In this paper, after briefly considering the reasons that prevented th
e development of hybrid systems for signal processing (SP), we point o
ut the requirements for their future exploitation. The need for a bett
er knowledge of different approaches in the scientific community and t
he definition of methodologies for designing hybrid systems' are highl
ighted as two key points. Then we suggest that the well-blown ''task-s
tructure analysis'' design technique should be modified to make it sui
table for hybrid systems. The proposed modification is based on the ma
in need to choose the roles of different approaches and the mechanisms
to integrate them. As an example, we describe the design of a hybrid
system for two-dimensional (2 D) image recognition; the system is base
d on the integration of a numerical, a symbolic, and a connectionist a
pproach. We detail the integration of the symbolic and connectionist a
pproaches to the generation of the models of the objects to be recogni
zed. We describe the main problems involved and the solutions adopted.
In particular, we exploit the synergistic aspects of the two approach
es in oi dei to overcome the bottleneck of knowledge acquisition. Fina
lly, we report experimental results on two applications to show some a
dvantages of the proposed hybrid system.