Vision and navigation for the Carnegie-Mellon navlab
IEEE Transactions on Pattern Analysis and Machine Intelligence - Special Issue on Industrial Machine Vision and Computer Vision Technology:8MPart
Ambient light array vision in three-dimensional
Ambient light array vision in three-dimensional
Point Light Source Estimation from Two Images and Its Limits
International Journal of Computer Vision
| Hi-index | 0.11 |
Theoretical foundations of an advanced vision system capable of detecting the three-dimensional orientation of planar surfaces from a single view of a scene are presented. The standard 2-D picture records the light intensity of points registered at the point of observation, whereas data for the new vision system is represented by the direction and intensity of light rays originating from points on objects in a scene, and registered at the point of observation. The collection of these rays, i.e., light intensities and associated directions, is called an Ambient Light Array and is modelled by a vector function. In the new system the image of a flat surface is a sphere. An algorithm for the detection of orientation of a planar surface exhibiting the Lambertian reflectance characteristic and illuminated by a point light source is described, and the results of simulation in the presence of random noise are shown.