Advanced Imaging

AdvancedImagingPro.com

   

Advanced Imaging Magazine

Updated: January 12th, 2011 09:49 AM CDT

Solving a SIPHER

A spatial intensity phase evaluator (SIPHER) for perceptual object detection in images
CACI Geospatial Technologies
Figure 1. A nominal out-of-phase condition for two signals. Because they are out of phase with one another, but otherwise identical, their relative amplitudes vary with time synchronously. The amplitude of signal 1 may be equal to or different from that of signal 2 at any given
Figure 2a.
Figure 2b.
Figure 2c. Flat or plane-faceted surfaces have more of a flashing effect. These figures include synthetic images of a sphere, a cylinder and two flat patches illustrate this effect as the threshold is increased across a range of values.
Figure 3a.
Figure 3b.
Figure 3c.
Figure 4a. This is a scene with a very small board partially hidden by some tree trunks, twigs and leaves (the brighter spot between the lower trunk areas of the two medium-sized trees in a cluster of three near the left side of the image—also shown in a magnified inset). It could be mistaken for a sunlit patch of ground.
Figure 4B. With SIPHER at a threshold of 60, the object in 4A can be seen with a different intensity than the sunlit patch to its right, and a dark border. With a threshold of 80, we get a contrast reversal in parts of the tree trunks, some tree-trunk shadows and most of the ground shadows.
Figure 4C.
Figure 4D. With SIPHER at a threshold of 60, the object in 4A can be seen with a different intensity than the sunlit patch to its right, and a dark border. With a threshold of 80, we get a contrast reversal in parts of the tree trunks, some tree-trunk shadows and most of the ground shadows.
Advertisement

By Herb L. Hirsch

he basis for the SIPHER algorithm was discovered within a U.S. Air Force Research Laboratory (AFRL)-sponsored imagery research project. It is based on the concept of objects being in or out of spatial intensity phase with one another. We first define this “spatial intensity phase” quantity mathematically, then compare it to conventional signal phase relationships, and finally apply it to some images to demonstrate its behavior and utility for discriminating objects. Applications include all forms of image interpretation, from airborne reconnaissance to medical image interpretation.

We define spatial intensity phase (fSI) as an independent variable which contributes, along with other independent variables, to producing an amplitude in image pixels. This is similar to phase relationships producing time-dependent and phase-dependent amplitudes in signal processing situations. Hence, our general equation for this behavior is:

A = f (fSI, V1, ... Vn)

Where: f is a function of this spatial intensity phase (fSI) and possibly other independent variables, V1, ... Vn

A is the amplitude

Analogously, in signal processing terms, for a simple sinusoid we can express some voltage amplitude, V, as a function of the independent variables phase, f, frequency, f, and time, t. Hence, a comparable signal equation is:

V = f ( f, f, t)

V = V0sin(2πft + f)

Where: V is time-varying signal amplitude in volts, f is frequency in Hz (constant if not frequency-modulated), t is time in seconds, f is phase in radians, and V0 is the peak or maximum voltage amplitude.

1 2 3 4 5 6 7 next


Subscribe to our RSS Feeds