Scalable geometric hashing on MasPar machines

Scalable data parallel algorithms for geometric hashing are presented. Implementations of the proposed algorithms are performed on MasPar MP-1/MP-2. New parallel algorithms are designed and mapped onto MP-1/MP-2. These techniques significantly improve upon the number of processors employed while achieving superior time performance. The authors´ implementations run on a P processor machine, such that 1⩽P⩽S, where S is the number of feature points in the scene. The results show that a probe of the recognition phase for a scene consisting of 1024 feature points takes less than 50 ms on a 1-K processor MP-1/MP-2