Recognition of shiny dielectric objects by analysing the polarization of reflected light
Image and Vision Computing
Constraining Object Features Using a Polarization Reflectance Model
IEEE Transactions on Pattern Analysis and Machine Intelligence
Polarization methods in computer vision
Polarization methods in computer vision
| Hi-index | 0.00 |
Because of the polarization camera -- a novel imaging device being developed with support from the ARPA Image Understanding Program -- polarization vision is rapidly emerging as a new general approach to image understanding and computer vision. Though human vision is oblivious to components of light polarization states, which provide an important visual extension to intensity and color, humans can directly visualize polarization as a particular hue and saturation. This article shows how such an approach can serve to present images of ordinary scenes as never before seen by humans in the domain of polarization vision.Metaphorically, humans are color blind with respect to the perception of polarization. Though this does not appear to inhibit human visual performance, under certain conditions sensory augmentation through polarization vision can significantly enhance both automated image understanding and possibly even improve human visual performance.Polarization cameras automatically sense components of polarization, as well as computationally process these components to produce polarization images. Recent advancements in polarization camera design enable low-cost modular components to interface with almost any existing imaging device, converting it into an automatic polarization camera. This compatibility with small, portable imaging devices for the first time makes polarization imaging accessible to a number of outdoor and underwater applications. As we will see, this vast new visually augmented domain includes applications in natural object recognition, automatic target detection and recognition, inspection of ship hulls for damage, and marine biology.