia, J.; Djarfour, Noureddine; Baddari, K.; Guérin, R.
It is well-known that when inverting two-dimensional (2D) electrical resistivity data, a major source of errors is the presence of noise and in particular noise spikes. The popular median filter is often applied to the removal of single spikes. However, when the signal is highly corrupted with successive spikes, the median filter performance is poor. This paper deals with the use of the signal dependent rank-order mean filter for the detection and removal of noise spikes from highly corrupted 2D electrical resistivity imaging data. In addition to its computational simplicity, this filter is shown to be extremely robust, even in the presence of very strong noise, especially when it is applied recursively. The signal dependent rank-order mean filter was tested on 2D synthetic resistivity data contaminated by near-surface inhomogeneities and the results confirmed efficient removal of the disturbances normally associated with near-surface inhomogeneities. The signal dependent rank-order mean filter was also applied to field data and demonstrated its ability to significantly improve the accuracy of the inversion process and to produce good visual results in the inverted electrical sections
A fuzzy logic-based filter for the removal of spike noise from 2D electrical resistivity data
(Elsevier, 2012) Ferahtia, J.; Djarfour, Noureddine; Baddari, K.; Kheldoun, Aissa
In this paper, a filter based on fuzzy logic is proposed to remove spike noise from 2 dimensional electrical resistivity data. The noise detection used in this paper is based on differentiating noisy samples from the central sample inside a moving window. These fuzzy derivatives are used by the fuzzy inference system to detect corrupted samples. To assess the performance of the proposed filter for the removal of spike noise, the root-mean squared error as well as the signal-to-noise ratio were used as an objective criterion. It has been demonstrated by synthetic and real examples that the proposed filter achieves quite good results compared to the standard median filter as well as to the very effective SD-ROM filter
Image-based processing techniques applied to seismic data filtering
(Elsevier, 2013) Ferahtia, J.; Aïfa, Tahar; Baddari, K.; Djarfour, Noureddine; Eladj, S.
Impulse noise reduction in 2D electrical resistivity imaging data based on fuzzy logic
(IEEE, 2011) Ferahtia, J.; Djarfour, Noureddine; Baddari, Kamel; Khaldoun, Asmae
Seismic attributes combination to enhance detection of bright spot associated with hydrocarbons
(Taylor & Francis, 2012) Farfour, M.; Yoon, W.J.; Ferahtia, J.; Djarfour, Noureddine
Tomographic velocity images by artificial neural networks
(2007) Djarfour, Noureddine; Ferahtia, J.; Baddari, K.
The present study deals with the use of Elman artificial neural network (feedback connexion) to reconstruct the velocity image from a traveltime in the seismic tomography experiment. This recurrent connection provides the advantage to store values from the previous time step, which can be used in the actual time step. The backpropagation algorithm has been used to learn the suggested neural network. Efficiency of these networks has been tested in training and generalization phases. A comparative reconstruction with two classical methods was performed using backprojection and Algebraic Reconstruction Techniques (ART). The obtained results clearly show improvements of the quality of the reconstruction obtained by artificial neural networks