Cross-Phase Based Multi-Camera Video Synchronization Calibration with Sub Frame Rate Accuracy

Abstract

In auroral research, it is necessary to accurately measure the time-shift differences between varying light sources. Such measurements can be carried out using two or more digital cameras, or between different regions of an image obtained using a single camera. An example of this is measurements of the time-shifts between prompt auroral emissions originating from different altitude regions during pulsating or flickering aurora, which can give information about auroral electron transport and its chemical compositions. Such measurements require a camera synchronization with an accuracy better than the time delay.

In this thesis, a calibration technique was developed to simulate the time-shifts, and a cross-phase based method was later used to estimate the time-shift, with sub-frame rate precision being designed and tested. This method is based on analyzing the frequency variance of the cross-phase between time-shifted signals. The results showed that with time-series with 60 fps, it is possible to measure the time-shifts as small as 500 microseconds accuracy, and with a time delay uncertainty better than 10 microseconds. The calibration technique was developed to enable measurements of electron transport related to optical emission time differences with high frame rate auroral imaging systems, but the technique can also be applied to a broad range of other applications - such as camera timing calibrations. The calibration technique met the design requirements, and further demonstrated that time-shifts much smaller than the frame rate can be determined.