An advanced genetic algorithm for designing 2-D FIR filters

In this paper, a new genetic algorithm is proposed for designing 2-D FIR filters with the objective of being relevant and tractable. The key point of our approach stems in the capacity of our GA to adapt the genetic operators during the genetic life while remaining simple and easy to implement. It hybridizes the use of conventional and dedicated processes. The first are capable of covering a wide range of optimization problems without being sufficient. The second includes the filter design specificities and takes feedback from the current state of the genetic search. We demonstrate the method efficiency on various cases of filter design.