The Research Of Detection System Of Helmet-mounted Displays Based On Machine Vision

As an image display and observation device,the helmet mounted display(HMD)uses its optical system to magnify the image of the image source and the target image to into the distant field of view to achieve full image vision.The way it is used is head-mounted,and the user can act freely in space,not only to emancipate the hands and control the action freely,but also to have a high sense of experience.Along with the development of the helmet display,HMD has been widely used in virtual reality,medicine,visualization research,education,electronic games,entertainment,aviation flight training and other fields.Due to the HMD is playing an increasingly important role in our daily life,many researchers make a further exploration on the optical system and manufacturing technology,and make the HMD had a bigger step development.Optical system of the HMD is directly related to the design of the display of the image quality,the weight of the helmet,and user experience,so in the process of design of optical system,the basic performance parameters such as the field of vision,accuracy,aberration,binocular alignment and distortion need to be considered.Therefore,these basic parameters need to be measured in the production of the HMD,which will provide the basis for the next adjustment work.The detection system of Helmet display was developed which combined the technology of visual in this paper.The CCD camera was used to simulate the human eye observe images,and in the case of ensuring the full field of vision to measure the parameters of the field of view,precision,parallax,binocular alignment and distortion completely.In this process,the main research contents are as follows.(1)Design of the overall scheme of detection system of the HMD based on machine vision.According to the application field and using value of the HMD,the overall scheme of the detection system is been designed combined with the basis of forefathers' research,and the suitable and high precision equipment components are been chosen.(2)The detection principle of the system is studied in detail.First of all,the detection principle of HMD optical system is analyzed,then the detection methods of each parameter are introduced in detail.(3)Completion of the equipment construction of the detection system.The hardware equipment of the detection system is mainly divided into five parts: the parallel optical tube system,the image acquisition system,the helmet mounting system,the displacement platform system,and the corresponding user software.(4)Research on image algorithm.The calculation of performance parameters is based on the test pattern collected by CCD camera.The image features are analyzed by denoising,binarization,threshold segmentation and edge detection.The corresponding relation between camera view field and parallel optical tube field is obtained by using method of subpixel reorientation and gray center of gravity.(5)Development of the corresponding software of the detection system.Under the environment of Visual Studio2015,the relative software of the detection system of HMD based on machine vision is developed by using C# language,and the basic functions such as the starting,operation,parameter setting and closing of the camera and motion controller are realized.The center of CCD camera is automatically aligned with the eye loci of helmet mounted display,and the measurement based on the full field of view is ensured by pattern search algorithm.The measurement functions of field of view,accuracy,parallax,binocular alignment and distortion are realized.Then the data reports and graphics of the measured results are stored in the corresponding folders to be used for reference and the relevant experimenters.(6)Provide complete fault diagnosis module.It mainly monitors the normal start-up and operation of the camera and the displacement platform.This system provides the user manual and the use method of the detection system.And this section gives the possible failure phenomenon and the solution.This paper has completed the development of the detection system of HMD based on machine vision and the corresponding user software successfully,and the accuracy of the system is evaluated,and the positioning accuracy is18?.And the possible error is analyzed,the motion accuracy of the guide rail is less than1?,and the parallax of the paralleled optical tube is not more than1.5?.The detection system is used to calibrate the eye loci and measure the five performance parameters in the case of the integrity of the camera's vision,and the results are expressed in a visual form(data reports and graphics).In the user software,the initial diagnosis function is included,which makes the system intelligent,safe,accurate and reliable.At present,the system has been put into production.