In modem manufacturing technology with increasing automation of manufacturing processes
and operations, the need for automated measurement has become much more apparent.
Computer measuring machines are one of the essential instruments for quality control and
measurement of complex products, performing measurements that were previously laborious
and time consuming. Inspection of sculptured surfaces can be time consuming since, for exact
specification, an almost infinite number of points would be required. Automated measurement
with a significant reduction of inspected points can be attempted if prior knowledge of the part
shape is available. The use of a vision system can help to identify product shape and features but,
unfortunately, the accuracy required is often insufficient. In this work a vision system used with
a Coordinate Measuring Machine (CMM), incorporating probing, has enabled fast and accurate
measurements to be obtained. The part features have been enhanced by surface marking and a
simple 2-D vision system has been utilised to identify part features. In order to accurately identify
all parts of the product using the 2-D vision system, a multiple image superposition method
has been developed which enables 100 per cent identification of surface features. A method has
been developed to generate approximate 3-D surface position from prior knowledge of the product
A probing strategy has been developed which selects correct probe angle for optimum accuracy
and access, together with methods and software for automated CMM code generation. This has
enabled accurate measurement of product features with considerable reductions in inspection
Several strategies for the determination and assessment of feature position errors have been investigated
and a method using a 3-D least squares assessment has been found to be satisfactory.
A graphical representation of the product model and errors has been developed using a 3-D solid
modelling CAD system. The work has used golf balls and tooling as the product example.
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.