SEARCH
MENU

History of Robotics Research and Development of Japan1982SensingObject Recognition Using Three-Dimensional Information

Masaki OshimaElectrotechnical Laboratory (ETL)
Yoshiaki ShiraiElectrotechnical Laboratory (ETL)
The vision group of the Electrotechnical Laboratory took the initiative in the world to develp a three-dimesional measurement system for vision (a range finder) which consists of a stripe light projector and a television camera.  Research has performed to extract so-called a surface element of each point of objects from a set of local three-dimensional points. These elements are used to find regions of objects. The research described in the followings is based on these results: This is an approach to the recognition of stacked objects with planar and curved surfaces. The system works in two phases. In the learning phase, a scene containing a single object is shown one at a time. The range data of a scene are obtained by a range finder. The description of each scene is built in terms of properties of regions and relations between them. This description is stored as an object model. In the recognition phase, an unknown scene is described in the same way as in the learning phase. Then the description is matched to the object models so that stackd objects are recognized sequentially. Efficient matching is achieved by a combination of data-driven and model-driven search process. Experimental results for blocks and machine parts show that this approach is successful. In this research shows:
  1. It was shown by using three-dimensional data that conventionally difficult solids can be recognized with a vision system. The three-dimensional data are obtained by a range finder consists of a stripe light projector and a television camera.
  2. Flexible and efficient recognition is achieved with the consistent techniques, such as a preprocessing, description by regions, recognition by matching, and learning.
IEICE Best Paper Award in 1983.
Fig. 1 Range data.<br /> (for a scene consists of planar and curved surfaces)
Fig. 1 Range data.
(for a scene consists of planar and curved surfaces)
Fig. 2 Surface elements.
Fig. 2 Surface elements.
Fig. 3 Regions
Fig. 3 Regions
Fig. 4 Recognition result.
Fig. 4 Recognition result.
Fig. 5 Range data.<br /> (for a scene consists of machine parts)
Fig. 5 Range data.
(for a scene consists of machine parts)
Fig. 6 Surface elements.
Fig. 6 Surface elements.
Fig. 7 Regions
Fig. 7 Regions
Fig. 8 Recognition result.
Fig. 8 Recognition result.

Related Article

記事はありません。