A cross-connected components-based layout analysis algorithm for Chinese business card
Huiying Zhu, Yuexian Zou · 2008
In this paper, a cross-connected components-based layout analysis algorithm for Chinese business card (CBC) is presented. The major aim of our scheme is to extract important personal information such as the holder's name, title, address, phone number, etc. from the CBC image. Though much work has been done on layout analysis at present, this task is still very difficult considering the characteristics of the CBCs, which is generally complex layout, mixed Chinese-English characters and diverse typesetting. Conventionally, detection of connected components (CCs) of CBCs is based on 4 or 8-connectivity, both of which have a problem of producing lots of small CC pieces, making it difficult to merge them into a line. To solve this problem, we proposed a novel cross-connected component extraction algorithm (CCCE) along with a Center-Height Component Mergence (CHCM) algorithm for CBC layout analysis. We implemented our method on a common personal computer. The experimental results show that our method can greatly reduce the computational complexity and exhibits high CBC layout analysis accuracy.