A Model of Saliency-Based Visual Attention for Rapid Scene Analysis
IEEE Transactions on Pattern Analysis and Machine Intelligence
Recognizing Characters in Scene Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Robust Extraction of Characters from Color Scene Image Using Mathematical Morphology
ICPR '98 Proceedings of the 14th International Conference on Pattern Recognition-Volume 2 - Volume 2
Efficient text detection in color images by eliminating reflectance component
ICCSA'07 Proceedings of the 2007 international conference on Computational science and its applications - Volume Part III
Hi-index | 0.00 |
In this paper, we propose the method that can anticipate the changes in illumination, filter out the reflectance component, clustering the colors and use the Saliency map to extract the character region, for the purpose of acquiring the character region extraction which is robust against the variation of illumination. We use outline information at down sampling in difference image by current image and reference images to have worked at continuous image in order to efficiently perceive whether or not there is by the occurrence of camera motion. We judge so that camera motion occurred if a action the average block which you defined is larger the critical value that decided experimentally, and carry out algorithm for global motion compensation. Natural scene images normally have an illumination component as well as a reflectance component. It is well known that a reflectance component usually obstructs the task of extracting and recognizing objects like character in the scene, since it blurs out an overall image. We propose an approach that efficiently removes reflectance components while preserving illumination components. We decided whether an input image hits Normal or Polarized for determining the light environment, using the histogram which consisted of a red component. In the normal image, we acquired the character region without additional processing. Otherwise we removed light reflecting from the object using homomorphic filtering in the polarized image. And then this decided the each character region based on the color clustering technique and the Saliency Map. Finally, we localized character region on these two candidate regions.