Text & Non-Text Segmentation in Colored Images
Nitesh Kumar Singh · 2014
The purpose of this paper color images with complex background for text and non-text segmentation is to propose a new system. The existing text extraction methods in the case of images with complex background do not work efficiently. Locating text in case of variation in style, color, as well as complex image background makes text reading from images challenging. Here the approach used is based on pre- processing steps, edge detection, CC-analysis, bounding rectangles, segmentation and finally extraction of only those blobs which consist of textual part. This approach is tested successfully across various images taken manually and from internet.