In this paper, we present a variation on the LRX (Local RX) algorithm for detecting anomalies in multi-temporal images.
Our algorithm assigns a relative weight to the Mahalanobis distance according to the number of times it appears in an
image. Standard transitions between pixels are therefore not viewed as anomalous; unusual transitions are assigned
proportionally higher weights. Experimental results using our proposed algorithm vs previous algorithms on multitemporal
datasets show a significant improvement.
I. Dayan, S. Maman, D. G. Blumberg, and S. Rotman, "Multi-temporal anomaly detection technique," Proc. SPIE 9987, Electro-Optical and Infrared Systems: Technology and Applications XIII, 99870G (Presented at SPIE Security + Defence: September 28, 2016; Published: 21 October 2016); https://doi.org/10.1117/12.2239530.
Conference Presentations are recordings of oral presentations given at SPIE conferences and published as part of the proceedings. They include the speaker's narration with video of the slides and animations. Most include full-text papers. Interactive, searchable transcripts and closed captioning are now available for 2018 presentations, with transcripts for prior recordings added daily.
Search our growing collection of more than 16,000 conference presentations, including many plenaries and keynotes.