Online Detection Method For Appearance Defects Of Ice Scoop

Ice spoon is an indispensable tool for eating cold drinks such as ice cream,its main raw material is birch.Affected by milling cutter wear and defects in the birch material itself,the finished ice spoons have different defects after processing.In market research,the ice cream bar online detection system needs to be improved in hardware design.After consulting relevant literature,the existing ice cream stick defect detection algorithms cannot effectively detect some defects of ice spoons,such as slab,buns,split and burr defects,the existing algorithms cannot distinguish between mineral blocks and pollution defects,the existing algorithms cannot distinguish light chromatic aberration and heavy chromatic aberration defects well.In order to facilitate the analysis,the above defects are divided into two categories: surface defects and side defects.Surface defects include slabs,burrs,mineral blocks,contamination,and color differences.Taking slab texture features as the detection object,this paper proposes a detection method based on Gaussian line detection and vertical projection,which effectively implements slab defect detection.The burr is shown as expenditure burr and split burr,this paper uses fixed threshold segmentation and morphological processing to implement burr defect detection,the algorithm has a fast detection speed.This article analyzes mineral blocks and pollution defects in color space,and uses the characteristics of different saturation of these two types of defects to distinguish the two types of defects.For chromatic aberration defects,this paper uses the maximum inter-class variance method to extract the chromatic aberration areas,and calculates the average gray value of the defective areas to distinguish the severity of chromatic aberration.Lateral defects include two types of defects: buns and split.The side light source is illuminated at a high angle with the horizontal plane,so that the beam is unevenly irradiated on the head of the ice spoon,at this time,the image of the buns defect can be divided into two categories: when the buns defect is facing away from the light source,the buns defect appears on the image,it is a dark area.When the buns is facing the light source,a bright line with a certain width appears in the image where the upper edge of the buns head is connected to the surface.Based on this imaging characteristic,this paper designs an edge extraction model to achieve two type of buns defect detection.The split appears as an obvious dark crack on the side of the head or the dislocation of the area caused by the split,this paper uses the method of Gaussian line detection and edge detection to detect the split defect.In this paper,the prototype of the scoop appearance defect is improved,and the algorithm verification is achieved by collecting images and establishing a gallery.Finally,the field test results show that the proposed scoop appearance defect detection algorithm proposed in this paper has reached the standard of scoop quality detection,and the detection speed is 6 per second,the root ice spoon has a detection recognition rate of 99.45%,which has certain practical application value,and has laid a good foundation for the development of the appearance defect detection of the ice spoon.