Address block localization for Chinese postal envelopes with clutter background

Author

Meiling Cheng ; Jinhua Xu

Author_Institution

Dept. of Comput. Sci. & Technol., East China normal Univ., Shanghai, China

fYear

2014

fDate

19-21 Aug. 2014

Firstpage

638

Lastpage

643

Abstract

In this paper we propose a novel supervised model to localize the address block for Chinese postal envelopes. The problem is formulated as a binary classification problem. We get the probability map via joint Conditional Random Field (CRF) training and dictionary learning. Histograms of Oriented Gradients (HOG) are used as descriptors. We evaluate our model on a challenging Chinese postal envelope database with clutter background. Experiment results demonstrate our model performs well and is robust to appearance variations in illumination, rotation, and clutter background.

Keywords

gradient methods; image classification; learning (artificial intelligence); postal services; probability; random processes; CRF training; Chinese postal envelope database; HOG; address block localization; appearance variations; binary classification problem; clutter background; conditional random field; descriptors; dictionary learning; histograms of oriented gradients; illumination; probability map; rotation; supervised model; Clutter; Computational modeling; Computer architecture; Dictionaries; Hidden Markov models; Histograms; Training; Address block localization; Conditional random field; Dictionary learning; Histogram of oriented gradient;

fLanguage

English

Publisher

ieee

Conference_Titel

Fuzzy Systems and Knowledge Discovery (FSKD), 2014 11th International Conference on

Conference_Location

Xiamen

Print_ISBN

978-1-4799-5147-5

Type

conf

DOI

10.1109/FSKD.2014.6980909

Filename

6980909