Design Of Gear Size Measurement System Based On Image Mosaic

Gear has a very high position in the mechanical industry,and it is indispensable to measure the size of gear.The traditional contact measurement has higher labor cost and lower efficiency.In recent years,machine vision measurement technology has developed rapidly,and it has been applied to gear size measurement.But at present,machine vision measurement is mostly used to measure gears whose size is less than the effective field of vision of the camera.For gears beyond the effective field of view of the camera,it is only solved by increasing the object distance or using wide-angle lens,but increasing the object distance will cause the loss of gear edge details,reduce the measurement accuracy,and wide-angle lens will lead to distortion,distortion will also lead to the reduction of measurement accuracy.Based on the above problems,this paper develops the research of gear dimension measurement based on image mosaic.This paper studies spur gear.The main research contents include image mosaic,image segmentation,edge detection and size measurement.Image mosaic: The SURF algorithm with rotation invariance and scale invariance is selected as the feature point detection method,and the FREAK algorithm with binary bit descriptor is used to describe the feature points,which speeds up the operation of the algorithm.Violent matching is used to match feature points,the ratio of nearest matching to next-nearest matching,effective feature point density clustering and RANSC algorithm are used to screen feature points,and zero-error feature points are obtained.The feature matrix is generated and modified to ensure that perspective transformation is rigid transformation,and image fusion is carried out by local weighted fusion method.Image segmentation: By comparing the commonly used segmentation methods,the combination of threshold segmentation and edge segmentation is selected to separate the gear target from the background.Edge detection: The common edge detection operators are analysed,and using OTSU algorithm to optimize Canny algorithm to get adaptive threshold edge detection algorithm.Find out the gear edge through edge tracking,eliminate the useless edge,use parallel stretching thinning algorithm to refine the gear edge,get single-pixel edge.Dimension measurement: Subpixel corner detection is carried out on the refined edge,sub-pixel corner points of the apex circle and the root circle are extracted,and the apex circle and the root circle are fitted by least square method.The apex circle diameter and the root circle diameter of the sub-pixel level are obtained.According to the apex circle diameter,the root circle diameter and the number of teeth,the dividing circle diameter and the tooth spacing are calculated,and the measurement knots are also calculated.Error analysis was carried out.The experimental analysis shows that the gear size measurement system based on image mosaic designed in this paper can accurately and efficiently measure the gears beyond the effective field of vision of industrial cameras.