Leak-detection system and methods for leak detection

Abstract

A leak-detection system comprises networked imaging packages including an infrared camera based on microbolometer arrays, one or more filters, and a processor. The processor uses one or more novel processing methods for identifying gas leakages in a monitored facility. In some embodiments, the imaging packages are installed at various locations in the monitored facility. In some other embodiments, the leak-detection system is embodied as a mobile platform, wherein the networked imaging packages are coupled to drones that fly throughout the monitored facility to monitor for gas leakages. The flight path of the cameras may be pre-programmed and may further be alterable in real time. The novel processing technique(s) employed by the leak-detection system disclosed herein involve the use of multiple unique filters that facilitate enhancing the signal-to-noise ratio of captured images; that is, enhancing the contrast of a monitored gas against other objects. The novel processing techniques include one or more of the following: (i) two-image subtraction; (ii) time-resolved imaging; and (iii) multi-filter detection. Additionally, in some embodiments, the imaging packages include a visible (light) camera in addition to an IR camera.