A simple approach for calibrating solid-state imaging systems

Abstract

Imaging by solid-state array sensors has become an indispensable tool in various modern scientific and engineering applications. In such instrumentation, a linear response of the sensor output to the incident light intensity is often very much desirable. The system linearity can be affected by both the sensing element and electronics. The linearity also needs to be accurately determined in order to assess and correct measurement errors. Some imaging systems, especially those of upper grade with high resolution, allow precise control of exposure time. If so, it is easy to check the linearity in response to the incident energy with changes of exposure time under a constant intensity of illuminating light. For imaging systems without precise time control, especially those of intermediate- or lower-grade, it is highly desirable to have some means for assessing linearity and applying corrections to improve accuracy for photometric uses. Here, we present two simple approaches, one for quickly checking linearity and the other for correcting the response, for those imaging systems without exact exposure-time control. The former allows a quick assessment of the linearity while the latter permits an accurate calibration with minimal hardware preparation. In initial testing, the calibration method works well with an intermediate-grade sensor interfaced to a frame grabber. The nonlinearity of the imaging system tested can be reduced from about 5.1% to much less than 1% in standard deviation.