Background Invariant Laser-Spot Detection and Tracking for Embedded Systems | Consortium for Embedded Systems | SIU

Abstract: The aim of this project is to develop a strategy to accurately detect a laser-spot in low-resolution images and to track the laser-spot in varying background and illumination conditions. Detection will be based on feature extraction and classification. The tracking techniques will be based on motion-related segmentation as well as the Kalman and particle filters. The initial focus will be on detecting the laser-spot in static and simple backgrounds. Subsequent efforts will focus on dynamic, complex, and noisy backgrounds. The final goal will be to improve the computational efficiency by embedding the detection and tracking strategy into a multi-core processing architecture.

Problem: This project will focus on developing a strategy to detect and track laser-points in varying background and illumination conditions.

Rationale / Approach: Lasers are used extensively in the development of smart munitions for targeting and for guiding munitions. It is equally important to be able to detect and track lasers in counter-measure applications. The approach is to formulate the detection problem in terms of classification and develop tracking algorithms based the Kalman and particle analysis filters.

Novelty: The topic of this project was proposed by Rockwell Collins. They anticipate numerous applications that will benefit from the successful development of novel algorithms to detect and track laser-spots.

Potential Member Company Benefits: Rockwell Collins has indicated that they envision several novel applications resulting from the development of the proposed laser-spot detection and tracking algorithms.
Deliverables for the proposed year: Detection and tracking algorithms.

Milestones for the proposed year: Develop algorithms to:
(a) Detect laser-spots in a sequence of images in varying illumination and backgrounds.
(b) Track laser-spots in a sequence of images in varying illumination and backgrounds.
(c) Evaluate the detection and fusion algorithms on a wide range of images.