A Finding Method of Shape Theme Based on Wavelet Time Series

Finding shape theme has raised great attention in the database of shapes. According to the problem of incompatible about accuracy and complexity in the shape theme search algorithm, this paper proposed a finding theme algorithm using the multi-resolution analysis of wavelet and the processing capability of reduction dimension of time sequence, accurately calculated the similarity between different objects combining with the Euclidean distance formula, and achieved satisfactory results. Through the comparison between the real data sets to test and traditional shape theme algorithm, it shows that the method has good stability and reliability, and ensure the real-time processing ability of the closed contour shape's overall matching.