This chapter derives a formula used to determine the measurement accuracy of K ADC (e â âDN) . To show the error's origin graphically, a shot noise PTC plotted on linear scales is presented in Fig. 6.1. The data were simulated using a random number generator based on a 500-pixel subarray. The data points are scattered about the ideal solid shot noise curve; the variance in data about the curve reflects upon the error in measuring K ADC (e â âDN) . A histogram of K ADC (e â âDN) with a standard deviation of 0.645 is shown in Fig. 6.2. The histogram in Fig. 6.3 shows an improved standard deviation of 0.141 DN when 10,000 pixels are sampled. These results suggest that the uncertainty in ADC sensitivity decreases by the square root of the number of pixels sampled. It would also be informative to know how read noise and signal level influence measurement accuracy. The analysis below includes these variables when K ADC (e â âDN) variance is calculated.
Online access to SPIE eBooks is limited to subscribing institutions.