Exploration of advancements in handwritten document recognition techniques

Handwritten document recognition and classification are among the many computers related issues being studied for digitizing handwritten data. A handwritten document comprises text, diagrams, mathematical expressions, numerals, and tables. Due to the variety of writing styles and the intricacy of the written language, it has proven difficult to recognize handwritten material. As a result, numerous handwritten document recognition systems have been developed, each with unique benefits and drawbacks. The paper reviews the evolution of handwritten document recognition in qualitative and quantitative ways. Initially, the bibliometric survey is presented based on the number of articles, citations, countries, authors, etc., on handwritten document recognition in the Scopus database. Later, a survey is done on the learning techniques used for handwritten documents: text recognition, digit recognition, mathematical expression recognition, table recognition, and diagram recognition. This paper also presents the directions for future research in handwritten document recognition.