CHAMDoc Dataset

This website provides the details of CHAMDoc dataset which we used to evaluate in the paper [1].

CHAMDoc dataset is a collection of inscription images which collected from Sound East Asia. This dataset is done as the part of the work in the ChAMDOC Project.

The dataset is composed of:

·       26 clean inscription images

·       395 line by line images segmented by our annotation.

You may download CHAMDoc dataset from here.


[1]: An effective method for text line segmentation in historical document images.  T.N. Nguyen, J.C. Burie, T.L. Lan, A. V. Schweyer. ICPR 2022.