Application Of The Machine Vision Technology On Appearance Detection-the Case Of Injection Molding Parts (Disc) And Battery

With the continuous development of intelligence industry,machine vision technology was extensively applied for its strong function,high precision,good stability,and so on.Obviously,appearance detection based on machine vision technology is a research focus in recent years.Products' appearance detection is generally divided into two types(flat and curved),as we all know,the hardware structures and software systems are different according to different types.Besides,the traditional positioning algorithm and the regional segmentation algorithm can't work well when detection accuracy is high needed and the detection area is not fixed.Moreover,there are some key points in dimensions measurement apart from defect detection(e.g.how to convert 3D images into 2D image processing).Overall,it is necessary to study the application of machine vision technology on appearance detection.The present investigation studied the appearance detection of injection molding parts(Disc,the flat detection)and battery(a cylindrical surface detection),which is intended to dealing with the problems of high precision positioning,unfixed in the detection area,3D image processing technology,etc.Also,the relevant software were developed at the same time.The main contributions of this paper are the followings:1.Introduce the main hardware components(camera,lens,light source)and software image processing algorithms of machine vision.2.In order to solve the problem of high precision positioning,based on the Disc visual inspection system and template geometry theory,a visual localization algorithm was proposed,which can achieve the appearance detection within 0.02 mm.And based on the visual localization algorithm,the Disc visual inspection software was developed.3.In order to solve the problem that the detection area is not fixed,based on the online visual inspection system of 360 o surface defection of cylindrical battery,HV-VHS adaptive region segmentation and defect detection algorithm was proposed.After its robustness and timed performance were validated,the battery appearance detection software was developed.4.For 3D dimension measurement,a 3D image processing algorithm that combines 2D image processing technology was proposed.By using 2D algorithms,3D images can be positioned quickly with a high degree of detection accuracy of 0.01 mm.The localization algorithm based on template geometry theory,HV-VHS adaptive region segmentation and defect detection algorithm,and 3D images processing technology using 2D algorithms,all of them were applied in engineering practice.The results show that all these algorithms and software can improve production efficiency,moreover,promote the application of machine vision technology on relevant industry.