Sparsity based signal processing is a relatively new research area which has attracted tremendous interest from researchers. Application areas for sparse signal processing include but are not limited to image processing, pattern recognition and computer vision. This work considers the joint application of sparsity and kernel methods to classification problems. Novel sparsity based classifiers have been effectively utilized in classification. Variants of sparse classifiers utilizing kernel functions on the other hand have garnered limited interest. Here we will examine the combination of non-dictionary learning sparse classifiers with kernel based methods. Simulations in face and digit recognition applications demonstrate competitive performance for classifiers utilizing sparsity and kernel methods concurrently.