Detection of table structure and content extraction from scanned documents

Tables are one of the efficient information conveying methods used now days in larger extent. This paper report a fast, language independent (English and Tamil), skilled technique for table structure detection and its content extraction from a scanned document image based on morphological operation, connected components and labeling. From the conducted exhaustive experimentation, it is observed that the proposed method is the fastest approach because of its simple operations. In addition with that it is noticed that it does not lead to any kind of degradation in the extracted table content since after detecting contents location it is retrieved from the original image. More over it is also very interesting to note that the presented approach works well for documents with different font´s size and font styles.