Wave Front Tracking using Template Matching and Segmented Regression

Abstract

An experimental study was designed to measure shock waves (supersonic gas flow) in a shock tube. A high-speed camera captured images of the shock waves, at a framerate up to 500,000 frames per second. with respect to the huge number of images to be analyzed, an image-processing algorithm was developed for automatic tracking of the shock waves. However, each shock wave might be divided into to two parts; a normal shock (the shock wave is perpendicular to the flow direction), and an oblique shock (the shock is at an oblique angle relative to the flow direction). The proposed framework calculates the characteristics of the wave front, i.e. the angle and velocity of normal and oblique shocks. A technique based on Template Matching and an extended version of Segmented Regression is developed to track the wave front in the high-speed videos. To our understanding, the proposed framework is novel, and our findings are in accordance with results derived from pressure sensors within the test tube.