slide 43

To get useful information, we need to remove the random component phi scatters. Interference is the key to removing it. By comparing equations and facts from two different star images, we can approximate phi scatters 1 to phi scatters 2. Interfering the two images allows us to remove phi scatters, leaving only delta r, which contains topography and displacement information. Interference is important because it removes random components in the face. So, in order to get useful information, we have to figure out how to remove the random component phi scatters. Interference is the key to removing it. We can look at the equations to know the process. Assume we have a fact at pixel P inside image 1, phi 1, consists of the distance, we call it is range, and phi scatters 1. We also have a fact at the same pixel P in the star image 2, phi 2, consists of the range and phi scatters 2. Note that the range can be converted to fact. The relation is fact is equal to minus 2 times 2 pi times range over lambda. The reason for a 2 is that the range is two-way path. We then look at the right figure. Because two star images are acquired at different positions, their range must be different. We denote the range difference as delta r. If scatters properties in the pixel P are similar in the two star images, phi scatters 1 is approximated to phi scatters 2. So, if we do interference between two star images, which means phi 1 minus phi 2, we can remove the phi scatters. The remaining component is delta r, and delta r consists of components of topography and displacement between two star acquisition times. So, the important reason to do interference is that interference can remove the random components in the face.