A new shadow detection and depth removal method for 3D text recognition in scene images

Document Type

Conference Article

Publication Title

ACM International Conference Proceeding Series


Text recognition in 3D scene images is challenging and interesting due to presence of shadow, which hinders problems for extracting distinct features of recognition. In this paper, we present a new method for shadow detection and removal to enhance recognition performance for 3D texts in natural scene images. The proposed method convolves Gabor kernel with gray values of each image for separating shadow pixels from texts and background pixels. Based on the responses of Gabor filters, the proposed method analyzes characteristics of shadow and background pixels to remove depth information, which results in a 2D image. Experimental results on 2D and 3D text images show that the proposed method is useful in improving performance of 3D text recognition in terms of recognition rate. To our knowledge this is the first to attempt towards 3D text recognition by using shadow detection and depth removal method.

First Page


Last Page




Publication Date


This document is currently not available here.