H.3.11. Vision and Scene Understanding
S. Bayatpour; M. Sharghi
Abstract
Digital images are being produced in a massive number every day. Acomponent that may exist in digital images is text. Textual information can beextracted and used in a variety of fields. Noise, blur, distortions, occlusion, fontvariation, alignments, and orientation, are among the main challenges for ...
Read More
Digital images are being produced in a massive number every day. Acomponent that may exist in digital images is text. Textual information can beextracted and used in a variety of fields. Noise, blur, distortions, occlusion, fontvariation, alignments, and orientation, are among the main challenges for textdetection in natural images. Despite many advances in text detection algorithms,there is not yet a single algorithm that addresses all of the above problemssuccessfully. Furthermore, most of the proposed algorithms can only detecthorizontal texts and a very small fraction of them consider Farsi language. Inthis paper, a method is proposed for detecting multi-orientated texts in both Farsiand English languages. We have defined seven geometric features to distinguishtext components from the background and proposed a new contrast enhancementmethod for text detection algorithms. Our experimental results indicate that theproposed method achieves a high performance in text detection on natural images.
H.3.11. Vision and Scene Understanding
Sh. Foolad; A. Maleki
Abstract
Visual saliency is a cognitive psychology concept that makes some stimuli of a scene stand out relative to their neighbors and attract our attention. Computing visual saliency is a topic of recent interest. Here, we propose a graph-based method for saliency detection, which contains three stages: pre-processing, ...
Read More
Visual saliency is a cognitive psychology concept that makes some stimuli of a scene stand out relative to their neighbors and attract our attention. Computing visual saliency is a topic of recent interest. Here, we propose a graph-based method for saliency detection, which contains three stages: pre-processing, initial saliency detection and final saliency detection. The initial saliency map is obtained by putting adaptive threshold on color differences relative to the background. In final saliency detection, a graph is constructed, and the ranking technique is exploited. In the proposed method, the background is suppressed effectively, and often salient regions are selected correctly. Experimental results on the MSRA-1000 database demonstrate excellent performance and low computational complexity in comparison with the state-of-the-art methods.
