Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4972870 | ISPRS Journal of Photogrammetry and Remote Sensing | 2017 | 9 Pages |
Abstract
Due to its sensitive nature against illumination and noise distributions, contrast is not widely used for image description. On the contrary, the human perception of contrast along different spatial frequency bandwidths provides a powerful discriminator function that can be modeled in a robust manner against local illumination. Based upon this observation, a dense local contrast descriptor is proposed and its potential in different applications of computer vision is discussed. Extensive experiments reveal that this simple yet effective description performs well in comparison with state of the art image descriptors. We also show the importance of this description in multiresolution pansharpening framework.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Information Systems
Authors
Sohaib Ali Syed, Muhammad Zafar Iqbal, Muhammad Mohsin Riaz,