Vision Based Displacement Sensor with Heat Haze Filtering Capability

Abstract

The vision based sensors enable easy-to-install, non-contact, and low-cost structural displacement monitoring with high accuracy. The heat haze, however, will affect the accuracy of the vision based displacement sensors therefore affect the fidelity of the structural dynamic analysis. Heat haze effect was observed in the field tests and the image distortions due to heat haze were clearly seen in the videos. To eliminate the heat haze is vital in enabling vision displacement sensor to be used for long-term monitoring in hot weather, especially for long-distance multi-point monitoring. Few studies have been conducted on eliminating the effect of heat haze in vision based displacement sensor. This article presents a vision based displacement sensor incorporated with heat haze filtering capability. A heat haze filtering technique for the vision based displacement monitoring is presented based on a distortion rectification method and a new heat haze distortion base. The filtering technique is a statistical technique that first produces many sample images with simulated distortions with known warping parameters from an undistorted reference image using a distortion base. Then a distorted image obtained by the video camera is matched with the most similar sample image in terms of the shortest Euclidean distance. The warping parameters of the matched sample image is used to warp back the distorted image and to obtain a rectified image. The matching and warping process is repeated until the distance between the rectified image and the reference image falls below a threshold. The rectified image is used to extract heat haze filtered displacement measurement. In this paper, the effect of heat haze will first be detected and quantified. Then the heat haze filtering technique based on the new heat haze distortion base, will be introduced and implemented in the vision sensor. The performance of the heat haze filtering technique will be evaluated through laboratory tests and a field test with heat haze, by comparing the displacement results with and without the heat haze, with and without the heat haze filtering technique