2D Continuous Wavelet Transform for Pattern Recognition: Application to the Azadi Tower

This paper explores the unique capabilities of the Two-Dimensional Continuous Wavelet Transform (2D-CWT) in Structural Pattern Recognition (SPR). It uses the historic Azadi Tower as a case study due to its diverse structural patterns across various scales and angles. Thus, the novelty of the work is to apply the 2D-CWT for pattern recognition of a historic structure. It examines the effects of scale and angle indices on target pattern extraction and the influence of different wavelet functions on recognition outcomes. Results show that reducing the angle index enhances edge detection, allowing for identifying edges and patterns from multiple directions. Moreover, higher scales of 2D-CWT are sensitive to local abrupt changes in images, underscoring the transform’s dual capabilities. The paper also highlights 2D-CWT’s proficiency in detecting directional edges and singularities.