Design of Image Recognition Algorithm for Critical Flight Parameters in Civil Aircraft Display System

Abstract

In the current civil aircraft display system design, complex COTS devices such as CPU and GPU have been widely used. While improving product performance, it also greatly increases the difficulty of airworthiness certification. Taking the display system as an example, the data integrity of the critical flight parameters which is displayed on the display unit must reach 1.0E-7 or higher, and the development assurance level of each software and hardware module must reach level B or higher. However, according to industry experience, the data integrity of CPU and GPU can only reach the level of 1.0E-6. Therefore, for the critical flight parameters in the display system, it must be monitored to cover the entire data path from data processing to image generation to prove the data validity. Aiming at the numeric critical flight parameters of civil aircraft display system, this paper proposes a numeric image recognition technology realized by FPGA image processing function, which can be used to monitor the integrity of numeric critical flight parameters, and this technology has been performed and verified in real display unit.