High-Speed Incoming Infrared Target Detection by Fusion of Spatial and Temporal Detectors

Title
High-Speed Incoming Infrared Target Detection by Fusion of Spatial and Temporal Detectors
Author(s)
김성호
Keywords
IR IMAGE SEQUENCES; DETECTION ALGORITHM; POINT-TARGETS; CLOUD CLUTTER; DIM TARGETS; FILTER; CLASSIFICATION; REJECTION; CONTRAST; TRACKING
Issue Date
201503
Publisher
MDPI AG
Citation
SENSORS, v.15, no.4, pp.7267 - 7293
Abstract
This paper presents a method for detecting high-speed incoming targets by the fusion of spatial and temporal detectors to achieve a high detection rate for an active protection system (APS). The incoming targets have different image velocities according to the target-camera geometry. Therefore, single-target detector-based approaches, such as a 1D temporal filter, 2D spatial filter and 3D matched filter, cannot provide a high detection rate with moderate false alarms. The target speed variation was analyzed according to the incoming angle and target velocity. The speed of the distant target at the firing time is almost stationary and increases slowly. The speed varying targets are detected stably by fusing the spatial and temporal filters. The stationary target detector is activated by an almost zero temporal contrast filter (TCF) and identifies targets using a spatial filter called the modified mean subtraction filter (M-MSF). A small motion (sub-pixel velocity) target detector is activated by a small TCF value and finds targets using the same spatial filter. A large motion (pixel-velocity) target detector works when the TCF value is high. The final target detection is terminated by fusing the three detectors based on the threat priority. The experimental results of the various target sequences show that the proposed fusion-based target detector produces the highest detection rate with an acceptable false alarm rate.
URI
http://hdl.handle.net/YU.REPOSITORY/33058http://dx.doi.org/10.3390/s150407267
ISSN
1424-8220
Appears in Collections:
공과대학 > 전자공학과 > Articles
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE