논문명 | Upper-Nonconvergent Objects Segmentation Method on Single Depth Image for Rescue Robots |
---|---|
논문종류 | 기타저널 |
저자 | Hoang Vu, Seoungjae Cho, Seongjo Lee, Kyungeun Cho |
Impact Factor | |
게재학술지명 | International Journal of Latest Research in Engineering and Technology |
게재일 | 2003.01 |
In this study, we proposed an algorithm to segment a special type of objects which we called the Upper-Nonconvergent Objects (UNOs) from the depth image. This type of objects plays an important role in our system. In that system, we use a robust algorithm based on height map to segment objects in three dimensional data. However, this algorithm cannot be segment the Upper-Nonconvergent Objects. Our algorithm works on two dimensional depth image which produced from 3D data. We first segment all the objects from the depth image including the ground using splitting and merging approach, then use morphological erosion method to extract the boundary of every objects, finally, we extract the Upper-Nonconvergent Objects using mathematical requirements which we defined. Information extracted from this algorithm can be used as additional process to the whole algorithms works on three dimensional data. |