Word based multiple dictionary scheme for text compression with application to 2D bar code

Abstract

Summary form only given. Research on text compression mainly concerns documentation applications; it has seldomly considered other applications. Significant efforts have previously been made to increase both the data capacity and the information density of bar code symbologies. The results of these efforts created the formats of 2D bar codes. We take PDF417 (Pavlidis et al. 1992) developed by Symbol Technologies as a example. PDF417 is the most popular of the 2D bar code symbologies. However the storage capacity in PDF417 has limited its wider application. Here, we propose a text compression technique with the back searching algorithm and new storage protocols. Studies on how a word-based multiple-dictionary text compression technique can be used to increase the storage capacity in a 2D bar code are described. In order to speed up the search of the text, a hashing function is also described. For application in data base retrieval the proposed technique is particularly useful. For data stored in 2D bar codes which are in the form of limited forms such as part numbers, location, name and reference, the compression ratio can be as high as 2 because the hit ratio can be 100%. For the decoder design, the complexity need not be complex as the decoder just requires to know the 'light' and 'dark'. To let the dictionaries become more 'intelligent', a sub-dictionary is proposed which allows the encoded text to be more independent.