Foresight Imaging and Image Accuracy
Foresight Imaging makes the highest accuracy frame grabbers. Period. There are several reasons for this. First, we have the best pixel jitter specification in the industry (±0.5 ns). This ensures that you are sampling each pixel where it should be sampled. Every time. Every line. Second, we have an excellent signal-to-noise ratio: 59 dB. This provides you with the confidence that noise will not corrupt your image measurements. Third, we employ superior analog design. Among other things, this ensures a fast settling time (even at high pixel rates) which results in true pixel value readings during extreme transitions from black to white. Other frame grabbers can introduce incorrect gray level values because they cannot settle quickly enough. And since most image measurements involve finding an edge characterized by a sharp change in gray level, this level of precision is essential in attaining accurate measurements. Fourth, we use 10 bit A/D converters on our AccuStream product family frame grabbers for additional precision. Lastly, our engineers have been designing frame grabbers for 12 years and have the experience necessary to put this all together to produce the highest accuracy frame grabber available. If your imaging application involves any type of measurement or diagnosis, you owe it to your company and your customers to obtain the highest accuracy possible. Only Foresight Imaging frame grabbers deliver this accuracy.
Figure 1 - Foresight Imaging Frame Grabber Image
Figure 2 - Foresight Imaging Frame Grabber Histogram
Figure 3 - Other Frame Grabber Image
Figure 4 - Other Frame Grabber Histogram
The Image Accuracy Test. At Foresight Imaging, we are constantly testing our frame grabbers (as well as those of other manufacturers) for image accuracy. One of the most difficult tests of a frame grabber's accuracy is to input the following pattern: a one-pixel wide black vertical line followed by a one pixel wide white vertical line repeated all the way across the image. On some pattern generators, this is known as the Grill_11 pattern. It is also commonly referred to as a "one pixel on - one pixel off pattern." Above are examples of the images (zoomed to show detail) actually acquired from the Grill_11 input pattern by a Foresight Imaging frame grabber and another popular frame grabber. From the corresponding histograms, you can see that the Foresight Imaging frame grabber acquired a pristine image with sharp histogram peaks at both black and white. The other frame grabber failed miserably with significant phase and settling time problems. A useful measurement could never be obtained from a frame grabber with this level of accuracy. The image was not even visually useful until the pattern lasted for at least three pixels, but still showed phase and settling time problems in the histogram data resulting in significant measurement errors.