The Polish Coins Denomination Counting by Using Oriented Circular Hough Transform

Summary

This paper concerns the coins recognition method, where the modification of the Circular Hough Transform (CHT) has been used. The proposed method allows to recognize denomination of coins in still, clear, blurred or noised images. This paper shows that the Hough transform is an effective tool for coins detection even in the presence of noises such as “salt and pepper” or Gaussian noise. It has been stated, that the proposed approach is much less time consuming than the CHT. In the proposed application, also computer memory requirement is profitable, in contrast to the CHT. In the test procedures, the Polish coins were used and have been recognized and counted. Experiments shown that the proposed modification, achieves consistently high performance compared to commonly used Hough’s techniques. Finally, the proposed approach was compared to the standard CHT dedicated for circular objects. Significant advantages proposed method arise from simplification and reduction of the Hough space. It is necessary to emphasize, that introduced modifications do not have the influence on the objects recognition quality. Presented investigations were carried out for Polish customer.