Fusion model of inductive thermography and ultrasound for nondestructive testing

Aluminum alloy is widely used in industry. Generally, the ultrasound testing as one of the non-destructive testing technologies has been applied to the metal sheet inspection for guarantying the quality of the production. However there exist challenges if only single ultrasound testing is applied. The near-field area of ultrasonic method remains limitation for detection. This paper investigates a robust data fusion model for comprehensive defects detection by hybriding inductive thermography and ultrasound testing in dynamic scanning. The proposed data fusion architecture is based on feature layer construction and it is able to cover the entire inspection from the surface, near surface and inside of the specimen. The experiment result demonstrates that the measurement of defects position and contour is accurate and reliable. It is easier to achieve automatic inspection through visualization and feature extraction. This method is effective and reliable for comprehensive detect and measurement metal plate.