How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
For centuries, the primary focus in microscopy was achieving higher and higher spatial resolution. As technology improved, optical microscopy approached, then surpassed the diffraction limit. Meanwhile, scientists perfected non-optical techniques such as atomic-force microscopy (AFM) and scanning-electronic microscopy to delve ever deeper into matter.
Although researchers have not given up the hunt for ever better resolution, the trend in microscopy today is less about perfecting a single technology and more about developing hybrid systems that leverage two or more approaches to gather more and better data about a sample. Groups are augmenting optical microscopy with contact techniques like AFM, or using two different optical methods, or even going further afield. “The trend is to get more and different properties,” says Arizona State University (ASU; Tempe, Ariz.) professor Nongjian Tao. “For example, most microscopes do not have the capability to provide chemical information. In general, microscopy’s advantage is to provide spatial resolution. The main goal here is to get additional information.”
JUST A PHASE
At the University of Illinois, Urbana-Champaign, a new technique called diffraction-phase microscopy (DPM) allows researchers to simultaneously image a sample and capture quantitative information such as structural thickness or refractive index. Somewhat akin to a hybrid of phase-contrast microscopy and holography, the system involves an ordinary optical microscope operated with dual beams to introduce phase-contrast capabilities. Unlike a conventional interferometer configuration, however, the beams do not split at the source to travel spatially distinct optical paths. Instead, a diffraction grating at the output port of the microscope splits the image field into two beams: one that passes through a spatial filter and one that does not.
As with holography, the biggest challenge is mechanical noise such as phase disturbances introduced by shifting air currents, vibration, etc. Although active stabilization techniques like feedback loops can cancel out noise, they add complexity and latency to the process. The approach used in DPM is far simpler—eliminate the noise physically by having the two beams travel almost the same optical path, through a single set of objectives. Because they effectively encounter the same noise sources, the noise introduced is nearly identical. As a result, the phase difference between them, which is the primary objective, becomes essentially noise free. “Now you can get any imaging system that gives you an image plane and split the beam there as opposed to starting with two beams,” says team leader and assistant professor Gabriel Popescu. “You can transform any imaging system into a phase-imaging system, as long as the light is spatially coherent.”
The system can achieve nanoscale accuracy in conjunction with motion. It is important to note, however, that this is not a measure of the system’s resolution. “We do not see objects that are nanometers in size,” says Popescu, “Our sensitivity to the nanoscale comes from motion or topography changes. Let’s say the membrane of the cell goes up by 1 nm, we would pick that up, but that’s completely different from resolving two objects separated by 1 nm, which is what the super resolution methods are aiming at.”