Robust Static Vehicle Detection Method Based on the Fusion of GPS SNR and Magnetic Signal
Autor: | Yong Xiong, Yanliang Jin, Jinyi Zhang, Liangliang Lou |
---|---|
Rok vydání: | 2019 |
Předmět: |
Fusion
business.product_category Computer science business.industry 010401 analytical chemistry Detector Real-time computing Energy consumption 01 natural sciences Wireless access point 0104 chemical sciences Feature (computer vision) Global Positioning System Wireless Satellite Electrical and Electronic Engineering business Instrumentation Energy (signal processing) |
Zdroj: | IEEE Sensors Journal. 19:10111-10120 |
ISSN: | 2379-9153 1530-437X |
DOI: | 10.1109/jsen.2019.2927297 |
Popis: | The wireless vehicle detectors (WVDs) based on the magnetic sensor for outdoor parking spaces have been studied in many papers. However, the magnetic signal has a blind zone between the front and rear wheels of vehicles, thus it is difficult for the WVDs to judge whether the magnetic signal is affected by the vehicle on the current or adjacent parking space. This paper proposes a robust vehicle detection method, which combines the data feature of global positioning system (GPS) satellites’ signal-to-noise ratios (SNRs) and magnetic signals to achieve vehicle detection. Since the high energy consumption of GPS receivers, the reference background of GPS-based method is calculated in the wireless access point (WAP), and the WVDs spend no energy on the background calculation in the proposed method. Moreover, the GPS receiver is activated to assist in the vehicle detection, when the data feature of magnetic signals is insufficient for the WVDs to make an accurate decision on the vehicle status, which improves the vehicle detection accuracy and reduces the average energy consumption of WVDs. Experiments show that the vehicle detection accuracy of our proposed method is up to 99.83%, when the sampling rate of magnetic sensor is 1 Hz, and it has a strong practical significance for the battery-powered WVDs. |
Databáze: | OpenAIRE |
Externí odkaz: |