Abstract: Degraded document images is very difficult to segment in text format due the high inter /intra variation between the document background and foreground text from different document images. In this paper we propose a binarization technique for recovering degraded document images. In the proposed technique firstly adaptive image contrast map is constructed by giving input degraded document images and detect the text stroke edge pixel. Text of the document is segmented by using local threshold estimation then further applying the post processing to improve document binarization quality. The proposed method is simple, effective and involves minimum parameter tuning.