Computer Science and Control
Аuthors
Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia
Abstract
Modem vectorizers have low efficiency of recognition of the design documentation of electronic. In this paper I propose an algorithm that increases the efficiency of tracing through the proper labeling of text and graphic components of the binary images oj the design documentation. Application of quadtrees for labeling and extracting text and graphics components on binary images with the following clustering of labeled components to text and graphics is described in the article. Proposed algorithm gains an advantage over sequential application of image components labeling algorithms and algorithms of text extracting in the image. Described algorithm allows to select the text components in noisy images with high reliability.
Keywords:
vectorization, design documentation for electronic, quadtrees, text components extracting, binary images labeling
mai.ru — informational site of MAI Copyright © 1994-2024 by MAI |