Image Patch Extraction in Text Area of Balinese Palm Leaf Manuscripts with Gabor Filters
Made Windu Antara Kesiman, Gede Aditra Pradnyana · 2020
In an effort to build a word spotting and indexation system for the Balinese palm leaf manuscripts, text area detection and image patch extraction in the text area must be done effectively.However, there is no space between words in writing with Balinese scripts.This paper aimed at describing a complete scheme to detect the text area in a Balinese palm-leaf manuscript image and to extract all possible image patches in the manuscript.Gabor filters provide initial information about the document's of text textures.A sliding window algorithm is proposed and is optimized to be able to optimally extract the image patches only in the text line area of the Gabor filtered images.The results show that the combination of Gabor filter with the optimized sliding window algorithm is effectively able to detect and to extract image patches from the text area of the Balinese palm leaf manuscripts.