Filter-based feature selection for rail defect detection
Machine Vision and Applications
CLAIRE: A modular support vector image indexing and classification system
ACM Transactions on Information Systems (TOIS)
The usage of soft-computing methodologies in interpreting capsule endoscopy
Engineering Applications of Artificial Intelligence
Content-based image retrieval methods
Programming and Computing Software
Minimum explanation complexity for MOD based visual concept detection
Proceedings of the international conference on Multimedia information retrieval
Human-computer intelligent interaction: a survey
HCI'07 Proceedings of the 2007 IEEE international conference on Human-computer interaction
An artificial imagination for interactive search
HCI'07 Proceedings of the 2007 IEEE international conference on Human-computer interaction
Interactive feedback for video tracking using a hybrid maximum likelihood similarity measure
HCI'07 Proceedings of the 2007 IEEE international conference on Human-computer interaction
Visual information retrieval: future directions and grand challenges
VISUAL'07 Proceedings of the 9th international conference on Advances in visual information systems
| Hi-index | 0.00 |
Textures are one of the basic features in visual searching and computational vision. In the literature, most of the attention has been focused on the texture features with minimal consideration of the noise models. In this paper, we investigated the problem of texture classification from a maximum likelihood perspective. We took into account the texture model, the noise distribution, and the inter-dependence of the texture features. Our investigation showed that the real noise distribution is closer to an Exponential than a Gaussian distribution, and that the L1 metric has a better retrieval rate than L2. We also proposed the Cauchy metric as an alternative for both the L1 and L2 metrics. Furthermore, we provided a direct method for deriving an optimal distortion measure from the real noise distribution, which experimentally provides consistently, improved results over the other metrics. We conclude with results and discussions on an international texture database.