Stochastic Systems Group
Home Research Group Members Programs  
Demos Calendar Publications Mission Statement Alumni

SSG Seminar Abstract


Registration and Fusion of Optical and Ladar Imagery

Andrew Mastin
CSAIL, MIT


Recent advances in airborne laser radar (ladar) sensor technology have made the acquisition of large scale digital elevation maps efficient and cost effective. The resulting availability of urban ladar data has motivated the construction of 3D models via fusion of traditional optical imagery and ladar imagery. The chief challenge in generating such models is registering the two images, as the image modalities are inherently different. In this talk, we introduce a novel approach for performing 2D-3D registration that is based on information theoretic similarity measures. We present preliminary registration results and demonstrations of 3D model generation.



Problems with this site should be emailed to jonesb@mit.edu