So you’re building a color inspection system that requires high magnification. It might be a traditional microscope-based system where the combination of objective and projector lens or eyepiece gives you a total magnification of 40x to 1000x. Or it might be an inspection system using a high-powered telecentric lens to achieve a magnification of 100x or more. Either way, your goal is to be able to clearly capture and analyze some very small details.
When it comes to selecting a camera for this type of system, one important thing you need to consider is something that isn’t actually part of the camera. Or rather, it shouldn’t be part of the camera, but all too often it is. We’re talking about dust or dirt – also called Foreign Object Debris (FOD) – on the camera’s sensor or elsewhere in the camera’s optical path.
Let’s take a look at how this can impact your system and what to look for when selecting a camera in order to minimize any issues.
Dust in the image
Dust specks act like noise in the image. But unlike digital noise, which tends to be random throughout the image and can be minimized by adding light/reducing gain and several other techniques, dust or FODs can be scattered throughout the image, can be clustered in certain areas, or can be in the form of streaks or smears in the image area. More importantly, FODs are physical particles. They aren’t electronic noise that can be tuned out by adjusting the settings.
High magnification tends to make FODs more noticable than in normal machine vision systems, as do the typical lighting techniques used in these systems which silhouette the FODs against the target image.
It is possible to reduce some of the effect of FODs on image quality by making optical adjustments. For example, dust/FODs tend to be less noticeable when larger apertures are used. However, larger apertures also mean shallower depth of field, and that can cause focusing issues for high magnification applications, resulting in issues that are far worse in terms of image quality than the impact of the FODs in the first place.
And for applications where apertures do need to change for different samples, this means the visibility of any FODs will also change from image to image, creating uncertainty in image quality where FODs need to be compensated in one image but not in another.
Most standard machine vision cameras provide reasonably “clean” images suitable for a wide range of applications. Applications that need to detect a specific color in a sample may have no trouble doing so even if there are a few dust specks or FODs in the image. But other types of applications can be much more sensitive to FODs.
In particular, applications involving the counting of cells or other types of small particles can be seriously impacted by FODs in the optical path. For example, an application that counts red blood cells and white blood cells could mistakenly count some FODs as one of the target elements, producing misleading and potentially dangerous results.
Similarly, a semiconductor system that looks for small bridges or breaks in device level geometries could mistakenly classify FODs in the image as defects on the chip or wafer.
Based on the type of system you are creating, you may be able to easily tolerate the low levels of dust/FODs found in typical machine vision cameras. But if your application could possibly be susceptible to the types of issues described above, then it wil be worth your while to look for cameras where additional screening and preventive measures have been taken to maximize dust/FOD suppression. You will also need to make sure that the way these cameras are handled ensures that the high level of dust suppression is maintained.
What you can do
If you need to minimize the impact of dust/FODs on your application, you should start by finding out what sort of specifications your camera vendor uses to classify their cameras. Do they inspect each camera for dust/FODs? Is this done on a pixel-by-pixel basis to fully characterize any potential issues? Is it done under high magnification conditions? Do they guarantee a certain level of dust/FOD suppression?
It is also good to ask whether steps have been taken to ensure that quality is maintained from the time the camera is inspected to the time that you receive it. For example, is the electronics compartment of the camera sealed off from the optical front-end to prevent any migration of FODs from the circuit boards to the optical path? Has the lens opening been carefully sealed and is it designed to inhibit the creation of FODs when lenses are attached?
You’ll also need to make sure that you take the proper precautions on your end to avoid introducing any dust into the equation. This means only attaching/detaching lenses in a cleanroom environment. It also means making sure that the lens or lenses you are using are also kept clean.
If you see specks in your image, it is often these other optical elements that are the culprit, rather than the camera. The objective lenses and the camera adapter optics in a microscope based system, or the front and rear surfaces of a high magnification telecentric lens should be checked before the camera is attached and prior to each critical use. Eyepieces in a microscope-style system also need to be kept clean. This won’t affect the quality of images captured by the camera, but it will provide peace of mind for the observer.
There are many articles available online to guide you in keeping your optics clean. This should be done whether you are using a standard machine vision camera or a camera featuring increased levels of dust/FOD suppression.
Of course, the only way to know for sure what level of dust/FOD suppression is right for you is to evaluate. JAI, for example has standard prism color cameras offering excellent image quality, as well as special models that have been pre-screened to provide even higher levels of dust/FOD suppression.
For additional information or help with an evaluation, contact JAI.