Visual Observables and Invariance

This paper presents the visual measurement of physical object properties that characterize the perceived object including: size, shape, surface properties, cover reflectance properties, distance, and motion. We provide an overview of complete set of local spatial, spectral and temporal measurements. From local visual measurements and a physical model of the visual stimulus formation, we derive complete sets of photometric, geometrical, and temporal invariants to counteract unwanted transformations in the observation including: illumination spectrum and intensity, scene setting causing shadow, shading and highlight effects, and variation due to object position, pose and distance. Experiments show the different invariants to be highly discriminative, while maintaining invariance properties. The presented framework for invariant measurement is well-founded in physics as well as in measurement science. Hence, the local visual measurements and their invariant representations are considered theoretically better founded than existing methods for the measurement of invariant features.