Alper Basturk

Impulse Noise Removal From Digital Images by a Detail-Preserving Filter Based on Type-2 Fuzzy Logic

A novel image filter based on type-2 fuzzy logic techniques is proposed for detail-preserving restoration of digital images corrupted by impulse noise. The performance of the proposed filter is evaluated for different test images corrupted at various noise densities and also compared with representative conventional as well as state-of-the-art impulse noise filters from the literature. Experimental results show that the proposed filter exhibits superior performance over the competing operators and is capable of efficiently suppressing the noise in the image while at the same time effectively preserving thin lines, edges, texture, and other useful information within the image.

Application of Type-2 Fuzzy Logic Filtering to Reduce Noise in Color Images

In this paper, we present a novel application of type-2 fuzzy logic to the design of an image processing operator called an impulse detector. The type-2 fuzzy logic based impulse detector can be used to guide impulse noise removal filters to significantly improve their filtering performance and enhance their output images. The design of the proposed impulse detector is based on two 3-input 1-output first order Sugeno type interval type-2 fuzzy inference systems. The internal parameters of the type-2 fuzzy membership functions of the systems are determined by training. The performance of the impulse detector is evaluated by using it in combination with four popular impulse noise filters from the literature on four different popular test images under three different noise conditions. The results demonstrate that the type-2 fuzzy logic based impulse detector can be used as an efficient tool to effectively improve the performances of impulse noise filters and reduce the impulse noise undesirable distortion effects.In this paper, we present a novel application of type-2 fuzzy logic to the design of an image processing operator called an impulse detector. The type-2 fuzzy logic based impulse detector can be used to guide impulse noise removal filters to significantly improve their filtering performance and enhance their output images. The design of the proposed impulse detector is based on two 3-input 1-output first order Sugeno type interval type-2 fuzzy inference systems. The internal parameters of the type-2 fuzzy membership functions of the systems are determined by training. The performance of the impulse detector is evaluated by using it in combination with four popular impulse noise filters from the literature on four different popular test images under three different noise conditions. The results demonstrate that the type-2 fuzzy logic based impulse detector can be used as an efficient tool to effectively improve the performances of impulse noise filters and reduce the impulse noise undesirable distortion effects.

Detail-preserving restoration of impulse noise corrupted images by a switching median filter guided by a simple neuro-fuzzy network

A new operator for the restoration of digital images corrupted by impulse noise is presented. The proposed operator is a simple recursive switching median filter guided by a neuro-fuzzy network functioning as an impulse detector. The internal parameters of the neuro-fuzzy impulse detector are adaptively optimized by training. The training is easily accomplished by using simple artificial images that can be generated in a computer. The most distinctive feature of the proposed operator over other operators is that it offers excellent detail- and texture-preservation performance, while effectively removing noise from the input image. Extensive experiments show that the proposed operator may be used for efficient restoration of digital images corrupted by impulse noise without distorting the useful information in the image.

Performance analysis of the coarse-grained parallel model of the artificial bee colony algorithm

Despite the efficiency of evolutionary algorithms is prominent for large scale problems, their running times in terms of CPU time are quite large. Multi processing units served by recent hardware developments can be employed to overcome this drawback reducing the running time and sharing the total workload. However, evolutionary algorithms cannot be directly distributed to processing units due to their cooperative working models. These models need to be modified to be able to run them on distributed environments without causing deterioration in performance. In this study, a detailed performance analysis of a parallel model for the artificial bee colony algorithm, which is one of the recently developed swarm based evolutionary algorithms and a promising numerical optimization tool, is proposed. For this purpose large-scale benchmark problems are solved by the proposed model and also its original sequential counterpart model. The model is also applied to a real-world problem: training of neural networks for classification purposes. Comparative results show that the artificial bee colony algorithm is very suitable to use in parallel architectures since it has the ability to produce high quality solutions with small populations due to its perturbation operator. The proposed model decreases the running time in addition to improving the performance and convergence rate of the algorithm. It can be said that the speedup gained over its sequential counterpart is almost linear.

Parallel Implementation of Synchronous Type Artificial Bee Colony Algorithm for Global Optimization

Evolutionary algorithms often need huge running times when solving large-scale optimization problems. One of the solutions for this issue is to introduce parallelization into the algorithm. To benefit from this approach for the artificial bee colony optimization algorithm, we present a new synchronous and parallel version of the algorithm. Performances of the proposed version and the original asynchronous algorithm are compared in terms of efficiency and speedup. Algorithms are competed to solve 20 large-scale global optimization problems. Comparative results show that the proposed parallel algorithm is still efficient as asynchronous version while it requires much less time to solve complex and large problems.

A Detail-Preserving Type-2 Fuzzy Logic Filter for Impulse Noise Removal from Digital Images

A novel filtering operator based on type-2 fuzzy logic techniques is proposed for detail preserving restoration of impulse noise corrupted images. The performance of the proposed operator is tested for different test images corrupted at various noise densities and also compared with representative conventional as well as state-of-the-art impulse noise removal operators from the literature. Experimental results show that the proposed operator exhibits superior performance over the competing operators and is capable of efficiently suppressing the noise in the image while at the same time effectively preserving the useful information in the image.

A new efficient training strategy for deep neural networks by hybridization of artificial bee colony and limited–memory BFGS optimization algorithms

Abstract Working up with deep learning techniques requires profound understanding of the mechanisms underlying the optimization of the internal parameters of complex structures. The major factor limiting this understanding is that there exist only a few optimization methods such as gradient descent and Limited–memory Broyden–Fletcher–Goldfarb–Shannon (L-BFGS) to find the best local minima of the problem space for these complex structures such as deep neural network (DNN). Therefore, in this paper, we represent a new training approach named hybrid artificial bee colony based training strategy (HABCbTS) to tune the parameters of a DNN structure, which includes one or more autoencoder layers cascaded to a softmax classification layer. In this strategy, a derivative-free optimization algorithm “ABC” is combined with a derivative-based algorithm “L-BFGS” to construct “HABC”, which is used in the HABCbTS. Detailed simulation results supported by statistical analysis show that the proposed training strategy results in better classification performance compared to the DNN classifier trained with the L-BFGS, ABC and modified ABC. The obtained classification results are also compared with the state-of-the-art classifiers, including MLP, SVM, KNN, DT and NB on 15 data sets with different dimensions and sizes.

Inspection of defects in fabrics using Gabor wavelets and principle component analysis

In this paper, a new method for inspection of textile defects in fabrics is presented. The method is based upon the extraction of fabric features by Gabor wavelets. The Gabor wavelets transform provides an effective way to analyze images and extract features of textures. Principal component analysis using singular value decomposition is used to reduce the dimension of feature vectors. Performance of the method has been tested with defective fabric images taken from TILDA textile texture database. Experiments show that these defects are detected accurately.

Parallel population-based algorithm portfolios

Although many algorithms have been proposed, no single algorithm is better than others on all types of problems. Therefore, the search characteristics of different algorithms that show complementary behavior can be combined through portfolio structures to improve the performance on a wider set of problems. In this work, a portfolio of the Artificial Bee Colony, Differential Evolution and Particle Swarm Optimization algorithms was constructed and the first parallel implementation of the population-based algorithm portfolio was carried out by means of a Message Passing Interface environment. The parallel implementation of an algorithm or a portfolio can be performed by different models such as master-slave, coarse-grained or a hybrid of both, as used in this study. Hence, the efficiency and running time of various parallel implementations with different parameter values and combinations were investigated on benchmark problems. The performance of the parallel portfolio was compared to those of the single constituent algorithms. The results showed that the proposed models reduced the running time and the portfolio delivered a robust performance compared to each constituent algorithm. It is observed that the speedup gained over the sequential counterpart changed significantly depending on the structure of the portfolio. The portfolio is also applied to a training of neural networks which has been used for time series prediction. Result demonstrate that, portfolio is able to produce good prediction accuracy.

Clonal selection algorithm based cloning template learning for edge detection in digital images with CNN

A cellular neural network (CNN) based edge detector optimized by clonal selection algorithm is presented. Cloning templates of the proposed CNN is adaptively tuned by using simple training images. The performance of the proposed edge detector is evaluated on different test images and compared with popular edge detectors from the literature. Simulation results indicate that the proposed CNN operator outperforms competing edge detectors and offers superior performance in edge detection in digital images.