Mohsin Bilal

Solution of n-Queen problem using ACO

In this paper, a solution is proposed for n-Queen problem based on ACO (Ant Colony Optimization). The n-Queen problem become intractable for large values of ‘n’ and thus placed in NP (Non-Deterministic Polynomial) class problem. The n-Queen problem is basically a generalized form of 8-Queen problem. In 8-Queen problem, the goal is to place 8 queens such that no queen can kill the other using standard chess queen moves. So, in this paper, the proposed solution will be applied to 8-Queen problem. The solution can very easily be extended to the generalized form of the problem for large values of ‘n’. The paper contains the detail discussion of problem background, problem complexity, Ant Colony Optimization (Swarm Intelligence) and a fair amount of experimental graphs.

Estimation and optimization based ill-posed inverse restoration using fuzzy logic

Intelligent systems ranging from neural network, evolutionary computations and swarm intelligence to fuzzy systems are extensively exploited by researchers to solve variety of problems. In this paper focus is on deblurring that is considered as an inverse problem. It becomes ill-posed when noise contaminates the blurry image. Hence the problem is very sensitive to small perturbation in data. Conventionally, smoothness constraints are considered as a remedy to cater the sensitivity of the problem. In this paper, fuzzy rule based regularization parameter estimation is proposed with quadratic functional smoothness constraint. For deblurring image in the presence of noise, a constrained least square error function is minimized by the steepest descent algorithm. Visual results and quantitative measurements show the efficiency and robustness of the proposed technique compared to the state of the art and recently proposed methods.

Novel Optimization Framework to Recover True Image Data

This paper focuses on the restoration of spatial degradations that appear in images due to invariant or variant blurs and additive noise. It is one of the basic problems of visual information processing systems. The problem possesses issues of complexity, huge volume of data, uncertainty and a real-time response in critical applications. In this paper, a new optimization framework for restoration is proposed to solve the problem effectively. The proposed solution is modeled as constrained optimization of huge vectors, each representing a grayscale image in spatial domain. In the proposed framework, particle swarm optimization-based evolution is adopted to minimize the modified error estimate (MEE) for better restoration. The framework added hyperheuristic layer to combine local and global search properties. Therefore, randomness in the evolution, augmented with apriori knowledge from the problem domain, assisted in achieving the objective. In addition, an adaptive weighted regularization scheme is proposed in MEE to cater with the uncertainty due to ill-posed nature of the inverse problem. The visual and quantitative results are provided to endorse the effectiveness of the proposed framework in maximizing signal-to-noise ratio and minimizing well-known error measures in contrast to existing restoration methods.

Image Restoration Using Modified Hopfield Fuzzy Regularization Method

This paper addresses one of the primary problems of visual information processing known as image restoration. Image restoration is a challenging task because of its ill-posed inverse nature. A modified Hopfield neural network with fuzzy adaptive regularization is proposed that shows potential to minimize constraint mean square error in order to guarantee the optimized results. Adaptive regularization was achieved by using fuzzy quasi-range edge detector. The visual results along with the statistical measurements of the resultant images are presented in the paper. Improved SNRs show that the fuzzy regularization method is superior to other statistical and neural network methods when used along with the modified Hopfield neural network.

A solution to bipartite drawing problem using genetic algorithm

Crossing minimization problem in a bipartite graph is a well-known NP-Complete problem. Drawing the directed/undirected graphs such that they are easy to understand and remember requires some drawing aesthetics and crossing minimization is one of them. In this paper, we investigate an intelligent evolutionary technique i.e. Genetic Algorithm (GA) for bipartite drawing problem (BDP). Two techniques GA1 and GA2 are proposed based on Genetic Algorithm. It is shown that these techniques outperform previously known heuristics e.g., MinSort (M-Sort) and BaryCenter (BC) as well as a genetic algorithm based level permutation problem (LPP), especially when applied to low density graphs. The solution is tested over various parameter values of genetic bipartite drawing problem. Experimental results show the promising capability of the proposed solution over previously known heuristics.

Adaptive filter and morphological operators using binary PSO

Mathematical morphology is a tool for processing shapes in image processing. Adaptively finding the specific morphological filter is an important and challenging task in morphological image processing. In order to model the filter and filtering sequence for morphological operations adaptively, a novel technique based on binary particle swarm optimization (BPSO) is proposed. BPSO is a discrete PSO, where the components values of a particle position vector are either zero or one. The proposed method can be used for numerous types of applications, where the morphological processing is involved including but not limited to image segmentation, noise suppression and patterns recognition etc. The paper illustrates a fair amount of experimental results showing the promising ability of the proposed approach over previously known solutions. In particular, the proposed method is evaluated for noise suppression problem.

Modified particle swarm optimization and fuzzy regularization for pseudo de-convolution of spatially variant blurs

We propose a modified particle swarm optimization (MPSO) based method for Pseudo De-convolution of the ill-posed inverse problem namely, the space-variant image degradation (SVD). In this paper, SVD is simulated by the pseudo convolution of different sub-regions of the image with different known blurring kernels and additive random noise with unknown variance. Two heuristic modifications are proposed in PSO: 1) Initialization of the swarm and 2) Mutation of the global best. Fuzzy logic is applied for the computation of regularization parameter (RP) to cater for the sensitivity of the problem. The computation of RP is crucial due to the additive noise in the SVD image. Thus mathematical morphology (MM) is applied for better extraction of spatial activity from the distorted image. The performance of the proposed method is evaluated with different test images and noise powers. Comparative analysis demonstrates the superiority of proposed restoration, in terms of quantitative measures, over well-known existing and state-of-the-art SVD approaches.

A revised framework of machine learning application for optimal activity recognition

Data science augments manual data understanding with machine learning for potential performance increase. In this paper, data science methodology is examined to enhance machine learning application in smartphone based automatic human activity recognition (HAR). Eventually, a modified feature engineering and a novel post-learning data engineering are proposed in the machine learning framework as the alternate of data understanding for an effective HAR. The proposed framework is examined on two different HAR data sets demonstrating a possibility of data-driven machine learning for near an optimal classification of activities. The proposed framework exhibited effectiveness and efficiency when compared with the existing methods. The modified feature engineering resulted in 42% fewer features required by support vector machine to yield 97.3% correct recognition of human physical activities. However, the addition of post-learning data engineering further improved the model to perform 99% accurate classification, which is an almost optimal performance.

Image restoration by multivariate-stochastic optimization using improved particle swarm algorithm

Image restoration is a multivariate-stochastic optimization challenge. In this paper, an improved particle swarm algorithm is proposed for image restoration. The proposed method - Optimal Image Restoration using Improved Particle Swarm Algorithm (OIRIPSA) - is analyzed with an approximated (constrained least square error) and a true cost (mean squared error) measures, respectively. Initial swarm of heuristic solutions is constructed arbitrarily along with problem specific knowledge. An optimistic hyper layer is further integrated to enhance the swarm search procedure by constructing an incipient solution in the neighborhood of the generations best. OIRIPSA is engendering better restoration than Richardson-Lucy algorithm and a state-of-the-art restoration method.