作者: Asma OujiYann LeydierFrank LeBourgeois
DOI:
关键词:
摘要: In this paper, we introduce a novel color seg- mentation approach robust against digitization noise and adapted to contemporary document images. This system is scalable, hierarchical, versatile completely automated, i.e. user independent. It proposes an adaptive binarization/ quantization without any penalizing information loss. model may be used for many purposes. For instance, rely on it carry out the first steps leading advertisement recognition in Furthermore, segmentation output localize text areas enhance optical character (OCR) performances. We held tests variety of magazine images point up our contribution well-known OCR product Abby Finer- Reader. also get promising results with ad detection large set complex layout testing