Identification of Persian handwritten cultivars using the bat algorithm

Recognition of handwritten numbers is a complex issue in the field of image processing. Due to being in the digital age, speed and accuracy of intelligent systems have always been developing, which is why image processing has become particularly important. On the other hand, due to the importance of handwritten numbers recognitions and its various applications, in this study, an algorithm has been proposed to recognize Persian handwritten numbers. In this research, two methods of “framing and wavelet transform” have been used to extract the feature. The KNN classification method is also used. After testing different dimensions of the mold, the accuracy percentage of the framing method has been obtained above 87% By selecting the features of the third level of wavelet transform, 95% accuracy is obtained. Using the evolutionary bat algorithm, it has been tried to reduce the features and improve the system.