Research On High-speed Acquisition And Real-time Processing Technology Of Molten Pool Image

Laser additive manufacturing technology of metal is a new processing mode developed rapidly in recent years.Compared with the traditional "subtracted manufacturing" using CNC machine tools,laser additive manufacturing has the advantages of greatly reducing material waste,reducing production cost,large-scale production and processing complex workpiece.At present,metal additive manufacturing is mainly used in military industry,aerospace and other fields with high requirements for workpieces.In recent years,it is also widely used in medical fields with high requirements for precision.In addition,with the development of machine vision technology,the industry also tends to detect the processing process to avoid the waste of human and material resources after forming.At the same time,improve the level of intelligence and automation of the system.Therefore,there is an urgent need for an effective way to monitor the processing process in real time.It is a widely used detection method to collect the image of the processing process by CCD and analyze it by image processing algorithm.It is more and more frequently used in the detection of complex industrial processing.The thesis mainly focuses on the change of molten pool morphology in the process of metal laser additive manufacturing.Combined with image collection,hardware acceleration,image processing and other technologies,the detection of molten pool characteristics is completed.Then,a new way to detect spatter and other defects is proposed.This thesis mainly completes the following work:(1)In the project,the vision monitoring system were designed of the Laser Engineered Net-shaping system(LENS)based on coaxial powder feeding and the Selective Laser Melting system(SLM)based on powder spreading.Based on the principle of optical detection,the optical path is designed to monitor the additive manufacturing process.(2)In view of the coaxial powder feeding laser engineered net-shaping system,the infrared camera with lower frame rate is selected according to the characteristics of slow processing.The collected coaxial analog signal is converted into digital image signal by BNC to USB high-definition video acquisition card,and the SDK of the acquisition card is used as the platform,VC++ is used as the programming language,combined with filtering,morphological processing,segmentation,feature extraction and other image processing technology,design online monitoring algorithm for molten pool morphology of LENS system,and finally the image of molten pool is classified as normal trailer and spatter.(3)In view of the powder spreading Selective Laser Melting system,the super high frame rate high-speed camera is selected according to the characteristics of fast processing speed and small melting pool area.In order to solve the real-time processing problem under high-speed,the CameraLink acquisition card with FPGA acceleration chip is selected,and the Runtime6.0 of the acquisition card is used as the platform and Visual Applets is used as the programming language,combined with image processing technology such as filtering,morphological processing,segmentation,feature extraction,etc.,to design online monitoring algorithm for molten pool morphology of SLM system,and finally the image of molten pool is classified as normal trailer and spatter.