A Vision-Based Technique for in-Flight Measurement of Helicopter Blade Motion

Abstract

The measurement of helicopter main rotor blade angles during flight is a key capability to implement advanced applications, such as strategies for the reduction of emitted noise and to develop innovative flight control laws. The approach proposed in this work for the real-time estimation of blade angles is based on a stereoscopic system mounted on the top of the main rotor and pointing to an optical target placed on the blade root. An advanced image-processing algorithm was developed to match the target features in the left and right camera images, which was required for the 3D reconstruction of the target based on a triangulation method. This algorithm was customized for the target used in this specific application, in order to implement a procedure that is both reliable in blob matching and characterized by a very low computation effort. This allowed the system to speeding up the triangulation procedure aimed at obtaining the 3-D coordinates of the features, in view of real-time applications, even with very compact processing units that can be accommodated on the main rotor head. An inverse problem for the 3-D rotation of the target was solved using the Singular Value Decomposition technique, thus improving the robustness of the measurement. The stereoscopic system was developed in order to be integrated on board an AW139 helicopter main rotor hub, equipped with a synchronous lighting device and a pre-processing unit. The latter enabled the system to automatically extract the