Title :
Object recognition using polar-exponential grid technique
Author :
Zhang, Hai-yan ; Wang, Dong-mu ; Song, Keou
Author_Institution :
Coll. of Comput. Sci. & Technol., Harbin Eng. Univ., China
Abstract :
A new polar-exponential grid technique for object recognition is proposed in this paper. An object in the rectangular coordinate is mapped into the polar-log coordinate, and the closed profile of the object is transformed a one dimension curve. The structure of the curve has rotate, scale, translate invariance, which can be used in object recognition. Better recognition result using the structure invariant is got.
Keywords :
computer vision; object detection; object recognition; polar-exponential grid technique; polar-log coordinate; rectangular coordinate; Computational modeling; Computer science; Computer simulation; Eyes; Image coding; Image sampling; Object recognition; Optical signal processing; Pixel; Signal sampling;
Conference_Titel :
Image and Signal Processing and Analysis, 2003. ISPA 2003. Proceedings of the 3rd International Symposium on
Print_ISBN :
953-184-061-X
DOI :
10.1109/ISPA.2003.1296390
