Building Earth Observation Cameras

145

6

Pushbroom Imagers

6.1 Introduction

We have seen in the previous chapter the limitations of an optomechanical

imaging system in generating high spatial resolution imagery from space

platforms. The limitation primarily arises from the mode of imaging on

an instantaneous geometric eld of view (IGFOV) by IGFOV basis, thereby

limiting the time to collect the radiation. The situation can be substantially

improved if data are collected for an entire strip by imaging the complete

swath at a time, since all the pixels of the strip get the integration time cor-

responding to the time taken by the satellite to move through one IGFOV.

Such a mode of operation is possible due to the development of long array

detectors.

6.2 Principle of Operation

A strip of the terrain is focused by an optical system, say a lens, onto the

linear detector. At any given instant of time, only those points on the ground

that lie in the plane dened by the optical center of the imaging system and

the line containing the sensor array (the optical center is the imaginary point

at which all the rays intersect are imaged). This plane may be referred to as

the instantaneous view plane. The radiation from the strip on the ground is

received simultaneously by every detector element of the sensor and each

detector element produces electrons proportional to the radiant ux received

by that detector element and the duration for which the detector is exposed,

also referred to as the integration time. Depending on the number of elements

in the detector, one line of information generates that many picture elements.

When such a system is mounted on a moving platform (usually such that the

array length is at a right angle to the velocity vector of the platform), as the

platform advances the view plane sweeps out a strip of the terrain that is con-

tinuously projected onto the charge-coupled device (CCD) array. Normally,

146 Building Earth Observation Cameras

the detector is exposed for a duration equal to the time taken by the subsatel-

lite point to move through one IGFOV, referred to as the dwell time. Thus, due

to the platform motion every successive exposure produces contiguous image

strips. Hence, a two-dimensional image is produced—the linear array detec-

tor producing one image line across track (XT) and successive image strips

along track (AT) by the motion of the platform (Figure 6.1). When compared to

optomechanical scanning systems, the scan mirror is avoided and instead the

detector array is used to produce one scan line of information. This mode of

scanning is referred to as pushbroom scanning. In this case, the total time taken

by the subsatellite point to move through one ground resolution element is

available for integration of the signal for all the pixels along the linear array.

The dwell time for a pushbroom scanner is given by

τ=

β

p

v

h

(6.1)

where β is IFOV, v is the velocity of the subsatellite point, and h is the satellite

height.

Swath

T

T+τ

Optics

View plane

Linear CCD

Direction of spacecraft motion

FIGURE 6.1

Illustration of the pushbroom scanning technique. τ is the integration time.