English Capital Letters Recognition Depends On Computing The Seven Moments

Abstract

There are many digital images for printed documents and the research will continue to arrive to the best
algorithm for identifying the English characters symbols and those letters more frequently. There has been a great
area for these researches. This research uses an image that contains single character each. The proposed
recognition process begins by converting that image into frequency domain by applying discrete cosine
transform(DCT) , computing the seven moments for this image as a features for that character and then building a
database which depends on these features for recognition task. The aim of this research is to prove that each capital
letter has been written with the same font and size but in different directions is also one of the English capital letter
because it has the same seven moment’s values.