VOL. 4, NO. 5, JULY 2009 ISSN 1819-6608 ARPN Journal of Engineering and Applied Sciences © 2006-2009 Asian Research Publishing Network (ARPN). All rights reserved. www.arpnjournals.com 30 DEVELOPMENT AND APPLICATION OF A MACHINE VISION SYSTEM FOR MEASUREMENT OF SURFACE ROUGHNESS D. A. Fadare and A. O. Oni Mechanical Engineering Department, University of Ibadan, Ibadan, Oyo State, Nigeria E-Mail: fadareda@yahoo.com ABSTRACT Monitoring of surface roughness is an essential component in planning of machining processes as it affects the surface quality and dimensional accuracy of machined components. In this study, the development and application of a machine vision system suitable for on-line measurement of surface roughness of machined components using artificial neural network (ANN) is described. The system, which was based on digital image processing of the machined surface, consisted of a CCD camera, PC, Microsoft Windows Video Maker, frame grabber, Video to USB cable, digital image processing software (Photoshop, and MATLAB digital image processing toolbox), and two light sources. The images of the machined surface were captured; analyzed and optical roughness features were estimated using the 2-D fast Fourier transform (FFT) algorithm. A multilayer perceptron (MLP) neural network was used to model and predict the optical roughness values. Tool wear index and five features extracted from the surface images were used as input dataset in training and testing the ANN model. The results showed that the ANN predicted optical roughness values were found to be in close agreement with the calculated values (R 2 -value = 0.9529). Thus, indicating that the proposed machine vision system and ANN model are adequate for online monitoring and control of surface roughness in machining environment. Keywords: measurement, surface roughness, machining, image processing, machine vision system, artificial neural network. 1. INTRODUCTION The demand for improved flexibility, productivity, and product quality in modern machining industry has necessitated the need for high-speed, non- contact and on-line monitoring and measurement of surface roughness of machine components. The quality of components produced is of main concern in planning of machining processes as it affects the surface quality and dimensional accuracy of the products [1]. Therefore, critical examination of surface roughness of the components is required as a quality control measure. The conventional method for assessing surface roughness is normally by using stylus type devices, which correlate the vertical displacement of a diamond-tipped stylus to the roughness of the surface under investigation. This method is widely accepted and has been used for many decades in the manufacturing industry [2]. However, this method requires direct physical contact with the surface of the workpiece, which necessitated halting of the machining operations in order to measure the roughness. Hence, this method is time consuming and cumbersome and therefore, not suitable for high-speed and high volume production systems. Another disadvantage of this method is the resolution and the accuracy of the instrument, which depends mainly on the diameter of the tip of the probe of the stylus device. As a solution to these limitations, other non- contact methods such as atomic force microscopy, phase shifting interferometry, stereo scanning electron microscopy, and laser scanning microscopy has been developed with reasonable success and commercial application of these methods is becoming increasingly popular in manufacturing industry [3-5]. However, all these developed non-contact methods are off-line-based. Hence, they can not be used for on-line and real-time monitoring and control surface roughness in machining environment. The application of machine vision system offers better solution in on-line and real-time monitoring surface roughness. Machine vision involves the use of camera, frame grabber, computer system and image processing software to acquire, analyses, monitor, and assess surface roughness parameters. Machine vision systems play an important role in the monitoring and control of automated machining systems. It has generated a great deal of interest in the manufacturing industry [6]. Researchers have shown that the application machine vision has the advantage of being non-contact and has well faster than the contact methods [7]. Using machine vision, it is possible to characterize, evaluate, and analyze the area of the surfaces of machined components. Several investigations have been carried out using the non-contact optical methods for the assessment surface roughness. These methods are based on statistical analyses of the gray-scale images in the spatial domain [6]. The intensity histograms of the surface image have been utilized to characterize surface roughness and quality [8]. They utilized statistical parameters, derived from the grey level intensity histogram such as the range and the mean value of the distribution and correlated them with the centre line average (R a ) value measured with a stylus instrument. Statistical methods such as co-occurrence matrix approach, the amplitude varying rate statistical approach and run length matrix approach have also been used to monitor the texture features of machined surfaces [9]. A 2-D fast Fourier transform (FFT) of the digitized surface image in which the magnitude and frequency information obtained from the FFT are used as measurement parameters of the surface finish has developed by Hoy and Yu [10]. Hisayoshi et al. [11] has