LOCAL BINARY PATTERN BASED EDGE- TEXTURE FEATURES FOR OBJECT RECOGNITION

In object recognition, there are two sets of edge-texture features and discriminative Robust Local Binary Pattern (DRLBP) and Ternary Pattern (DRLTP). By knowing the limitations of Local Binary Pattern (LBP), Local Ternary Pattern (LTP) and Robust LBP (RLBP). DRLBP and DRLTP are proposed with new features to solve the problem of discrimination between a bright object against a dark background and vice-versa inherent in LBP and LTP. DRLBP also solves the problem of RLBP whereby LBP codes and complements in the specific block are mapped to the same code. Furthermore, the proposed features maintain contrast information for representation of object contours discard by LBP, LTP, and RLBP. These features are tested on seven data sets like INRIA Human, Caltech Pedestrian, UIUC Car, Caltech 101, Caltech256 and Brodatz, with KTH-TIPS2. Results shows that the proposed features outperform the compared approaches on most data sets.