Regeneration of Binary Images From Spoiled Documents A Robust Method for Binary Image Extraction Shekar Kosna Raja1, Teja Sai Krishna1, Kumar Gowru. Rohit1, Balachander Bhuvaneswari2 1Student, Electronics and communication Engineering, Saveetha School of Engineering, Saveetha University Thandalam, Chennai, India 2Assistant Professor, Electronics and communication Engineering, Saveetha School of Engineering, Saveetha University Thandalam, Chennai, India Online published on 16 January, 2018. Abstract This paper proposes a method that retrieves lost information from spoiled documents. Historical documents are of the great importance which has suffered the high degree of degradation. Sometimes these materials undergo contrast problems such as foregrounds ink with different background colour, faded ink, uneven illumination; etc. This paper proposes a model that consists of steps such as pre-processing, main binarisation and finally post-processing. Phase derived feature is used in the first two steps. Specialized adaptive weighted trimmed median filter is used in the post-processing step. Binarisation results are also compared with previously available binarisation methodologies to prove the efficiency of the proposed model. Top Keywords Pre-processing, phase derived feature, Binarization, Adaptive weighted trimmed median filter, post-processing. Top |