Wear Monitoring of Cemented Carbide Tools Based on Raster Projection Image
Tool wear monitoring is one of the core technologies of automatic production.In order to ensure the re-liability of monitoring results,this article presented a method of monitoring cemented carbide tool wear based on raster projection images.Firstly,we used the grating projection detection system to collect the pattern of the gate line of the measured object and calculated the length of the grating on the projection wall.Secondly,we used the gray coding al-gorithm to divide the grating strip information,and thus to obtain the tool profile data through three-dimensional infor-mation reconstruction.After scanning all image data,we regarded the sparse data as isolated points,and clustered the same feature images by the balanced iterative algorithm based on hierarchical structure.Secondly,we used three indi-cators,namely the intensity mean,attribute and clarity,to optimize the image quality.On the basis of the visual salient feature,we used the gradient of the gray density function to find out the abrupt change point of the grating projection image.Moreover,we divided the resolution layer of the tool wear image by wavelet decomposition,and solved the problem of image noise and uneven illumination by the diffusion operation.Finally,the tool wear orientation was locked by the Gaussian function.Simulation results show that the proposed method has high tool wear monitoring ac-curacy and reliability,as well as wide application space.
NETL
NSTL
万方数据
