A code error correction system for PDF documents using regex and similarity matching

This paper designs a code error correction system for PDF documents to promote the informatization and automation of work. Its core function is to regularize the text extracted from PDF drawings, and then to correct the error of the extracted contents using similarity matching method, thereby improving the accuracy of PDF document code extraction. The system proposed in this paper is based on regular expression (i.e. regex) and similarity matching method. Regex can realize the preliminary filtering of the extracted contents, and remove the character interferences outside the character set of the presetting code database. The improved edit distance algorithm is used to obtain the correct code closest to the extracted content in code database. The experimental results show that the system can be applied to engineering practice. The regex and the improved similarity matching method can effectively correct the extracted contents that are inconsistent with the code database.