Image processing in an enhanced and sythetic vision system

Kurzfassung

“Synthetic Vision” (computer generated images using airport databases and the actual aircraft position and attitude data) and “Sensor Vision” (images generated from IR, MMW, video sensors, ...) complement to an ideal system for the pilot´s situation awareness. To fuse these two data sets the sensor images are first segmented by a k-means algorithm and then features are extracted by blob analysis. These image features are compared with the features of the projected airport data using fuzzy logic in order to identify the runway in the sensor image and to improve the aircraft navigation data. This process is necessary due to inaccurate input data i.e. position and attitude of the aircraft. After identifying the runway, obstacles can be detected using the sensor image. The extracted information is presented to the pilot’s display system and combined with the appropriate information from the MMW radar sensor in a subsequent fusion processor. A real time image processing procedure is discussed and demonstrated with IR measurements of a FLIR system during landing approaches.