Colorizing Paper Texture of Green-scale Images of Historical Documents

C.A.B. Mello, C.S.V.C. Cavalcanti, and C.A.M. Carvalho (Brazil)


Document processing, Texture Analysis, Artificial Neural Networks.


This paper presents a new method for colorizing images of paper texture of historical documents. The algorithm presented is based on a 256-color image of the texture, some features inherent to the paper texture itself and on Artificial Neural Networks. The final image is perceptually close to the original by visual inspection and quantitatively by the application of ANOVA. Using this coloring process it is possible to create a cluster of textures and to use it to generate different textures from just one sample.

Important Links:

Go Back