What is super smear rejection?
The photo on the left shows an image of a back illuminated rectangular aperture with an exposure time of 1ms in a stretched scaling (48-100 counts correspond to 0-255 gray levels). The white line indicates the position of the readout pixel row shown in the next graph. The right photo shows an image of the same aperture with the same scaling at 100ms exposure time and reduced back illumination. The resulting right image has the same intensity values like the left image.
If an interline-transfer-CCD-image sensor is read out, the generated charge carriers are transferred from the light sensitive part (photodiode) of a pixel to the shift register, which is next to the photodiode and acts like a potential bucket. Because of the weak light sensitivity of these shift registers, they are shaded to prevent the additional generation of charge carriers. When the charge transfer to the shift registers is complete, they are shifted vertically row by row to the horizontal readout line, which is read out in a serial way. This vertical shift of row by row can be best compared with an endless conveyor belt process. With every shift step all registers are shifted by one register location including register rows that were already drained from the image.
If an intense light source is imaged to the CCD-image sensor, this can generate unwanted charge carriers in the shadowed shift registers (mainly due to scattering of light on the CCD chip) in every line or row, which is shifted below the spot. This results in unwanted additional light signals, called “smear”, which can be recognized as bright vertical bands above and below the bright image spot. This phenomenon is more often seen at very short exposure times (because the light signal has to be intense to achieve a good image with a sufficient signal-to-noise ration) and is a characteristic of the corresponding CCD-image sensor.