Logpolar sampling and normalization based on boundary crossing for handwritten numerals recognition

This paper presents a new logpolar sampling procedure for recognition of handwritten numerals. It is shown that this approach requires less computation than the logpolar sampling method employed by Duren and Peikari (1991). Furthermore, in addition to the ability of transforming rotational variation to translational variation, it can also reduce the scale variation. This logpolar sampling is used as a pre-processing stage in conjunction with various neural network structures. The results show that it can be used with a two layered sparsely connected neural network to obtain a better recognition rate than previous works. A normalization method based on boundary crossings is also introduced, it is shown that it requires even less computations than the logpolar sampling method and has the ability of reducing the deformation effect found in handwritten characters. Over 16500 character samples are used in conducting the experiments, recognition rates of 96.24% and 95.91% (96.9697% at 374th training epoch) are obtained using logpolar sampling and normalization method respectively with a 4:1 training/testing partition