PROFILE RECOGNITION AND MENSURATION IN MACHINE VISION

This paper presents a systematic profile recognition and mensuration a pproach in machine vision. It can be utilized to recognize and measure the profiles of industrial parts in an automated manufacturing proces s by machine vision systems. A new method od profile representation by sampling the data from the abject boundary in a digital image is pres ented. Autoregressive (AR) models are used to code the sampled data of the profiles into AR coefficients for profile recognition. Characteri zation of the profiles is accomplished by the Data Dependent Systems ( DDS) methodology. The AR coefficients and characteristic roots help co nstruct the AR and DDS descriptors to characterize the signatures of t he profiles. The frequency domain information about the profiles can b e extracted by DDS analysis. The measurement of the profile variation is obtained from the DDS results using optical mensuration method. Neu ral network and feature weighting method are utilized as reasoning mac hines for recognition. The illustrative examples in which the profile sampled data are corrupted by noise show that the profile recognition and mensuration approach is very effective and robust in a typical noi sy environment on the shop floor.