Angle weighted sign coherence factor-based ultrasonic fusion imaging for crack detection

Abstract

This paper proposes a novel ultrasonic fusion imaging method to enhance crack surface characterization while mitigating image distortion issues commonly associated with phase coherence imaging during signal-to-noise ratio improvement. The proposed method first applies Gaussian model-based correction to array data collected from multiple scanning positions and establishes a vector field corresponding to crack orientation. An Angle Weighted Sign Coherence Factor (AWSCF) is then developed based on this vector field to facilitate weighted image fusion, yielding the final crack visualization. Experimental validation was performed using a 5 MHz array probe on aluminum alloy specimens containing cracks of varying orientations. Results demonstrate that the proposed method achieves superior crack surface continuity characterization compared to conventional total focusing method (TFM) imaging with a single linear array. Furthermore, the method maintains stable imaging performance at high signal-to-noise ratios while effectively preventing crack image distortion.