Multiresolutional Multisensor Target Identiication
نویسندگان
چکیده
An algorithm for multiresolutional multisensor target identiication is presented. It uses the computationally eecient scale sequential approach to hypothesis testing during identiication and implements the inverse discrete wavelet transform to fuse data from diierent multires-olution sensors. We show that fusion leads to an increase in the probability of correct identiication without a signiicant increase in the number of computations.
منابع مشابه
Reflective Negotiating Agents for Real-Time Multisensor Target Tracking
In this paper we describe a multiagent system in which agents negotiate to allocate resources and satisfy constraints in a real-time environment of multisensor target tracking. The agents attempt to optimize the use of their own consumable resources while adhering to the global goal, i.e., accurate and effective multisensor target tracking. Agents negotiate based on different strategies which a...
متن کاملA Sketch of Multiresolutional Decision Support Systems Theory
Multiresolutional Decision Support Systems gain better performance and higher accuracy by the virtue of building highly efficient multiresolutional representation and employing multiscale Behavior Generation Subsystem (Planning and Control ). The latter are equipped by devices for unsupervised learning that adjust their functioning to the results of self-identification. We show planning and lea...
متن کاملMultisensor Multitarget Mixture Reduction Algorithms for Tracking
A single-sensor single-target Mixture Reduction (MR) data association algorithm is extended for use in multisensor multitarget tracking situations. MR is extended for tracking an arbitrary number of targets using an arbitrary number of sensors under the assumption that the sensor measurement errors are independent across sensors. Like the single-sensor single-target MR algorithm, which gives be...
متن کاملA Multiresolutional Coding Method Based on SPIHT
In this work, we incorporate a multiresolutional coding functionality into the SPIHT algorithm [1]. The multiresolutional coding can be considered as a kind of the region of interest (ROI) coding with multiple regions of interest. Therefore, the ROI coding proposed by authors can be extended for the multiresolutional functionality without any cost in performance. The parent of ROI (PROI) and th...
متن کاملModel Theory Based Fusion Framework with Application to Multisensor Target Recognition
In this work, we present a model theory based fusion methodology for multisensor wavelet-features based recognition called Automatic Multisensor Feature-based Recognition System (AMFRS). The goal of this system is to increase accuracy of the commonly used wavelet-based recognition techniques by incorporating symbolic knowledge (symbolic features) about the domain and by utilizing a model theory...
متن کامل