Character Recognition Algorithm Based on the Central Location and Polar Projection

It is an important approach to locate the centre in circle projection, the accuracy and speed of the central location will affect the recognition rate and the speed of the character recognition. In order to improve the accuracy and speed of the central location, this paper proposes a new algorithm, based on the theory of the triangles and quadrilateral. And also this algorithm implements an invariant descriptor for pattern recognition by using the polar projection and the Discrete Fourier transform. The experimental results show that the performance of the proposed algorithm has good rotate invariance, real-time and high precision property.