Visual Servoing on a Trianglar Primitive Using the Moment Approach
| Autor: | Malti Abed, Derrar Yasser |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Image moment
Orientation (computer vision) business.industry Computer science Angular rotation Degrees of freedom ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ComputerApplications_COMPUTERSINOTHERSYSTEMS Visual servoing Moment (mathematics) Center of gravity Image noise Computer vision Artificial intelligence business ComputingMethodologies_COMPUTERGRAPHICS |
| Zdroj: | 2018 International Conference on Signal, Image, Vision and their Applications (SIVA). |
| DOI: | 10.1109/siva.2018.8661150 |
| Popis: | Visual servoing based on points is not reliable since it is very difficult to detect and track points in real applications. However, visual servoing based on contours is more reliable since contours can be more reliably detected. Among the contour based approaches, there is the visual servoing based on moments. In this paper, we propose to study the visual servoing on a triangle primitive using the zero, first and second order image moments. We use the area (zeros order), the coordinates of the center of gravity (first order) and the orientation of the triangle (second order). With this setup we are able to control four camera’s degrees of freedom: the 3D velocity and the angular rotation around the camera axis. We compare this method to the visual servoing on three lines and show that it is more effective and reliable. We also show that when applied on a triangle, the moment based approach is more stable regarding image noise. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|