Algorithm Symmetric 2-DLDA for Recognizing Handwritten Capital Letters

Abstract

Statistical pattern recognition is the process of using statistical techniques to obtain information and make informed decisions based on data measurements. It is possible to solve the doubt inherent in the objective function of the 2-Dimension Linear Discriminant Analysis by employing the symmetrical 2-Dimension Linear Discriminant Analysis approach. Symmetrical 2-dimensional linear discriminant analysis has found widespread use as a method of introducing handwritten capital letters. Symmetric 2-DLDA, according to Symmetric 2-DLDA, produces better and more accurate results than Symmetric 2-DLDA. So far, pattern recognition has been based solely on computer knowledge, with no connection to statistical measurements, such as data variation and Euclidean distance, particularly in symmetrical images. As a result, the aim of this research is to create algorithms for recognizing capital letter patterns in a wide range of handwriting. The ADL2-D symmetric method is used in this study as the development of the ADL2-D method. The research results in an algorithm that considers the left and right sides of the image matrix, as opposed to ADL2-D, which does not consider the left and right sides of the image matrix. In pattern recognition, the results with symmetric ADL2-D are more accurate