The elementary picture element (or photosite) for this device is the photodiode, that is a light sensitive, reverse biased P-N junction. The operating principle is illustrated in Figure 1:
When the switch S is open the charge is reduced at a rate proportional to the incident radiation (the photocurrent Ipc) and the thermally generated hole-electron pairs (the dark current Ide). For modem, high quality P-N junction (where Ipc » Idc) the discharge time constant is of the order of several seconds for zero incident radiation and in the milliseconds region for typical irradiation levels (Hill, 1982): the charge stored within each photosite P-N junction is therefore inversely proportional to the incident irradiation level. Monitoring the rate of change of this charge (i.e. the capacitor current ic) while scanning the photosites provides a voltage Vout proportional to the grey scale image of the scene focussed on the array, as required. This technology favours uniformity and stability of response but does allow a lower array resolution than other solid state cameras in view of the more complex ‘on chip’ support electronics required.
The most common geometry for industrial photodiode arrays is in fact the linear array (or linescan) camera. This s due, as previously mentioned, to the technological difficulties involved with making large 2-D photodiode arrays and their relative high cost.
Linescan cameras offer a simple, low cost method of detecting edge position which, in turn, can be used to measure object size parameters such as width or length (IPL, 1984; Electronic Automation, 1985; Hollingum, 1984; Drews et aI., 1986), as shown in Figure 2 (for further information on how to separate the object image from the background and obtain its edges positions.
A device which uses a linear photodiode array to measure object width based on this technique is shown in Figure 3(a). This device can be used in arc welding path control vision systems where the metal channel width and its position with respect to the welding torch need to be measured in real time in the presence of severe electrical and optical interference.
It can also be seen from Figure 3 that if the lenses field of view is fixed and the light source is capable of producing a sharp dot on the measurand surface irrespective of its distance then the same arrangement can be used for measuring range. This is, in fact, the principle of operation of some single point rangefinder sensors.
These devices use laser diodes to produce the required collimated light beam and are employed in automated manufacturing applications where non-contact range measurement is essential, as is the case in the food and confectionery industry (Electronic Automation, 1984), an example of which is shown in Figure 3(b).