Adjustment of vision system parameters and image processing algorithms for measurement of miniature parts in automatic assembly

Automatic assembly in the small world presents many challenges. In order to fulfill precise measurement with limitation of vision field, diversity and severe surface condition of miniature parts, a two-layered architecture with operator interaction is proposed. These two layers specify for user and developer of the assembly application respectively. Based on part's feature, the former adjusts vision parameters with the interface of dynamic illumination control and the later develops reasonable feature extraction strategy for measurement of part's position and orientation with an image processing tool. With general image processing algorithms, it can not only be applied by itself, but also be integrated into an assembly system to facilitate developer to adjust image processing strategy. This operator interaction function can improve the accuracy of recognizing and locating part, promote the success ratio of product assembly and shorter the period of software development as well.