Camera Handoff for Multi-camera Surveillance
| Autor: | Vishal R. Satpute, Apurva Jirafe, Mayuri Jibhe |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Background subtraction
Computer science business.industry ComputerSystemsOrganization_COMPUTER-COMMUNICATIONNETWORKS Frame (networking) ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Tracking (particle physics) Object (computer science) Handover Feature (computer vision) Video tracking Computer vision Artificial intelligence business Feature detection (computer vision) |
| Zdroj: | Algorithms for Intelligent Systems ISBN: 9789813348615 |
| DOI: | 10.1007/978-981-33-4862-2_29 |
| Popis: | A handoff algorithm based on SURF is implemented in this paper. In multiple camera video surveillance systems, for consistently tracking an object, we require the concept of handoff. Using multiple cameras improves the flexibility and the area to be tracked which is why we need to consider handoff concept. Handoff or handover is transferring control of a mobile object from FOV of one camera to FOV of another camera. We have performed background subtraction by choosing one image as reference and one as a new frame. The resultant foreground image is used for extracting SURF features. SURF descriptors are saved for tracking the object by performing consecutive frame comparison. The presence of feature location within the handoff boundary indicates the need for handoff, either giving or accepting. The features found in the handoff boundaries of adjacent cameras are compared. If the features match, handoff will be initiated. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|