Geometrically modeled derivative feature descriptor aiding supervised shape retrieval

摘要

Recent research on shape retrieval render highly acute feature descriptors that are computationally intensive. Hence, a simple approach through a novel tessellated version of the Tetrakis Square tiling scheme for acute feature descriptor aiding supervised shape retrieval is contributed in this paper. The proposed descriptor labeled as Triangulated Second-Order Shape Derivative (TSOSD) performs feature characterization and abstraction by fusing hybrid geometrical concepts with image derivative operators. First, the mechanism tessellates the image into square tiles that are later organized as right-angled triangles. Secondly, the derivatives from the right-angled triangular neighbors interact locally using the trigonometric identities to produce an angle-based feature map. Finally, the feature descriptor is then formulated by local segmentation of the attained feature maps to produce the shape histogram. Experimental results on three standard benchmark databases demonstrate the effectiveness of the proposed approach, particularly rendering a consistent retrieval rate greater than 95% in comparison with the state-of-the-art methods.