Digital microscopes are being automated and computerized to make them easier to use, display more data with more detail and precision, and expand their areas of application. With traditional microscopes, you need years of experience in order to take full advantage of the instrument. Digital microscopes can reduce the learning period by automatically selecting many of the settings to optimize an image. They can focus, zoom, polarize light, move filters in and out, and utilize tools like differential interference contrast (DIC) prisms.
The Olympus (Waltham, MA) Model DSX, for example, has a feature called Best Image, in which the microscope takes several different images, trying different adjustments, then outputs a sample of nine of them. The user can select the one they like best by clicking a button on the GUI — the microscope settings are then automatically configured for the selected image. (Figure 1)
With all the variables controlled by the system, the microscope knows its chosen magnification and all of its calibrations. With a traditional microscope, if someone changes a zoom setting, or an objective and doesn't input those changes to the software, they're going to make mistakes. With a digital microscope, the microscope stores all of the settings and can repeat them any time in the future.
The controlling algorithms can reside in the camera, or, as in some of the older cameras, in computer software. If the processing takes place in the camera, it tends to run faster, which gives lower-latency live images. But regardless of where the algorithm is located, camera and software should be considered as a single integrated system.
Software can be trained to look for specific elements in the image. With repetitive measurements, since you're always looking at the same type of samples, you can teach the software what the image looks like and where the thresholding is. This could be based on a gray scale or a color value. The microscope can look for changes in the image. By teaching the software where those changes should be, it can learn how to detect edges and how to make measurements.
Electronics: An important application in electronics, is examining circuit boards. By measuring the trace, knowing the amount of material, spotting any scratches or defects, you can determine how much power can be run and at what signal speeds.
Metallurgy: With specialty metals, there's a need for critical measurements of welds: the size of the weld material, how solid it is, how good the contact is. One specialized weld measurement is throat thickness, where you can see how much into the corner the weld material is going and how far out it's coming, so you know how solid it is.
If destructive testing is an option, a cross-section of the weld can be inspected to ensure that it was done correctly. A good example is in automated assembly lines, for example, automotive. Once a day a car is pulled off the line, cut into pieces, and multiple welds are cross-sectioned. If an existing or potential problem is seen, where a weld is starting to be out of tolerance, you can go back and make corrective adjustments. (Figure 3)
Composites: There are new composite materials — carbon fibers, for example — where proper weave is critical for the strength of the material. You can view the strands and the orientations of the weaves. You can use the microscope to see that the layers are correctly in place. Then as you're gluing the fibers, you can see how well they're bonded together and that they are orientated correctly.
A useful feature for all of these applications, is the ability to put a sample up on the screen, spin it around, zoom in and out, pan across the image, and take measurements in two dimensions or in three dimensions. With a good 3D digital microscope, for example, you can even look down into a hole and use the software to measure its size, how deep it is, the curvature within it, and the roughness of the surface. It can also calculate the volume in order to determine the dimensions of a piece designed to mate with the hole.
Material microscopy presents challenges that don't exist with life science microscopy. Living tissues are usually so thin and translucent that light passes through the specimen. In material science, however, light is bounced off the sample. With a printed circuit board, a shiny piece of metal, a black piece of carbon fiber — because these materials are so dramatically different — light bounces off the samples differently. To obtain a good image in the microscope, is therefore a real challenge. The type of illumination and its control is, therefore, critical. When traditional halogen light bulbs heat up, or as their intensity is varied, the color temperature changes, so if you're not constantly monitoring that, your sample will look different at different light intensities. Heat is another problem: traditional light sources can heat the sample, which can cause the metal to expand and change measurement values. LED light sources overcome these problems. They're cooler, have a longer life, and very importantly, their color temperature stays constant even as the intensity changes.
Image clarity is improved by using advanced high-definition digital techniques, such as High Dynamic Range (HDR), which takes different images at multiple exposures. With a sample that has both dark objects and shiny reflective objects, the reflective surface will either be blinding white or if you dim down the light, everything else in the image will look black. HDR takes multiple pictures at multiple exposures to rework the image, optimizing different pixels. It'll make the dark spots look lighter and it'll dim down the shiny areas just enough to produce an image with a more even look.
A technique for providing an accurate image of metals, is to finely polish them. But that produces glare (halation). This effect can be removed by lighting the sample in different ways, in different directions. The lighting level can be varied directly or the camera exposure can be changed to alter the appearance of the lighting. With digital scopes, since lighting is typically integrated into the instrument, the GUI can control it. If you change one setting, the system will adjust another to compensate — always trying to produce the best image.
When an Inverted Microscope Will Help
For a traditional microscope, the size of the sample has to be limited to fit under the instrument. With an inverted microscope, all the optics are built into the base — it is essentially flipped upside-down. So, large samples can be placed on top of the microscope, since there are no physical obstructions. One application for placing the sample on top, is to do a fine measurement on a small part of a very large sample.
Another thing that can be challenging with microscopes, is finding focus. An inverted microscope is designed with a narrow focus range. This allows you to put samples upside-down and almost guarantee that they're in focus. If you cut and polish a piece of metal, by flipping it upside-down and putting it on the inverted microscope, it's not only easier to get it in focus, but to also keep it in focus as you move around on the sample.
There is important data that can be displayed in addition to the image itself. With a digital microscope, you can also be looking at height data. The microscope can keep track of where the image is in focus, which enables it to create a height map that can be stored and displayed within the image. Another level of data, can be the lighting conditions when the image was taken. At a research lab, you're looking at a lot of different materials, so if you come back to it a couple of weeks later, you want to be able to repeat the configuration parameters. That too, can be stored and displayed within the image. In addition, people can input sample information as part of the display: time and date stamps, text comments, pass/fail results. A great deal of metadata can be embedded in large image formats such as TIFF.
One of the most important advantages of digital microscopes is the ease of using them, which removes complexity for new users. Learning how to operate a microscope is an important step for getting a new employee up to speed more quickly. With easy-to-use GUIs and interfaces, it's easier to obtain an image and easier to take a measurement.
The current trend will continue to be improving the software to make processes easier through increased automation.
This article was written by Marc D. Silverstein, Product Manager, Industrial Microscopy, Scientific Solutions Group, Olympus Corporation of the Americas (Waltham, MA). For more information, contact