Correcting partial errors in faulty facsimile transmitted images with JBIG compression and V.29 modulation

Abstract EN

Facsimile transmission is a lossless process in which the presence of bit errors can severely reduce the readability of an image not only to humans but also would have more drastic effects on the performance of character recognition systems.

In this paper we introduce a correction scheme for bit errors occurring in facsimile transmission under JBIG compression and V.29 modulation.

Our approach relies on knowing the error patterns that might occur and implements an exhaustive replacement algorithm in order to correct erroneous frames.

We describe an error classification for the most occurring errors and their effects, error patterns, and an error correction module that incorporates all the above techniques to correct errors in each data segment.

Our results show a high percentage of success in "correcting" errors and image recovery, and the ability to extend our work to other error classifications and other modulation standards.