Paper Info
Title
Vehicle detection and recognition based on a MEMS magnetic sensor
Abstract
Because vehicles moving over ground can generate a succession of impacts on the earth's magnetic field, we can detect them by means of detecting magnetic perturbation using a magnetic sensor, and automatically recognize them by advanced signal processing and recognition method. Comparing to traditional devices, magnetic sensors fabricated with Micro-electro-mechanical system (MEMS) technology is a promising apparatus for vehicle detection, because it is low cost, low power, small volume and light weight. This paper proposes a vehicle detection system structure together with its installation method. The system is based on a MEMS magnetic sensor. This paper also proposes a vehicle detection and noise removal solution based on short-time transform (STFT). Magnetic signals of typical vehicles are researched in order to extract features and recognize targets. In order to enhance the recognition speed, a technique of improved Support Vector Machine is used for recognition. The system and the algorithm have been used for detection and recognition of magnetic signals of vehicle targets in outdoor environment. From experimental results, it can be proven that magnetic signals of moving ground vehicles can be sensed by the MEMS magnetic sensor and detected by the detection system and the detection algorithm. The feature extraction of target magnetic signal is correct and SVM classifier is effective to solve the recognition problem for moving ground targets.
Year
DOI
Venue
2009
10.1109/NEMS.2009.5068605
NEMS
Keywords
Field
DocType
vehicle detection,magnetic sensor,target magnetic signal,magnetic perturbation,detection system,mems magnetic sensor,detection algorithm,vehicle detection system structure,magnetic field,magnetic signal,magnetic resonance imaging,signal processing,support vector machine,support vector machines,classification algorithms,feature extraction
Computer vision,Signal processing,Object detection,Magnetic field,Composite material,Microelectromechanical systems,Support vector machine,Short-time Fourier transform,Feature extraction,Artificial intelligence,Statistical classification,Materials science
Conference
ISSN
ISBN
Citations 
2474-3747
978-1-4244-4629-2
5
PageRank 
References 
Authors
0.70
0
2
Name
Order
Citations
PageRank
Jinhui Lan1216.55
Yuqiao Shi250.70