Vision based orientation invariant measurement of metric screw thread parameters

Measurement of screw thread parameters is one of the most tedious procedure in conventional metrology practices since most parameters part of the screw thread nomenclature require exclusive equipment specific to each parameter. Conventional methods are contact-type and also time consuming. Vision aided metrology has been appreciated as one of the most powerful technique for non-contact measurement. The paper reports a research work involving development of a machine vision system based screw thread parameters measurement. Metric screws are considered for the measurement though the algorithm presented may be suitably adapted for most commonly used screw thread patterns. Image of a profile projection of the screw is acquired with a diffused transmitted lighting with LED as the light source. A contour processing algorithm is developed to estimate the screw thread parameters such as various diameters, angles included in the threads, linear distance metrics along the screw axis. The algorithm presented in this paper is insensitive to part presentation in terms of its position and orientation. The paper describes the various steps involved in the measurement system starting from selection of the imaging system, hardware configurations, image acquisition.