An Ancient Degraded Images Revamping Using Binarization Technique
Kaveri Jagtap1, Chandraprabha. A. Manjare2
1Kaveri Jagtap, Department of Electronics and Telecomm Engineering, JSPMs,Jayawantrao Sawant College of Engg, Hadapsar, Pune 28, India .
2Chadraprabha. A. Manjare, Department of Electronics and Telecomm Engineering, JSPMs,Jayawantrao Sawant College of Engg, Hadapsar, Pune 28, India
Manuscript received on January 02, 2014. | Revised Manuscript received on January 04, 2014. | Manuscript published on January 05, 2014. | PP: 1-7 | Volume-4 Issue-6, January 2014. | Retrieval Number: F2453014615/2015©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Revamping of ancient degraded document images is a grueling task due their foreground text and background which is degraded due to uneven illumination, dust, water marks, smear, strain, ink bleed and low contrast etc. The proposed Binarization technique addresses this problem by using adaptive image contrast which is a combination of the local image gradient and local image contrast that is stoic to text and background variation. In the proposed technique, for an input ancient degraded document image an adaptive contrast map is first constructed. The contrast map is then binarized and combined with Canny’s edge map to recognize the text stroke edge pixels. The text of document is further segmented by a local threshold that is concluded based on the intensities of detected text stroke edge pixels within a local window. Dataset of different languages like Modi, Marathi and English are used as input in handwritten and printed form. Modi, Marathi, English database are from year 1908, 1957, 1922.The proposed system is simple, required minimum parameter tuning, and give the superior performance compared with other techniques.
Keywords: Document Image Processing, Document Analysis, Pixel Classification, Degraded Document Image Binarization, Adaptive Image Contrast