Dr. Bull et Dw. Redmill, OPTIMIZATION OF IMAGE-CODING ALGORITHMS AND ARCHITECTURES USING GENETIC ALGORITHMS, IEEE transactions on industrial electronics, 43(5), 1996, pp. 549-558
This paper addresses the application of genetic algorithm (GA)-based o
ptimization techniques to problems in image and video coding, demonstr
ating the success of GA's when used to solve real design problems with
both performance and implementation constraints, Issues considered in
clude problem representation, problem complexity, and fitness evaluati
on methods, For off-line problems, such as the design of two-dimension
al filters and filter banks, GA's are shown to be capable of producing
results superior to conventional approaches, In the case of problems
with real-time constraints, such as motion estimation, fractal search,
and vector quantization codebook design, GA's can provide solutions s
uperior to those reported using conventional techniques with comparabl
e implementation complexity, The use of GA's to jointly optimize algor
ithm performance in the context of a selected implementation strategy
is emphasized throughout and several design examples are included.