Method for Character Domain Extraction from Image Using Wavelet Transform
Taiki Taniguchi, Yasunari Yoshitomi · Journal of Robotics Networking and Artificial Life · 2015
There has been an increase in the number of images on the Internet that contain private information and/or the URL of an illegal Web site.This information could result in a violation of human rights and/or a crime.We present a method for extracting from an image the region(s) that contain characters.Our method uses the discrete wavelet transform and the empirical knowledge that characters have strong vertical and/or horizontal elements.The experimental results show the usefulness of our method.