Research On Shape Detection Based On Digital Projection

The iron and steel industry is currently the basic industry of China's national economy.However,because products are mostly low-end products,product quality is not high,and competitive high-end products are lacking,they are in a state of "big but not strong".As an important category of steel products,cold-rolled strip steel has a very wide range of applications in the automotive industry,petrochemicals,mechanical products,and electrical kitchen utensils industries.With the rapid development of modern manufacturing industry in China,the quality requirements of strip steel are also increasing.The flatness is an important indicator to measure the quality of strip steel.Therefore,the flatness detection system has been proposed high precision requirements.Conducting inspection and study of strip flatness is of far-reaching significance in promoting the development of advanced manufacturing in China.In this paper,on the basis of summarizing the research of contact and non-contact flatness detection technology at home and abroad,taking cold rolled strip as the research object,a non-contact flatness detection method based on DLP digital projection is studied.Focused on the basic principles of the projection stripe detection method,the distortion correction of the image,the detection and extraction of the projection strip center,and the fitting of the projection curve,the theoretical analysis and research were carried out.The main contents of the thesis include the following points:(1)The definition of strip flatness and two commonly used representation methods are introduced,and common types of slab defects are introduced.Describes the principle of plate shape detection based on DLP digital projection technology and the method and procedure for shape detection in this paper.(2)The imaging model of the camera and the digital model of lens distortion in the digital imaging process are introduced.In the actual production process,due to the limitation of the lens processing technology and the assembly process,the lens will cause distortion of the captured image,and elaborates the image distortion correction algorithm.(3)The paper discusses the image processing algorithm in flatness detection.In the image preprocessing,the source of noise in the shape detection process is analyzed.The principles of spatial and frequency domain filtering algorithms are discussed,and the image is filtered and denoised.In the light bar center extraction algorithm,the light bar center extraction algorithm based on the gray center of gravity and the light bar center tracking algorithm are respectively studied to extract the center of the fringe in the projected image.In addition,a curve fitting algorithm based on Akima interpolation is used to fit the light curve to calculate the length of the light bar.(4)The situation of a variety of strip-shaped defects in strip steel was simulated by experiments,and the flatness was detected using the detection algorithm studied in this paper,and the detection results were analyzed.