Research And Implementation Of Key Technologies For Panoramic Interactive Monitoring Platform

Video surveillance systems play an important role in industrial production and social life,and are widely used in many fields such as urban security,education and healthcare,finance and transportation,and property management.With the development of technology and social economy,people have higher requirements for the resolution and field of view of surveillance video images.In many application scenarios today,multiple surveillance devices are often needed to obtain more comprehensive image information.These discontinuous,incomplete,and fragmented surveillance images can lead to limited visibility,weak immersion,low efficiency and concentration,especially in industrial production,which can easily cause safety accidents.This article focuses on the research of the overhead panoramic video and front panoramic video splicing algorithms in the panoramic interactive monitoring platform against the background of remote operation of engineering machinery.The following are the main research contents of this article.Firstly,the domestic and foreign remote driving engineering machinery video surveillance platforms were investigated,and the functional requirements,composition structure,and implementation methods of the panoramic interactive monitoring platform were analyzed.Secondly,the implementation method of the overhead panoramic video splicing was studied.In view of the image displacement and ghosting phenomenon that may occur during overhead panoramic video splicing,an improved feature point selection strategy and two geometric transformation combinations were proposed to obtain a good panoramic bird’s-eye view image.Two splicing methods based on different calibration board schemes were compared and analyzed,and the overhead panoramic video splicing architecture was constructed by analyzing relevant methods such as image interpolation,overhead transformation,and image geometric transformation alignment.The feasibility and good splicing effect of this splicing architecture were verified in experiments,and there was no obvious image distortion in the calibration area.Finally,the implementation method of the front panoramic video splicing was studied.To solve the problem of image splicing seam caused by color and brightness differences in front panoramic video splicing,an image calibration method based on weighted linear function fitting was proposed,which is simple and efficient and can eliminate the image splicing seam in the overlapping area while making the image colors more realistic.By comparing and analyzing relevant methods such as image feature registration and image fusion,the front panoramic video splicing architecture was constructed.A combination of improved weighted linear fusion and optimal seam line method was proposed to solve the problems of image ghosting and displacement caused by linear weighted fusion.The reverse cylindrical projection method was used to solve the barrel distortion of spliced images.The feasibility and effect of the splicing architecture were verified in experiments,and good splicing effects were obtained in environments with different lighting conditions and scene complexities,and there was no obvious image distortion in the spliced images.