Research on the measurement of bubble velocity based on cross-correlation algorithm
Min Chen,Junhua He,Liangyi Chen
DOI: https://doi.org/10.1117/12.694435
2006-01-01
Abstract:Cross-correlation algorithm is based on information theory and stochastic process theory. It has been widely used in many research fields. such as, medical ultrasound, ocean engineering, signal detection, modal parameters under ambient excitation, etc., but its applications in research of bubble curtains arc seldom seen in domestic released periodicals, so our work is attempting to use cross-correlation algorithm. Through computing the velocity vector of bubble curtains. the bubble movement character can be known, i.e. the details of the bubble curtains can also be known. After comparing the differences between it and Doppler method, the cross-correlation algorithm has been applied to the measurement of bubble curtains parameters from a new aspect. The He-Ne laser and high-speed CCD camera are used to acquire the images of dancing bubbles, the velocities of bubbles are computed from image post-processed. Through improving conventional cross-correlation algorithm commonly used for analysis of flow field, the Fast Fourier Transform (FFT) has been used to implement the cross-correlation algorithm rapidly. In order to enhance the computing accuracy, Gaussian curve fitting is used to modify the correlation peak location and the fitting equations are listed, so the bubbles displacement with subpixel accuracy is obtained. Noises are stochastically added from hardware when acquiring images and the cross-correlation algorithm may also introduce errors. The character of velocity vectors result can be entirely wrong with ambient vectors, so they must be corrected. In order to calibrate the cross-correlation algorithm, images with universal displacement are used to validate its feasibility and reliability. The algorithm is applied to the computation of parameters in bubble curtains, yielding the vector graph of bubble motion. The algorithm is expected to be a valuable tool in acquiring the real-time velocity information in bubble curtains.