Detection and segmentation of tables and math-zones from document images

We propose an algorithm to separate out tables and math-zones from document images. The algorithm relies on the spatial characteristics of tables and math-zones in a document. It has been observed that tables have distinct columns which imply that gaps between the fields are substantially larger than the gaps between the words in text lines and in math-zones the characters and symbols are less dense in comparison to normal text lines. These deceptively simple observations have led us to design a simple but powerful table and math-zone detection system with low computation cost.