Pixels over-illumination

The upper dynamic range is limited by the amount of charge that can be quantitively stored in a single pixel well. Over-illumination of a pixel causes the charge to spill over into surrounding pixel wells, a phenomenon known as "blooming". 

Figure 7.1 Land placed two sheets of papers with reflectances of 40% (a) and 80% (b) next to each other. The illumination was chosen such that the amount of light reflected from the center of one sheet of paper was equivalent to the light reflected from the center of the other sheet of paper. The left sheet appears darker than the sheet on the right. However, if a pencil is placed over the edge where both sheets join, then both sheets seem to have the same reflectance. The image shown here was generated artificially. The graph on the right shows the pixel intensities for a horizontal line of the image.

Apart from producing a constant output irrespective of the illuminant, the entire color space should be used. We can compute the standard deviation of the pixels of a single image in RGB chromaticity space and average the standard deviation over all images of the image set. This will measure how well the available color space is used. The color... 

Figure 10.4 (a) IR illumination over the 64 x 64 FPA detector in reflection mode from a clean Keveley Ag-Sn02 coated microscope slide, (b) The figure shows representative spectra of a protein-based sample from four pixels along the same pixel column closest to the periphery of the detector and designated by the arrow. 

In essence, HR-CS AAS offers only advantages over LS AAS, and eliminates most of its notorious limitations. Firstly, as the instrument uses a CCD array with several hundred pixels, it is equipped with hundreds of independent detectors (200 of which are typically used for analytical purposes), which are all illuminated and read out simultaneously. Since only a few pixels are necessary to measure atomic absorption, the rest of them are available for other purposes. The... 

Practically this means that high signal frequencies are detected at bright illumination while poor illumination results in low-pass filtering in each pixel. Log amplification inserts an amplification noise proportional to the amplification factor. In contrast, exponential attenuation results in a constant SNR over most of the range because the noise is proportional to the square root of Af Fig. 7.8.6 shows the difference in dynamic behavior of a CCD and a nonlinear CMOS sensor. 

---