In automated object recognition systems, measurements from a single source of information do not always suffice for the reconstruction of the underlying scene. Incompleteness, inaccuracy and unreliability of the information often leaves room for multiple interpretations of the world which are consistent with the available data. Combination of scene descriptions from multiple sensors can in these cases help to resolve the ambiguity. This thesis discusses the issues involved in the design of a multi-sensor system for object recognition and illustrates these issues in the light of a system for recycling of electronic components mounted on printed circuit boards. Integrated circuits, electrolytic capacitors, batteries and oscillators are the main electronic components of interest that are encountered on printed circuit boards, because these objects are either valuable for re-use or harmful for the environment when they are incinerated. The shape and surface material of these electronic components contain sufficient clues for the discrimination of the different object classes. In order to capture these properties, a pilot system was built which incorporates a range image sensor, a colour camera a high-resolution grey-level camera and an eddy current sensor.
|Award date||23 Apr 1999|
|Place of Publication||Enschede|
|Publication status||Published - 23 Apr 1999|