Range image segmentation
Range images are used more and more as a means for extracting spatial features from the surrounding world. In a storage structured environment module a teleoperated vehicle circulates, carrying among other equipment a Laser Range Finder (LRF), which is used to determine its exact location within the module, as well as to generate range images as "seen" from this vehicle. From the analysis of these images a spatial model of the surrounding environment is built in order to monitor, aid and validate the vehicle's operation. In this paper, the preliminary results of range images segmentation are described. The edge detection is achieved with a residual analysis algorithm proceeding in two distinct phases for jump edge and roof edge detection. Some results of segmenting noisy images are also presented.
Bibliographic Reference: Paper presented: REC PAD 90, Lisboa (PT), March 29-30, 1990
Availability: Available from (1) as Paper PT 35364 ORA
Record Number: 199010848 / Last updated on: 1994-12-01
Original language: pt
Available languages: pt