The Self-Aiming Camera
The self-aiming camera (SAC) is an automatic surveillance camera developed by a team of neuroscientists and computer scientists at the University of Illinois. SAC aims itself at the same types of items that a human security agent would aim at, using a computational method called intelligent multisensor fusion (IMF) developed by me. SAC and IMF have been patented.
IMF automatically locates items of interest by imitating the brain and the eye. In humans, the light-sensitive part of the eye is called the retina, and the retina contains a high-resolution patch called the fovea. When humans survey a scene, they use faculties including their multiple senses and their knowledge to aim their fovea at items in the scene that they need to see at high-resolution. IMF works the same way. It combines low-resolution images of a scene from multiple sensors to decide which items in the scene are important and which are simply part of the background. This smart operation is made even smarter by adjusting the low-resolution sensor inputs using specialized knowledge modules. Once an item of interest is selected and localized, IMF directs a movable, high-resolution camera at it and takes a snap-shot. IMF has already been used in real-world security applications.