Page 507 - The Mechatronics Handbook

0066_frame_C19 Page 129 Wednesday, January 9, 2002 5:32 PM

Lenses are manufactured to match standard image sizes such as the 36 mm × 24 mm 35 mm photo-
graphic format and the standard television sensor sizes 1″, 2/3″, 1/2″, 1/3″, and 1/4″. These sizes are
defined to be twice the horizontal dimension of a rectangular image with 4:3 aspect ratio so that, for
example, a 1″ sensor has a width of 12.7 mm, a height of 9.5 mm, and a diagonal length of 15.9 mm.
Lens sizes are similarly specified to allow easy matching of lenses to sensors. The maximum angular field-
of-view F OV of a lens focused at infinity is given by


F OV = 2tan −1 C F (19.86)
------
2f 

where C F is the diagonal of the sensor format. For example, a 35-mm lens with a focal length of 55 mm
has a field of view of 43°. Because image distortion and sharpness worsen towards the edges of the field
of view it is acceptable, for example, to use a 2/3″ lens with a 1/2″ sensor, but not the converse. A 35-mm
camera lens generally performs much better, at relatively low cost, than a corresponding C-mount lens
supplied for a TV camera but a C-mount to Pentax, Cannon, or Nikon mount converter is then required.

Image Sensors
The generation of an output signal from a standard image sensor involves up to four operations: the
conversion of the spatial distribution of light irradiance in the image plane into a corresponding spatial
distribution of charge; the accumulation and storage of this charge near the point where it is generated;
the transfer or read-out of this charge; and the conversion of the charge to an output voltage signal. Each
of these operations can be achieved in many ways. Vacuum tube sensors such as vidicons, for example,
use a photoconductive detector material and a scanning electron beam for read-out while CMOS sensors
use a photodiode detector and a readout bus. Solid state devices are currently the most widely used types,
so only CCD and CMOS sensors are considered here. In these devices the image irradiance is measured
on a one- or two-dimensional array of sample regions with positions fixed during fabrication. Each
sample is called a picture element or pixel and the greater the number of pixels, the higher the resolution
with which the image can be recorded. Area sensors are manufactured with numbers of pixels ranging
from tens of thousands to several million. Color sensors are achieved by placing color filters over the
individual pixels, in a mosaic or stripe pattern, and interpolating the color values at pixels where necessary
from the neighboring values. In such color devices the color resolution is lower than the luminance
resolution, but this is not important for many applications because the resolution of the human eye is
worse for color than for luminance. More expensive color cameras use three precisely aligned sensors,
one for each primary color. Cameras incorporating such sensors generally produce either a television
standard signal [8] (RS-170 monochrome and NTSC color for 525 American television or CCIR mono-
chrome and PAL color for European television) or a digital signal such as RS-423, USB or IEE 1394 Fire
Wire, which can be readily connected to a computer.

Charge-Coupled Devices
6
In a charge-coupled device [9] an isolated packet of charge, of between 10 and 10 electrons, is moved
through the semiconductor, from a position in one CCD cell to a position in an adjacent cell, by applying
a sequence of voltage pulses to gate electrodes. In CCD-based light sensors, photon generated charge
packets accumulate in photosites, which are modified CCD cells, and are then transported through other
CCD cells to another modified cell with a readout amplifier attached. A CCD is fabricated on a single
crystal wafer of P-type silicon and consists of a one- or two-dimensional array of charge storage cells,
on centers typically about 10 µm apart. The operation of a 3-phase CCD cell is illustrated in Fig. 19.107.
Each cell has three closely spaced electrodes (gates) on top, insulated from the silicon by a thin layer of
silicon dioxide. A positive voltage applied to one of these gates will attract and store any free charge
generated in the silicon due to light or thermal action while free holes are repelled and collected by the
substrate electrode. Lower voltages on the adjacent gates isolate it from the neighboring cells, creating a

©2002 CRC Press LLC