A Novel Similartiy Measure for Point Matching using Hilbert Curve

摘要

In this report, we present a novel similarity measure using Hilbert curve for point pattern matching. In our method, the similarity measure is computed in one-dimensional (1-D) sequence in stead of in two-dimensional (2-D) space by using Hilbert, curve. The experimental results show that our measure is fast and robust to noise than conventional similarity measures.