Figure 6: (a) A typical first image during the acquire phase. The image of the star through the beam splitter is the brighter spot to the left the fainter image to the right is the image from the mirror around the fiber entrance. The read-out stripes result from the shutter-less design. (b) The probability function used for identifying stars from their measured image elongation (y, to the back) and sharpness (x, to the right). The green, constant level at a probability of 0.446 is the threshold above which stars are identified. The two-dimensional function in red was derived by a manual training of the acquire system on 100 different images.