TY - JOUR AU - Hu, Xiaodan AU - Naiel, Mohamed A. AU - Azimifar, Zohreh AU - Ben Daya, Ibrahim AU - Lamm, Mark AU - Fieguth, Paul PY - 2018/12/24 Y2 - 2024/03/29 TI - Text Enhancement in Projected Imagery JF - Journal of Computational Vision and Imaging Systems JA - J. Comp. Vis. Imag. Sys. VL - 4 IS - 1 SE - Articles DO - UR - https://openjournals.uwaterloo.ca/index.php/vsl/article/view/331 SP - 3 AB - <p>There is great interest in improving the visual quality of projected<br>imagery. In particular, for image enhancement, we would assert<br>that text and non-text regions should be enhanced differently in<br>seeking to maximize perceived quality, since the spatial and statis-<br>tical characteristics of text and non-text images are quite distinct.<br>In this paper, we present a text enhancement scheme based on a<br>novel local dynamic range statistical thresholding. Given an input<br>image, text-like regions are obtained on the basis of computing the<br>local statistics of regions having a high dynamic range, allowing a<br>pixel-wise classification into text-like or background classes. The<br>actual enhancement is obtained via class-dependent Wiener filter-<br>ing, with text-like regions sharpened more than the background.<br>Experimental results on four challenging images show that the pro-&nbsp;<br>posed scheme offers a better visual quality than projection with-&nbsp;<br>out enhancement as well as a recent state-of-the-art enhancement<br>method.</p> ER -