The performance of digital cameras is improving at a considerable rate and so it is important to understand the factors that are likely to limit their use in the fields of surface inspection and measurement.We describe here test methods for comparing the relative quality of digital cameras using only a computer.
The choice of digital camera for a particular purpose requiring specified information content, includes consideration of the specification of the imaging lens, the number of pixels required in the image format, and the size of each pixel. This, together with the lens, determines the resolving power achievable. Unfortunately, the manufacturers usually only provide the nominal number of pixels, and even this information may include pixels providing color and edge smoothing rather than resolution. Fortunately, it is not usually necessary to measure the completeMTF of the camera systemto obtain an approximate measure of the limit of resolution. This is because the image processing electronics provide a measure of equalization of the MTF over the spatial frequencies resolved. This ensures good edge sharpness in images and we only need to determine the limiting spatial frequency or spatial frequency bandwidth (SFB) in cycles/mm. The inverse of this number will be the effective pixel or pixel cluster size in cycles/mm. In resolution terms a cluster of two pixels will be required to detect one cycle.
Online access to SPIE eBooks is limited to subscribing institutions.