AUTOMATIC RECOGNITION OF DEFECTS INSIDE ALUMINUM INGOTS BY X-RAY-IMAGING

The present paper refers to the non-destructive testing of aluminium i ngots by digital radioscopy while the ingots are moving. Internal defe cts are automatically detected and their type identified, while the ob ject is moving. The defect recognition procedure consists of three ste ps: defect detection, feature extraction and recognition. The detectio n method is based on mathematical morphology. A study of the ingot pro duction and of the defects forming allows one to model the defect type s, and to gather as much as priori information as possible. The select ed features for defect modelling are shape, orientation, width, contra st. The recognition method is based on an identification tree, of whic h inputs are the features, and outputs are the defect types. Once the objects are separated in shape, orientation and width families, a clus tering of aligned and neighbouring objects is performed, in such a way as to gather the objects that would have been disconnected by segment ation. Then, the features of the clustered objects are compared to tho se of the models, yielding the defect type identified.