Applying low rank representation based spatial pyramid matching in welding image classification

Spatial Pyramid Matching (SPM) and related methods have been found perform well in general image classification. A recently proposed improvement on this was LrrSPM, which used low rank representation of encode the SIFT descriptors, to achieve comparable recognition rates, with faster processing speeds. While image classification of this kind has been applied to many fields, an area where this has not been used is in industrial welding applications. This paper attempts to apply LrrSPM to welding image datasets, and show comparable classification results. It also proposes a cognitive architecture involving associative neural networks to perform classification on the images.