Loading...
Please wait, while we are loading the content...
Similar Documents
AnySURF: Flexible Local Features Computation
| Content Provider | CiteSeerX |
|---|---|
| Author | Sadeh-Or, Eran Kaminka, Gal A. |
| Abstract | Abstract. Many vision-based tasks for autonomous robotics are based on feature matching algorithm, nding point correspondences between two images. Unfortunately, existing algorithms for such tasks require signi cant computational resources and are designed under the assumption that they will run to completion and only then return a complete result. Since partial results a subset of all features in the image are often su cient, we propose in this paper a computationally- exible algorithm, where results monotonically increase in quality, given additional computation time. The proposed algorithm, coined AnySURF (Anytime SURF), is based on the SURF scale- and rotation-invariant interest point detector and descriptor. We achieve exibility by re-designing several major steps, mainly the feature search process, allowing results with increasing quality to be accumulated. We contrast di erent design choices for AnySURF and evaluate the use of AnySURF in a series of experiments. Results are promising, and show the potential for dynamic anytime performance, robust to the available computation time. 1 |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Signi Cant Computational Resource Anytime Surf Autonomous Robotics Point Correspondence Complete Result Flexible Local Feature Computation Partial Result Di Erent Design Choice Dynamic Anytime Performance Many Vision-based Task Rotation-invariant Interest Point Detector Su Cient Surf Scale Available Computation Time Additional Computation Time Feature Search Process Exible Algorithm Re-designing Several Major Step |
| Content Type | Text |
| Resource Type | Article |
