Research on digital indicator recognition algorithm of instrument based on improved LeNeT-5 network

Digital indicator is the common display information of civil aircraft integrated display instrument. Automatic identification of it is very important to automatically test and judge whether the working status of civil aircraft system is normal. At present, the test of civil aircraft system is mainly through manual interpretation of the comprehensive display instrument to display the image and record the display result. The automation degree is not high, and the traditional test record needs to be written by hand. In the case of small changes in the test and mass production, the efficiency of manual test is low. Therefore, taking the main flight display (PFD) in the civil aircraft integrated display instrument system as the research object, this paper studies the automatic recognition technology of digital representation in display images, proposes an intelligent recognition method of digital characters based on LeNet-5 network, and introduces the Dropout layer to improve LeNet-5 network. Finally, the main flight display (PFD) designed by GL studio simulation is used for experimental verification, which proves the accuracy and effectiveness of automatic recognition of displayed image information. Thus, the automatic recognition function of key digital indications and other information in civil aircraft integrated display instrument is realized, which reduces the defects of traditional manual recognition methods and improves the ability and efficiency of testers in processing information.