Font Size: a A A

Research On Stack Filters Optimi Zation Algorithms And Their Implementation

Posted on:2006-12-05Degree:MasterType:Thesis
Country:ChinaCandidate:Y CuiFull Text:PDF
GTID:2168360155968726Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Most of signal processing problems appearing from phenomenon of nature and society are nonlinear, and with the rapid improvement of the precision, flexibility and real time requirement for signal processing, nonlinear digital filters work well in these situations where the linear filters may fail. For this reason, nonlinear digital filtering theory and technology have been developed since 1970's.Stack filters which are a kind of nonlinear digital filters, have received considerable attention during recent years when their theory was presented. Stack filters virtues lie in: applying the architecture of threshold decomposition, suitable to parallel processing and VLSI realization; including many kinds of nonlinear filters, being an important tool for nonlinear research. Therefore, researching on stack filters has important theoretical and practical values. The research fields of stack filters are mainly focused on theirs optimization algorithms and analysis of the properties of their output signals. Beginning with stack filters' fundamental theory, optimization algorithms and their application of stack filters are studied in this dissertation. The main contents and contributions of this paper are showed as following:The fundamental theory of stack filters is systematically summarized in this dissertation. It introduces the threshold decomposition and stacking property of stack filters and the algorithm for generating positive Boolean function.Simulated annealing(SA) belongs to a class of so-called guided random search method. It is effective in optimization problems with multimodal and difficulty in calculating gradients.It requires no gradient and can achieve a globel optimal solution. In this dissertation, we apply it to optimize stack filters and study the performance of optimization algorithms with various noise percentages and MAE and MSE error criteria. The simulation experiments show that optimized stack filters can give satisfactory results.Optimizing Stack Filters by existing Genetic Algorithm easily relapse into local best result. Due to this aspect, in this dissertation, by introducing annealing mechanism to selection operator and selecting cross probability and mutation probability adaptively according to different individual, we bring forward a new optimizing algorithm-Adaptive whole annealing genetic algorithm. Results of experiments show that stack filters optimized by AWAGA under MSE error criteria can suppress noise and protect the details of image effectively.Existing optimizing algorithms of stack filters are mostly genetic algorithm and simulated annealing, but genetic algorithm and simulated annealing have mass calculation and long runtime. Due to this aspect, In this dissertation, we apply a new optimizing algorithm-Tabu search(TS) to optimize stack filters, It can search for the best all-around result in short runtime. By analyzing capability of stack filters optimized by TS under MSE error criteria, we validate the feasibility of this algorithm.
Keywords/Search Tags:Nonlinear filters, Stack filters, Image processing, Simulated annealing, Adaptive whole annealing genetic algorithm, Tabu search
PDF Full Text Request
Related items