研究目的
To develop an indoor navigation system that exploits peak intensities of unmodulated luminaries to create a virtual graph representation, avoiding the need for floor maps, localization systems, or additional hardware.
研究成果
PILOT provides an effective and efficient indoor navigation system without requiring floor maps or localization systems, leveraging unmodulated light intensities and mobile sensors. It demonstrates high accuracy in peak detection and overlapped segment identification, with significant time savings and low power consumption compared to existing methods.
研究不足
Potential issues with false positives/negatives in peak detection due to factors like device orientation, walking speed, and environmental interferences; reliance on crowdsourced data which may be incomplete; challenges in dynamic environments with sunlight or non-functional luminaries.
1：Experimental Design and Method Selection:
The system, PILOT, uses light intensity peaks from unmodulated luminaries detected by mobile device sensors to construct a virtual graph. Methods include IIR filtering, moving average for light intensity smoothing, and dynamic time warping (DTW) for path merging.
2：Sample Selection and Data Sources:
Experiments conducted in three indoor environments: a supermarket, a shopping mall, and an office building, with participants using various mobile devices.
3：List of Experimental Equipment and Materials:
Mobile devices (Huawei Mate 8, Huawei P9, Samsung Galaxy S5, Google Nexus 9) with built-in sensors (light sensor, gyroscope, compass, barometer).
4：Experimental Procedures and Operational Workflow:
Participants walk while holding devices to collect sensory data; data is uploaded to a cloud server for graph generation; navigation paths are computed and displayed on devices.
5：Data Analysis Methods:
Statistical analysis of peak detection accuracy, overlapped segment detection, time savings, delay measurements, and power consumption comparisons.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
