We are pulling together climate related resources.

Autofocus System and Autofocus Method for Focusing on a Surface

Stage: Prototype

This microscope autofocus method is directly applicable to the U.S. Department of Energy (DOE) Defense Nuclear Nonproliferation mission. Analysis of microscopic particle samples is critical to meeting the criteria of nonproliferation of nuclear weapons treaties. This system speeds up this process by automatically creating sharp microscope images for samples on transparent slides for automated analysis. Beyond its use in the DOE, it can be licensed to microscope manufacturers for inclusion in their systems since the technique can be optimized for all types of samples and illumination methods. In addition, it can be applied to other imaging techniques. A prototype of this system was delivered to NIST.

This autofocus system includes an imaging device, a lens system and a focus control actuator that is configured to change a focus position of the imaging device in relation to a stage. The electronic control unit is configured to control the focus control actuator to a plurality of predetermined focus positions and activate the imaging device to obtain an image at predetermined positions and then apply a spatial filter to the obtained images. Typical autofocus techniques find the best focus on an object of interest in the image and are optimized for typical biological samples. When a sample has multiple objects of interest on or near the surface and has other objects (such as labels) that are spatially larger and are substantially above or below the surface, the best focus is often ambiguous to the existing algorithms. This new algorithm works equally well on the typical samples and these more difficult samples. By applying a spatial filter to the image, both the edges of objects of interest (which are at the surface) and any surface mottling produce the optimum result, making this method effective for all types of samples and illumination methods. The filtering system can be employed in one of two methods, the first of which is a computationally very fast method; the second method is computationally slower and provides increased accuracy. The implemented filtering technique minimizes both memory and processing power making it comparable to some of the most efficient techniques investigated.

Applications and Industries

  • Nonproliferation
  • Detection of fissionable material
  • Any application using microscopes


  • Increased speed of obtaining clear images
  • Increased sharpness for ambiguous samples
  • Targeted objects for focus