An Augmented Reality-based Intelligent Maintenance Assistance System For Airborne Equipment

With the increase in computing ability of electronic equipment and the cost reduction brought about by the development of technologies,augmented reality technology is becoming more and more widely used in manufacturing.In the current field of airborne equipment maintenance,where a large number of instruments and equipment are used,these maintenance steps are numerous and cumbersome,in addition to problems such as difficulty in consulting manuals and human error.This has led to the fact that only highly trained and experienced experts are qualified for the job.Solving the difficulty in maintenance of airborne equipment,reducing human error,and liberating the generation of maintenance personnel has become a valuable research issue.In order to assist maintenance staff,reduce maintenance thresholds and improve efficiency,this paper proposes an augmented reality-based intelligent maintenance assistance system for airborne equipment.On the one hand,the system uses AR to mix virtual auxiliary information into the real working environment,and guides the maintenance staff to work;on the other hand,through the image detection and identification module,it has the intelligence to actively analyze the maintenance situation.In order to achieve the airborne equipment maintenance auxiliary system with natural interaction and intelligence,the main work of this paper includes:1.For intelligent maintenance assistance issues,an image detection and recognitionmodule is implemented,where the text detection and recognition model used as thecore,including EAST text detection and LSTM text recognition.And a fusion modelis established for the characteristics of airborne equipment.It can analyze the textand specific graphics in the scene,as the basis for the presented system to activelyrecognize the progress in current.2.For the problem of the location of virtual information and the semantic understandingof text,it is proposed to match the image of the instrument and locate its position inthe world space coordinate as a spatial reference point for virtual auxiliaryinformation.The mapping between instruments and text is established according tothe matching result.In addition,the world anchor is introduced to solve the virtualobject drifts problem,which due to the error of the target positioning information,and maintain the result of the target positioning stably.3.Developed an augmented reality-based intelligent maintenance assistant system forairborne equipment.The system integrates image detection/recognition and targetpositioning to achieve intelligent guidance.On the other hand,through augmentedreality display,the virtual information appears in the real world with convincingsensory effects,realizing natural human-computer interaction.This system is used toassist maintenance staff in airborne equipment maintenance,which helps to reducethe difficulty of maintenance work and improve the quality of maintenanceefficiency.