Automatic evaluation of form errors

Automating completely the process of geometric errors verification has some problems. The most critical ones are the preventive shape recognition and the lack of an agreement about criteria to be used to associate ideal features to the extracted points from the real surface. In this paper an approach for automatic metrological inspection of manufactured parts is proposed. Starting from a tessellated model, coming from a scanned workpiece, the methodology carries out automatically the partition of the acquired surface by a shape recognition approach. Then, the ideal features are associated to the corresponding non-ideal features, based on some criteria that permit to define the minimum zone solution. The proposed methodology stands out for the number and the types of the automatically recognized features. Finally, the form errors, related to all the recognized features, are evaluated. In this paper the procedure has been used to evaluate some specific form error typologies for several simulated case-studies and real objects. The results deriving from this experimentation are here critically discussed.