Research On Surface Defect Detection Of Refractory Bricks Based On Grating Projection

Refractory brick is a high temperature building material that can be used to build various thermal equipment and furnaces.In the production process,improper craftsmanship and operation will cause surface defects such as missing corners,missing edges,pits and protrusions.Such defects not only affect the service life and safety of the material,but also have a great impact on the production quality and economic benefit of refractory bricks.At present,the surface quality inspection of refractory bricks still mainly relies on traditional manual inspection methods.There are problems such as low product sampling rate,low inspection accuracy,high cost and low degree of automation.It is greatly affected by artificial experience and subjective factors and cannot meet the needs of modern production.Since the refractory brick surface contains some brown or black spots,its color background texture affects the detection accuracy in traditional image processing.Grating projection technology is not only unaffected by it,but also has the characteristics of fast detection speed,high detection sensitivity and non-contact.In order to realize the automation of refractory brick surface defect detection,this paper deeply studies the refractory brick surface defect detection system based on grating projection.The main contents are as follows:(1)Design of surface defect detection system for refractory bricks.Aiming at the defects with obvious depth information on the surface of refractory bricks,such as missing corners,missing edges,pits and protrusions,the overall design of the surface defect detection scheme is carried out from the level of software and hardware on the basis of clarifying the detection requirements for surface defects of refractory bricks.A set of surface defect detection system of refractory bricks based on grating projection is built.(2)Refractory brick surface image processing and defect identification.According to the detection requirements,the eight-step phase shift method is used to solve the phase of the acquired sinusoidal grating fringes.The multi-frequency heterodyne method is applied to the phase unwrapping to obtain a high-precision absolute phase map.Combined with image processing algorithms such as mean filtering,Canny edge detection and morphological processing to realize the identification and marking of surface defects.(3)Three-dimensional reconstruction and defect identification of refractory brick surface.Binocular stereo vision technology was introduced to solve the problem that grating projection combined with image processing could not identify the defect depth information.Zhang’s calibration method was used to calibrate the monocular camera and binocular camera respectively.Bouguet algorithm is used for stereo correction.The three-dimensional point cloud of the refractory brick surface is obtained by the phasebased stereo matching algorithm.The RANSAC algorithm is applied to the filtered point cloud for plane fitting.An improved point cloud dimensionality reduction algorithm is proposed.Morphological processing is applied to the image after dimension reduction to realize the extraction and identification of 3D defect features.(4)Experimental verification and analysis.An experimental platform was built to verify the feasibility of the combination of grating projection and image processing and the combination of grating projection and binocular stereo vision in the detection of refractory brick surface defects.The results show that the combined method of grating projection and image processing has a faster detection speed,but the depth information of defects cannot be obtained.Although the detection speed of grating projection and binocular stereo vision is relatively slow,the detection accuracy is higher,and threedimensional defect information including defect depth can be obtained.Therefore,both methods can identify defects with high accuracy and make a reasonable choice according to the requirements of application occasions,which promotes the development of automatic detection of surface defects of refractory bricks.