Smart Table

Object Recognition Algorithm

The object recognition algorithm computes position, orientation and identification of objects on the table surface. 'Object pad' denotes one of the magnets on the back of an object. First, the algorithm searches for pads based on sensor status information, which results in the position of these pads. Second, pads are grouped into clusters of five pads, where each cluster represents an object. The third step is identification of objects, i.e. object identification numbers are evaluated based on pad clusters. Fourth and final step is computation of object localization, which consists of positions and orientations of objects.

According to the four steps towards object recognition, the algorithm is split into four independent parts, which are 'pad localization', 'object pad grouping', 'object identification', and 'object localization'. The interfaces with associated data in- and outputs, and the corresponding algorithm parts are shown in the picture above.

This material is based upon work supported by the National Science Foundation under Grant No. ANI-0085773. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).