Classification of Similar Metallic Materials Based On Non-contact Ultrasound

To achieve accurate classification of similar metallic materials, this paper proposes a classification of similar metallic materials based on non-contact ultrasound. First, an experimental setup is used to collect ultrasonic echo signals from different metallic materials. Then, filtering and wavelet transform are applied to the original signals for noise reduction. Subsequently, a support vector machine (SVM) algorithm is used to classify the signals after wavelet transform. In the experiment, four different types of aluminum alloy plates were selected, and 100 ultrasonic echo signals were collected for each type of alloy plate. The signals were processed and classified using the proposed method. Experimental results show that the classification accuracy of this method for similar metallic materials can reach above 97.5%, effectively distinguishing different types of similar metallic materials.