Research On Surface Crack Recognition Of Carbonized Moxibustion Strip Based On Machine Vision

Carbonization moxibustion strips are developed from the ancient moxibustion of traditional Chinese medicine.At present,the production technology level of carbonization moxibustion strips is uneven and the degree of automation is relatively low,which restricts the development of carbonization moxibustion strip industry.In the process of carbonization moxibustion strip making,cutting,drying and transportation,it is easy to be affected by the environment and other external forces,resulting in cracking,extrusion and damage,so as to form cracks or defects.Cracks and defects not only affect the appearance of products,but also may cause fracture of carbonized moxibustion strips and unsustainable combustion.Based on machine vision,this paper detects and identifies the surface cracks of carbonized moxibustion strips,which has certain economic value,and can also reduce the labor intensity and the risk of occupational diseases.(1)The appearance characteristics of carbonized moxibustion strip were studied,and its crack characteristics were analyzed.Aiming at the problem of poor imaging effect,a special image acquisition equipment is built,and the best working illumination is found through shooting and testing under different illuminance.(2)In view of the image interference caused by debris or left marks in the detection of carbonized moxibustion strips,denoising was carried out.Several mainstream filtering algorithms are compared and analyzed,and finally the median filtering method is selected to eliminate its noise.The characteristics of carbonization moxibustion strip images in R,G and B channels are analyzed.Finally,the gray image is obtained by weighted average to remove the interference of redundant channels.(3)Combined with the influence of carbonization moxibustion strip surface roughness and irregular fine lines on recognition,the common edge detection algorithms are analyzed and compared,and Canny operator is used for edge detection.The contour of the detected edge is searched and drawn,and the crack image is obtained after double threshold optimization.(4)In order to realize the complete detection of the outer surface of the carbonization moxibustion strip,a detection device is designed in this paper.By using the cooperation and control between various mechanisms,the carbonization moxibustion strip can complete rolling,take photos and identify automatically,and finally remove the cracked carbonization moxibustion strip.Through identification test,the average identification accuracy of the detection device is 93.575%,and the average deviation of crack identification rate is 6.0%.In conclusion,the carbonization moxibustion strip recognition device based on machine vision algorithm can effectively identify the surface cracks of carbonization moxibustion strips.