Accelerated method for the optimization of quadratic image filter

Abstract. Quadratic image filter involves the second-order multiplications of an input image mask in addition to linear terms, and determining the weights of the quadratic filter using optimization methods requires intense computational power due to the cost of the resulting fitness function. A graphics processing unit (GPU)-based algorithm is proposed to determine quadratic image filter weights using genetic algorithms and particle swarm optimization methods. Since the most time consuming part of determining the mask weights using heuristic algorithms is the fitness computation process, the fitness computation process is designed to work on the GPU, and a significant acceleration is achieved. For this purpose, different designs such as direct method and population-based method have been developed within the GPU to minimize training time. According to experimental results, the proposed methods provide significant accelerations over sequential implementation. The first method computes the fitness function for each population member separately, and it produces about 43 times acceleration over the sequential implementation. The second method developed is based on the complete evaluation of the population, and it produces about 151 times acceleration over the sequential implementation.

[1]  Devrim Akgün,et al.  GPU accelerated training of image convolution filter weights using genetic algorithms , 2015, Appl. Soft Comput..

[2]  Adhesh Garg,et al.  Comprehensive Analysis of the Uses of GPU and CUDA in Soft-Computing Techniques , 2019, 2019 6th International Conference on Signal Processing and Integrated Networks (SPIN).

[3]  Xiang Zhai,et al.  Efficient Rendering of Large-Scale CAD Models on a GPU Virtualization Architecture with Model Geometry Metrics , 2019, 2019 IEEE International Conference on Service-Oriented System Engineering (SOSE).

[4]  Robert D. Nowak,et al.  Random and pseudorandom inputs for Volterra filter identification , 1994, IEEE Trans. Signal Process..

[5]  Bin Luo,et al.  Timing Channel in IaaS: How to Identify and Investigate , 2018, IEEE Access.

[6]  Amitava Chatterjee,et al.  Volterra kernel based face recognition using artificial bee colonyoptimization , 2013, Eng. Appl. Artif. Intell..

[7]  R. Gopikakumari,et al.  Enhancement of calcifications in mammograms using volterra series based quadratic filter , 2012, 2012 International Conference on Data Science & Engineering (ICDSE).

[8]  Parikshit Sakurikar,et al.  Comparison sorting on hybrid multicore architectures for fixed and variable length keys , 2014, Int. J. High Perform. Comput. Appl..

[9]  BanerjeeDip Sankar,et al.  Comparison sorting on hybrid multicore architectures for fixed and variable length keys , 2014 .

[10]  S. Acton,et al.  Image enhancement using a contrast measure in the compressed domain , 2003, IEEE Signal Processing Letters.

[11]  Vali Derhami,et al.  Parallel genetic algorithm based on GPU for solving quadratic assignment problem , 2015, 2015 2nd International Conference on Knowledge-Based Engineering and Innovation (KBEI).

[12]  John D. Owens,et al.  GPU Computing , 2008, Proceedings of the IEEE.

[13]  G. Ramponi.  Edge extraction by a class of second-order nonlinear filters , 1986 .

[14]  C. Willmott,et al.  Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance , 2005 .

[15]  M. Kanamadi,et al.  Alpha Weighted Quadratic Filter Based Enhancement for Mammogram , 2013 .

[16]  G. Ramponi,et al.  A computational method for the design of 2-D nonlinear Volterra filters , 1988 .

[17]  M. B. Meenavathi,et al.  Volterra Filtering Techniques for Removal of Gaussian and Mixed Gaussian-Impulse Noise , 2007 .

[18]  Dean J. Krusienski,et al.  Particle swarm optimization for adaptive IIR filter structures , 2004, Proceedings of the 2004 Congress on Evolutionary Computation (IEEE Cat. No.04TH8753).

[19]  Yvan R. Petillot,et al.  Image processing optimization by genetic algorithm with a new coding scheme , 1995, Pattern Recognit. Lett..

[20]  Gerald Schaefer,et al.  UCID: an uncompressed color image database , 2003, IS&T/SPIE Electronic Imaging.

[21]  Pao-Ann Hsiung,et al.  Efficient Workload Balancing on Heterogeneous GPUs using Mixed-Integer Non-Linear Programming , 2014 .

[22]  Yue Shi,et al.  A modified particle swarm optimizer , 1998, 1998 IEEE International Conference on Evolutionary Computation Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98TH8360).

[23]  S. Mukherjee,et al.  Gravitational search algorithm in digital FIR low pass filter design , 2012, 2012 Third International Conference on Emerging Applications of Information Technology.

[24]  D. Wiesmann,et al.  Evolutionary Optimization Algorithms in Computational Optics , 1999 .

[25]  Süleyman Uzun,et al.  An Accelerated Method for Determining the Weights of Quadratic Image Filters , 2018, IEEE Access.

[26]  Jayshree Ghorpade,et al.  GPGPU Processing in CUDA Architecture , 2012, ArXiv.

[27]  Rui J. P. de Figueiredo,et al.  Exponential nonlinear Volterra filters for contrast sharpening in noisy images , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[28]  S. Fakhouri.  Identification of the Volterra kernels of nonlinear systems , 1980 .

[29]  Kye-Sung Lee,et al.  Parallelized multi–graphics processing unit framework for high-speed Gabor-domain optical coherence microscopy , 2014, Journal of biomedical optics.

[30]  Helmar Burkhart,et al.  Algorithmic performance studies on graphics processing units , 2008, J. Parallel Distributed Comput..

[31]  Tianzi Jiang,et al.  A modified Gabor filter design method for fingerprint image enhancement , 2003, Pattern Recognit. Lett..

[32]  Bo An,et al.  GPU Scheduling for Short Tasks in Private Cloud , 2019, 2019 IEEE International Conference on Service-Oriented System Engineering (SOSE).

[33]  P. D. Sheth,et al.  Crossover Operators in Genetic Algorithms:A Review , 2015, SOCO 2015.

[34]  Roland R. Draxler,et al.  Root mean square error (RMSE) or mean absolute error (MAE) , 2014 .

[35]  Süleyman Uzun,et al.  A Literature Review on Quadratic Image Filters , 2018, 2018 2nd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT).

[36]  Zhang Bo-hao,et al.  Application of multi population genetic algorithm in traffic assignment problem , 2017, 2017 IEEE 2nd Advanced Information Technology, Electronic and Automation Control Conference (IAEAC).

[37]  A. Ghanbarzadeh,et al.  Application of PSO (particle swarm optimization) and GA (genetic algorithm) techniques on demand est , 2010 .

[38]  Tokunbo Ogunfunmi,et al.  Adaptive Nonlinear System Identification: The Volterra and Wiener Model Approaches , 2007 .

[39]  David Kirk,et al.  NVIDIA cuda software and gpu parallel computing architecture , 2007, ISMM '07.

[40]  John Strassner,et al.  DEN-ng: achieving business-driven network management , 2002, NOMS 2002. IEEE/IFIP Network Operations and Management Symposium. ' Management Solutions for the New Communications World'(Cat. No.02CH37327).

[41]  Guo Wei,et al.  Iteration Time Prediction for CNN in Multi-GPU Platform: Modeling and Analysis , 2019, IEEE Access.