|
RKF has developed software for various other applications besides the products and solutions listed elsewhere in this section. These software tools primarily fall in the areas of computer vision. Specifically, RKF has developed several computer vision technologies that can be used for:
- Target acquisition, identification and tracking
- Human identification
- Determining commonality between multiple images
- Identification using data fusion from many sensor types
- Optimization and test of alternative approaches
|
|
|
These technologies are primarily based on the "Shape Transform", an image transform developed by Dr. Jeffrey Freedman in his PhD dissertation at the University of Maryland. For that dissertation partially occluded 3d objects were detected in any 3d orientation in noisy range imagery in near real time. The algorithm was used to detect human faces, articulated limbs and other solid objects in a cluttered scene. In addition this approach was used to track an articulated hand as it moved through an image.
|