Paper Details: Downloads: 733
Serial Number: P1151133713
Title: Multi-Sensor Image Fusion Using Temporal Object Detection
Authors: M. Altun, M. Celenk
Abstract: This paper describes a new approach to surveillance video content enhancement using multi-spectral image data fusion. Multi-spectral video in this context consists of the visual and infrared (IR) bands. The method takes an unprecedented approach to visual and IR information integration by calculating the estimated background in both bands and detecting changes in the scene under surveillance. The foreground regions are then selected as candidate regions for image fusion at the decision level, which is performed by comparing the edges detected in the foreground with the ones detected in the estimated background. This allows the method to determine the regions which are occluded; hence, recovery is sought via information available in IR images. To highlight the objects of interest foreground regions in the IR frames are clustered and marked with bounding boxes in the resultant fused image.
Keywords: Image Fusion, Video Content Enhancement, Edge Detection, Adaptive Background Extraction, Surveillance, Occlusion Detection
Journal/Conference: ICGST Conference on Computer Science and Engineering, CSE-11
Volume:
Issue:
Submission Date: 8/13/2011 12:00:00 AM
Review Date: 11/8/2011 12:00:00 AM
Publishing Date: 12/19/2011 12:00:00 AM
Article Downloads: 733
Download:

Facebook