This paper proposes the use of image fusion for merging pulse-echo ultrasound and vibro-acoustography data of a metallic sample with machined damages employing algebraic, Boolean, and principal component analysis fusion methods, to enhance the measurement results. Image fusion is an interesting alternative in this sense since it is easy to use, may provide efficient results, and its processing can be made a posteriori, without the need for raw data. After binarization with a combination of image processing tools, the damages were measured using Hough transform and the measurement errors related to their nominal values were analyzed. Two fusion rules of the algebraic fusion method provided the smallest measurement errors, with values 49% and 39% smaller than the ones of the ultrasound technique which provided the best results in isolation. The improvement in the results encourages the use of the proposed fusion methods in non-destructive testing applications.

本文提出了利用图像融合技术融合具有机械损伤的金属样品的脉冲回波超声和振动声像图数据的方法，采用代数，布尔和主成分分析融合方法，以提高测量结果。从这个意义上讲，图像融合是一种有趣的替代方法，因为它易于使用，可以提供有效的结果，并且可以将其处理后验，无需原始数据。在使用图像处理工具组合进行二值化之后，使用霍夫变换测量损伤，并分析与标称值相关的测量误差。代数融合方法的两个融合规则提供了最小的测量误差，其值分别比超声技术的49％和39％小，后者在隔离方面提供了最佳的结果。结果的改进鼓励在无损检测应用中使用建议的融合方法。