During the integration phase (exposure time) all electrons released by the incoming light at the semiconductor-oxide interface are accumulated in a potential well. The accumulated capacity is proportional to the amount of the incident light and the exposure time. The capacity per pixel is then read out following several sensor read out techniques that are described in the following:
In addition to light sensitive pixels, interline transfer (ILT) sensors include vertical shift registers. Instead of shifting the charge through the light sensitive pixels, it is moved into a shielded shift register situated next to each line of pixels. The charge is moved to the vertical shift registers in one step before reaching the horizontal readout register and being clocked into the output amplifier, pixel by pixel. The reduction in fill factor imposed by the presence of the shift registers can be compensated for by microlenses to improve sensitivity. Although the full well capacity is reduced compared with other CCD architectures, well sizes of 30-50 KeV are commonly available on Sony and ON Semiconductor CCDs, which is adequate for 8-10 bits of digitisation. ILT CCDs are the most common type of sensor for modern vision cameras as the single step shift to the readout register allows for short exposure times and suitability for fast moving images. Over the next few years it is expected that the majority of applications that would have used interline CCDs will be using CMOS sensors.
The full frame CCD shifts each pixel through all of the pixels directly below on a line by line basis until it reaches the horizontal shift or readout register.
The frame transfer sensor uses the same method of transferring the charge through the pixels, but it can do so much more quickly as the charge is destined for a shielded readout area next to the active area where the time consuming pixel by pixel readout is performed in the same way as the full frame device.
Both, frame transfer and full frame architectures often reach 100% fill factors with large full well capacity that yields a high dynamic range. Both types of sensors frequently use large pixels with dimensions of 12 or 14 μm, resulting in better dynamics and higher sensitivities and are often found in scientific and high end professional photography applications where speed is less important than dynamic range and extremely low noise.