DIMENSIONAL AND ANGULAR MEASUREMENTS USING LEAST-SQUARES AND NEURAL NETWORKS

In this study a machine vision approach is developed for dimensional a nd angular measurements of manufactured components comprising straight line segments. We aim at the measurements of distance between two par allel lines and angle between two intersecting lines using both least mean square (LMS) and artificial neural network (ANN) techniques. LMS models estimate the line parameters based on the sum of squared perpen dicular distances, rather than the vertical distances, between the obs erved data points and the line. A set of 23 gauge blocks of varying si zes is used to evaluate the performance of the LMS line estimators. Ex perimental results show that the measurement errors of the LMS models are affected by the line length and orientation of digital images. ANN techniques are, therefore, used to adjust the measurement errors resu lting from the LMS models. Two back-propagation neural networks are de veloped, one for measuring the distance between two parallel lines, an d the other for measuring the angle between two intersecting lines. Ex perimental results show that the ANNs are very effective for correctin g the measurement errors regardless of line lengths and orientations o f digital images. A 90% improvement in measurement accuracy for the AN N compared to the LMS was achieved By using the ANNs, the measurement accuracy and flexibility in manufacturing applications can be signific antly improved.