Robust integration of thermal and visual imagery for outdoor scene analysis

The author presents a statistically robust approach for multisensory computer vision. Specifically, energy exchange model parameters used to interrelate thermal and visual imagery are reliably computed. The approach extracts physically meaningful estimates of internal object properties which are useful for automated object recognition. The robust technique minimizes sensitivity to outliers caused by segmentation errors and misregistration which are endemic to multisensor fusion.<<ETX>>