Application of Neural Network for Estimation of Surface Roughness Using Discrete Wavelet Transform

Machine vision for surface roughness, is a measurement of features in the branch of metrology. In many research schemes lead machine vision applications, as the advantages lie in non-contact and speed process rather than contact methods. Machine Vision, is a used to provide image inspection and analysis like process control, automatic inspection and robot guidance in industry, to make intelligent decision. In this research work, roughness estimation is done by Machine vision. The enhanced image's feature extraction done with Transform techniques in frequency domain. Neural network (NN) is trained with the inputs acquired from wavelet Transform to obtain R-t as output. The surface roughness (R-t) is estimated based on NN is compared with R-t values of the Stylus technique.