Algoritmo genético embarcado aplicado à indução de operadores morfológicos

Abstract

The use of genetic algorithms, evolutionary strategies and genetic programming has been notoriously addressed in the areas of image processing and computer vision and has contributed with diverse results. However, this approach requires large computing power. Some academic papers bring these algorithms for the areas of embedded and dedicated systems, improving in this way the performance of algorithms and minimizing the computational cost. This design addresses the implementation of these algorithms applied to morphological processing of images using FPGAs and Verilog-HDL. The genetic algorithm is adapted for a more intuitive and simple implementation for the development of a dedicated hardware and it may contribute to reduce logical resources and enhance performece. Studies and tests carried out during the project showed good results for the proposed design.