Fig. 4 | Scientific Reports

Fig. 4

From: An applied noise model for scintillation-based CCD detectors in transmission electron microscopy

Fig. 4

(a) Bias frame averaged across 30 unprocessed images without signal at zero exposure time. It can be seen that the image is offsetted by \(\mu \approx \text {252}\) counts indicated by the grey value in between a 5\(\sigma\) range displayed. Further, brighter areas can be spotted on the detector, where dark currents \(\sigma _{therm}\) are increased and vertical columns are visible, where the Ohmic resistance changes currents. The four quadrants can be seen at their boundaries, where image features are interrupted. (b) After subtracting an image without signal, the result appears without offset, brighter areas are cleared and the image in general has fewer features than before subtraction. Horizontal lines become visible, where the bias of the entire line varies between read-out, because image variation in general decreases. Most of the image variation now is given by the read-out noise \(\sigma _{read}\) changing between pixels on top of the row noise \(\sigma _{row,j}\). This procedure of getting from (a) to (b) is referred to as ‘dark frame subtraction’.

Back to article page