Automatic Retrieval of Shoeprints Using Modified Multi-Block Local Binary Pattern

A shoeprint is a valuable clue found at crime scene and plays significant role in forensic investigations. In this paper, order to maintain the local features of image place pattern block, novel automatic method was proposed, referred as Modified Multi-Block Local Binary Pattern (MMB-LBP). method, images are divided into blocks according two different models. The histograms all first second models separately measured stored feature matrices, respectively. performance evaluations proposed were carried out by comparing with state-of-the-art methods. evaluation criteria successful retrieval rates obtained using best match score rank one cumulative for five matches. comparison results indicated that performs better than other methods, terms complete incomplete shoeprints. That is, able retrieve 97.63% shoeprints, 96.5% toe 91.18% heel Moreover, experiments showed significantly resistant rotation, salt pepper noise, Gaussian white noise distortions