Offline Text-Independent Writer Identification Using Local Black Pattern Histograms

The problem of authenticating a writer from his/her writing samples has been the most important and prevalent one subject of active research in the field of handwriting biometrics for the last decade. In this paper, we have focused mainly on the forensic document analysis, more precisely, the offline automatic writer identification in a truly text-independent mode. A new and simple potential textural descriptor has been analyzed for characterizing the handwriting style of the writers, so as to be used to describe the intra and inter-writer variability by calculating the similarity measurements. In order to extract the textural properties from a scanned handwritten sample, an effective statistical texture descriptor is computed from binary connected-components: Local Black Pattern (LBLP). Classification is performed using k-Nearest Neighbors (k-NN) and the Chi-Square (X$2^{}$) distance in a Holdout strategy. The experimental results obtained on two well-known databases show that the proposed scheme achieves a very satisfactory performance and thus reflecting that our approach is still competitive against the state-of-the-art.