A mobile recording system featuring high-precision time synchronization

Abstract

In order to accurately integrate data measured and recorded by combining multiple data loggers and video/audio recording devices in a mobile environment and extract valid information, the location and time information at which data was recorded must be recorded with sufficient accuracy. In recent years, it has become relatively easy to obtain location information with accuracy on the order of cm by using high-precision satellite positioning such as RTK-GNSS. On the other hand, in many cases, a time accuracy of approximately 1 ms or less is required to integrate data recorded while the recording device or recording target is moving and to determine the position with cm-order accuracy. In data recording by simple data loggers, commercially available video cameras, smartphones, etc., it is not easy to achieve a time accuracy of 1 ms due to various limitations such as the accuracy of the built-in clock generator and the effects of communication delays related to synchronization. For example, built-in clock generators in consumer electronics can have errors on the order of 100 ppm due to various factors. This means that even a recording of only 10 seconds cannot guarantee 1 ms time accuracy, an error that cannot be ignored. The problem is further complicated by the addition of error factors based on software architecture and equipment operation. To cope with such problems, professional equipment uses a synchronization signal input to guarantee time accuracy, but this is not a common practice in consumer equipment. Several methods and system architectures proposed in this paper aim to achieve time synchronization between independent data loggers, motion sensors, and different media such as video and audio with errors within 1 ms. In order to record the sampling timing of the system clock and sensors related to the software with an accuracy of less than 1 ms, we explain a method that takes advantage of the fact that the 1-second timing pulse (1-PPS signal) output by the GNSS receiver has an accuracy of about 50 nanoseconds. Furthermore, we propose a method to precisely identify the time of image capture using an optical beacon (called GNSS Clock Beacon (GCB) in this paper) with multiple LEDs driven by the 1-PPS signal. Using this method, we show that it is possible to determine the time of capture from a single image with an accuracy equivalent to the exposure time or less than half of the exposure time.