Processing Infrared Images For High Speed Power Line Inspection
Reza Safabakhsh · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1989
This work examines automation of fault detection in infrared images of power transmission lines. Several levels of processing for semi- and full automation of the task are proposed and the first processing level is implemented. A global adaptive thresholding technique which adapts to each new tape is used for the real-time detection of potential fault spots in the taped imagery. Timing considerations for fault detection and image saving are discussed and the impact of hot background objects on the system performance is explained.