Journal of Computer Technology & Applications
Intelligent Document Scanner
Abstract
AbstractThere is an irresistible trend in the present world for scanning the paper documents and convert it into a digitalized format. Most of these scanning treat the whole document as an entire image. In this scenario, we propose a novel “Intelligent Document Scanner” which automatically segment and classify the contents of the image document including texts, tables and images and store it as a PDF document with three sections that include all the above said contents that were extracted from the image. Segmentation consists of three steps whichconsist of object extraction, object clustering and object filtering. Before object extraction, Mean-Shift filtering is performed for smoothening the document. To perform object clustering K-means algorithm is employed and to find the relationship between kernels, we employed Kernel Propagation algorithm. Classification includes seed point calculation and categorization of contents which is based on weighted priority. The categorized results are then stored as PDF document.Keywords: Mean-Shift filtering, Object extraction, Object Clustering, K-means, Object Filter, Kernel Propagation, Seed Point, Categorization, Weight priorityCite this ArticleDevi Devapal, Aravind MP, Athul P et al. Intelligent Document Scanner. Journal of Computer Technology & Applications. 2016; 7(2): 47–57p.References (9)
- Cheolkon Jung, Juan Liu, Tian Sun, et al. Key Lab of Intelligent Perception and Image Understanding of Ministry of Education of China, Xidian University, Xi’an 710071, China, C. Jung, et al. Automatic Image Segmentation using Constraint Learning and Propagation, Digit Signal Process. 2014; 24: 106–116p.
- Pushmeet Kohli, Philip HS. Torr, Dynamic Graph Cuts for Efficient Inference in Markov Random Fields, IEEE T Pattern Anal . 2007; 29(12): 2079– 2088p.
- Wenbing Tao, Hai Jin, Yimin Zhang, et al. Image Thresholding Using Graph Cuts, IEEE T Syst Man Cy A. September 2008; 38(5): 1181–1195p.
- Luca Bertelli, Baris Sumengen, Manjunath BS, et al. A Variational Framework for Multiregion Pairwise -Similarity-Based Image Segmentation, IEEE T Pattern Anal. August 2008; 30(8): 1400–1414p.
- Imen Karoui, Ronan Fablet, Jean -Marc Boucher, et al. Variational Region -Based Segmentation Using Multiple Texture Statistics, IEEE T Image Process . December 2010; 19(12): 3146–3156p.
- Ismail Ben Ayed , Amar Mitiche , A Region Merging Prior for Variational Level Set Image Segmentation, IEEE T Image Process . December 2008; 17(12): 2301–2311p. Journal of Computer Technology & Application Volume 7, Issue 2 ISSN: 2229-6964(online), ISSN: 2347-7229(print) JoCTA (2016) 47-57 © STM Journals 2016. All Rights Reserved Page 57
- Melih S. Aslan, Ahmed Shalaby, Hossam Abdelmunim, et al. A Probabilistic Shape- based Segmentation Method using Level Sets, IET Comput Vis . September 2013 ; 8(3):182–194p.
- Felipe Calderero, Ferran Marques, Region Merging Techniques Using Information Theory Statistical Measures, IEEE T Image Process . June 2010; 19(6): 1567– 1586p.
- Gökhan Bilgin, Sarp Ertürk, Tülay Yıldırım, Segmentation of Hyperspectral Images via Subtractive Clustering and Cluster Validation Using One -Class Support Vector Machines , IEEE T Geosci Remote. August 2011; 49(8): 2936–2944p. Cite this Article Devi Devapal, Aravind MP, Athul P et al. Intelligent Document Scanner. Journal of Computer Technology & Applications. 2016; 7(2): 47–57p.