Abstract: This paper presents the robust object recognition using Discriminative Robust Local Binary Pattern (DRLBP) and Discriminative Robust Local Ternary Pattern (DRLTP) methods for feature extraction. The system proposes new approach in extension with local ternary pattern called DRLTP and DRLBP. By using these methods, the category recognition system will be developed for application to image retrieval. The category recognition is to classify an object into one of several predefined categories. DRLTP & DRLBP is used for different object texture, edge contour and shape feature extraction process. It is robust to illumination and contrast variations as it only considers the signs of the pixel differences. The proposed features retain the contrast information of image patterns. They contain both edge and texture information which is desirable for object recognition. The DRLBP & DRLTP discriminates an object like the object surface texture and the object shape formed by its boundary. The boundary often shows much higher contrast between the object and the background than the surface texture. Differentiating the boundary from the surface texture brings additional discriminatory information because the boundary contains the shape information. These features are useful to distinguish the maximum number of samples accurately and it is matched with already stored image samples for similar category classification. Our proposed features are compared with two classifiers and results are tested on five datasets: WANG, Caltech 101, Caltech 256, VOC 2005, and UIUC.
Keywords: Test image, Preprocessing, Feature Extraction, Database Training, Classification, Parameter analysis.