Translator Disclaimer
Paper
12 October 2010 A space variant maximum average correlation height (MACH) filter for object recognition in real time thermal images for security applications
Author Affiliations +
Abstract
We propose a space variant Maximum Average Correlation Height (MACH) filter which can be locally modified depending upon its position in the input frame. This can be used to detect targets in an environment from varying ranges and in unpredictable weather conditions using thermal images. It enables adaptation of the filter dependant on background heat signature variances and also enables the normalization of the filter energy levels. The kernel can be normalized to remove a non-uniform brightness distribution if this occurs in different regions of the image. The main constraint in this implementation is the dependence on computational ability of the system. This can be minimized with the recent advances in optical correlators using scanning holographic memory, as proposed by Birch et al. [1] In this paper we describe the discrimination abilities of the MACH filter against background heat signature variances and tolerance to changes in scale and calculate the improvement in detection capabilities with the introduction of a nonlinearity. We propose a security detection system which exhibits a joint process where human and an automated pattern recognition system contribute to the overall solution for the detection of pre-defined targets.
© (2010) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Akber Gardezi, Ahmed Alkandri, Philip Birch, Rupert Young, and Chris Chatwin "A space variant maximum average correlation height (MACH) filter for object recognition in real time thermal images for security applications", Proc. SPIE 7838, Optics and Photonics for Counterterrorism and Crime Fighting VI and Optical Materials in Defence Systems Technology VII, 78380N (12 October 2010); https://doi.org/10.1117/12.865485
PROCEEDINGS
14 PAGES


SHARE
Advertisement
Advertisement
Back to Top