Icing detection and evaluation of the electro-impulse de-icing system based on infrared images processing.

• We suggest an icing detection method based on infrared thermal wave detection. • We introduce the original icing and icing damage detection methods. • We have defined an index for evaluating the de-icing performance of the electro-impulse de-icing system. The aircraft icing poses a significant hazard to aircraft safety and it is important to remove the ice in time. The electro-impulse de-icing (EIDI) system is a mechanical de-icing method, which prevents ice accumulation on that part of the aircraft where ice is more likely to form. Since its discovery research on the theory and application of the EIDI system continued to progress. In this paper, the formation of ice is detected using infrared thermography of the heat signal generated on the aircraft skin surface. Ice accretion specimen are produced to explore the effects of seven critical parameters. The performance of the EIDI system is evaluated using the infrared thermal wave test. The infrared thermography sequence of the ice accretion specimen is acquired by the infrared camera during de-icing operation. The ice detection of the EIDI system employs a method that achieves the original icing detection and icing damage detection based on infrared image processing, including principal component analysis (PCA) and Canny edge detection. The measurement of the original icing on center and radius is achieved by the three-point method, which may be further employed to evaluate the icing manufacturing error. An icing damage detection method is also designed and applied. An index based on the infrared thermal image processing results is introduced to evaluate the performance of the EIDI system under different experimental conditions. The quantitative behaviour of de-icing performance in terms of seven important parameters involved in manufacturing the ice has been verified and discussed. The results show that PCA and Canny edge detection may be effectively exploited to implement ice shape detection and, in turn, to improve the design of the EIDI system. [ABSTRACT FROM AUTHOR]