Research On Laser Intelligent Removal System Based On Machine Vision

Recent years,laser technology is developing rapidly.It has been applied in many fields.In terms of bridge maintenance,laser rust removal technology is replacing traditional one.Regretfully,the laser rust removal equipment all needs manual operation on site,which is not only ineffective but also dangerous to some extent.For example,the high strength luminance of laser can cause damage to workers' eyesight.The rusty granule produced while they are removed is harmful to workers' health to some degree.To solve the problems,this paper studies the laser rust removal intelligent control system.It takes advantage of machine vision inspection to identify surface rust.Then it cleans the identified rust by using laser removal system.It will lay a solid foundation to future intelligent unmanned laser removal industry.Below are the main studies.(1)Research on laser derusting process parameter library.Put forward rust grade adapting to laser rust removal,include floating rust,light rust,medium rust and heavy rust Use the laser removal to clean all grades rust.There are 2 technology demands for laser rust removal:1)There is no remain rust on the artifact surface after laser rust removal The quality is satisfied to Sa2.5 level;2)The roughness of substrate surface should be between 35 ?m to75 ?m to make sure adhesion when spraying again.In the progress of laser rust removal,this paper aims to explore the effect of basal surface Fe element weight percent and surface roughness by laser power,frequency,number of scans and the speed of Galvanometer scan through the single-factor experiment and orthogonal experiment.(2)Research on machine vision recognizing rust.There are 3 technology demands for machine vision:1)Identify the artifact surface rust and classify it by practical rust grade on image;2)Each grade rust images need connect as much as possible.Expand the area where there are many little holes to large and intact rust area;3)Rust image must include all artifact surface area.The missed detection must be less than 4%.Deal with the rusty object surface image by using Python and OpenCV open Source computer vision library.Recognize the rusty area by a series of image processing algorithm and obtain the rusty area position,rust grade,size and so on.(3)Integrate and test of rust removal intelligent control system.Integrate machine vision with laser derusting as a device.Machine vision provides information of artifact surface rust grade and profile.Different rust grade images is imported to different coverage by laser rust removal equipment.Use diverse parameters to confirm rust position and project derusting route to achieve fixed derusting.After that,machine vision,as a detection tool,will identify the artifact surface again to make sure the derusting quality.The research indicates that:machine vision can recognize entire rust area,and it can transmit information including rust position,rust grade and size to laser cleaning equipment.Then,the laser rust removal equipment uses the parameter according to rust grade to confirm rust position and project derusting route to achieve the goal that clean the artifact surface rust by fixed derusting.The substrate surface quality after laser intelligent rust removal is approximately equal to traditional laser rust removal.And it meets the demand of coating twice.It is worth mentioning that its efficiency far surpasses the traditional one.