A visualization tool for rotation-invariant boundary image matching

Abstract

In this paper, we present a visualization tool that efficiently supports rotation-invariant boundary image matching. Supporting the rotation invariance in boundary image matching is very important for the more accurate and more intuitive matching result. We use the client-server model to design and implement the rotation-invariant matching. First, the client converts a query image to a time-series and sends it to the server. Second, the server performs the rotation-invariant matching by evaluating the query on the database with indexes. Third, the client visualizes the result received from the server. Experimental results show that our system supports the rotation-invariant property efficiently and provides the intuitive analysis through line and polar charts.