| This thesis deals with the problem of recognition of Urdu characters using Fourier descriptors. In particular, the scope of this work has been to develop a robust Urdu characters pattern classification, representation, and recognition system, which can classify patterns even if they are deformed by transformations like rotation, scaling, and translation or any combination of these, in the presence of noise. Fourier descriptors are used to uniquely describe, classify, and recognize Urdu characters within one sub-class. The Fourier descriptors are used here to describe the shapes for representing Urdu characters. The polygonal approximations of a thinned character are used to accelerate the number of computations involving shape information such as shape recognition. Although current information transmission media i.e. cable, Ethernet etc. may still be used for communications but we proposed new technology i.e. WDM (Wavelength Division Multiplexing) due to its high speed and low loss transmission. Finally experimental results are presented to show the power and robustness of the proposed technique for Urdu character recognition. |
