Multi-Modal Sensor Selection with Genetic Algorithms
We develop Genetic Algorithms based method and the tool to select sensors, which provide the specified quality of data after fusion. In this paper, we concentrate on introducing multi-modal sensor fusion in the selection operation. To evaluate data quality, we consider the combination of diverse sensor's accuracy and security metrics. We modify data quality evaluation calculus that incorporates these major metrics to include the possibility of multi-modal sensor fusion. To evaluate Genetic Algorithm feasibility in sensor selection, we compare it against the conventional brute force search approach. To implement our approach and facilitate its use in practice, we produce and release an Android application that automatically selects multi-modal sensors based on the specified sensor types and required data quality.
Chuprov, S., Reznik, L., Khokhlov, I., & Manghi, K. (2022,October 30- November 2). Multi-modal sensor selection with genetic algorithms [Conference paper]. 2022 IEEE Sensors. Doi: 10.1109/SENSORS52175.2022.9967296.