An Invariant Approach for Image Recognition
Erick Zuniga-Segura, Guillermo Sánchez-Díaz · 2006
Shape-of-object representation has always been an important topic in image processing and pattern recognition. This work deals with representation of shape of objects, and approaches to recognize objects. Several invariant techniques are widely used to represent an object because they preserve information and allow considerable data reduction. In this paper, a new approach based on a code representation and testor theory is presented. The proposed method is invariant under translation, scaling and rotation. Also, the paper discusses the capabilities of the method in recognizing objects. In addition, results using simple figures classes are show