Paper Info
Title
Image Based Mail Piece Identification Using Unsupervised Learning.
Abstract
Next generation postal sorting machines reuse once extracted mail piece addresses in different sorting steps by means of the mail piece image. Based on the mail piece uniqueness, characteristics derived from the image guarantee the assignment of stored addresses. During the first sorting step mail piece characteristics are extracted and stored together with the target address in a database. In subsequent sorting steps the address is accessed by determining the corresponding mail piece characteristics in the database. Appropriate mail piece image characteristics and procedures for their distance measurement were presented in a previous work. Image based mail piece identification poses a challenge by a constantly changing and non-deterministic mail spectrum and the differentiation of nearly identical bulk mail. In particular, the rejection of unknown mail pieces requires the definition of carefully chosen rejection classes depending on the current mail spectrum. In this paper we present an approach for distance based mail piece identification using a two-stage classification process. Bulk and private mail are handled individually by an unsupervised learning process which clusters similar mail piece characteristics. Based on these clusters specific rejection classes can be estimated within each cluster. The first step in the identification process is the determination of the corresponding cluster for a given mail piece. Using the cluster specific rejection classes a mail piece is either identified or rejected. Experimental results obtained on real-world data sets prove the applicability of the proposed method.
Year
DOI
Venue
2008
10.1007/978-3-642-01044-6_35
ADVANCES IN DATA ANALYSIS, DATA HANDLING AND BUSINESS INTELLIGENCE
Keywords
Field
DocType
Document identification,Unsupervised learning,Adaptive rejection criterion
Distance measurement,Semi-supervised learning,Information retrieval,Computer science,Reuse,Image based,Sorting,Unsupervised learning,Artificial intelligence,Machine learning
Conference
ISSN
Citations 
PageRank 
1431-8814
0
0.34
References 
Authors
6
2
Name
Order
Citations
PageRank
Katja Worm100.68
Beate Meffert223.84