DEVELOPMENT OF SPACE TIME IMAGE VELOCIMETRY INTRODUCED FAST FOURIER TRANSFORM FOR IMPROVING ROBUSTNESS IN RIVER SURFACE FLOW MEASUREMENT

摘要

As one of the non-contact river flow measurement methods, an image analysis method using space-time image of river surface flow, namely space-time image velocimetry (STIV), has been developed by the authors. In this method, a space-time image (STI) for a line segment placed along the streamwise direction can be generated, whereby which streamwise surface velocity can be extracted from the mean gradient of a striped pattern that appeared in the STI. Originally, STIV applies an image smoothing filter to an STI for improving the measurement accuracy in order to remove non-negligible noise in the STI. However, this still poses difficulty in measuring deteriorated space-time images obtained during adverse weather conditions. In the method presented in this work, we transformed the STI into a frequency domain by using a 2D fast Fourier transformation (FFT) to extract only the information directly related to surface flow. The treatment of STI improves the measurement accuracy and also allows us to analyze wave length and celerity distribution of surface waves generated during floods.