Main Article Content
Abstract
Handwritten Alphabet Recognition is one of the artificial intelligence applications which provides an
important fundamental for various advanced applications, including information retrieval and human-computer
interaction applications. This paper elaborates a research which is performed to build a system which is able to
recognize handwritten latin alphabets in the form of images. The system is developed using the Hamming
Network method.
From the experiments of 100 prototypes of data, the system is able to identify handwritten latin alphabets
with 76.97% average accuracy. Mistaken recognitions are mainly caused by the similarity of the alphabet
patterns, such as the pattern of I, T and Z, or the pattern of C and G. The value of ε and the matrix b which are
given in the system may affect the recognition results.
Keywords: Handwritten Alphabet Recognition, Hamming Network
important fundamental for various advanced applications, including information retrieval and human-computer
interaction applications. This paper elaborates a research which is performed to build a system which is able to
recognize handwritten latin alphabets in the form of images. The system is developed using the Hamming
Network method.
From the experiments of 100 prototypes of data, the system is able to identify handwritten latin alphabets
with 76.97% average accuracy. Mistaken recognitions are mainly caused by the similarity of the alphabet
patterns, such as the pattern of I, T and Z, or the pattern of C and G. The value of ε and the matrix b which are
given in the system may affect the recognition results.
Keywords: Handwritten Alphabet Recognition, Hamming Network