Overview

Original Research

MANUSCRIPT TEXT RECOGNITION ALGORITHMS RESULTS ANALYSIS

ISKANDAROVA SN 1, and ADIROV TOLLIBOY KHASANOVICH 2.

Vol 17, No 12 ( 2022 )   |  DOI: 10.5281/zenodo.7427443   |   Author Affiliation: Tashkent University of information technologies named after Muhammad al-Khwarizmi 1; Republic of Uzbekistan under the State Tax Committee, fiscal institute Associate Professor of Mathematics and Information Technology 2.   |   Licensing: CC 4.0   |   Pg no: 352-367   |   To cite: ISKANDAROVA SN, and ADIROV TOLLIBOY KHASANOVICH. (2022). MANUSCRIPT TEXT RECOGNITION ALGORITHMS RESULTS ANALYSIS. 17(12), 352–367. https://doi.org/10.5281/zenodo.7427443   |   Published on: 12-12-2022

Abstract

In the case of handwritten text recognition, the image processing, line, word, and letter segmentation algorithms, their processing structures are given, CNN neural network operation mode description is described. Modern hardware has been described as a more improved view of a high-performance product that achieves high results through neural networks. The results are based on the images of the data set available in the neural networks and the data set formed for the Uzbek language. Through this neural network structure, 93% kb recognition efficiency was achieved.

Keywords

initial processing, recognition, segmentation, neural network, CNN