Portal de Periódicos da CAPES

On the Best Evolutionary Wavelet Based Filter to Compress a Specific Signal

2010; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-642-16773-7_34

1611-3349

Oscar Herrera-Alcántara,

Image Enhancement Techniques

This work presents an application of a genetic algorithm in the design of digital filters used to implement the discrete wavelet transform. The best compression of a transformed signal is achieved when its power is described by the smallest number of transformation coefficients. The individuals of the genetic algorithm aim to reach this target, at the same time that they reduce the error of the reconstructed signal. In the experiments we worked with grayscale images, and we compared the performance of evolutionary and Daubechies filters. Experimental results show the feasibility and convenience of finding custom wavelets for each image, and support the idea that there is a suitable wavelet to compress any given signal.