I guess you could call me a die-hard when it comes to what I consider to be a real image. For me, it comes down to whether a real optical image was formed or whether the image itself was generated from some other source, signal or phenomenon. Sure, some will argue that this is all semantics, but in a very real way, the difference is as significant to me as Direct versus Indirect Particle Size Measurement. Part of it has to do with my personal background: I got into this field first as a photographer (who realized it was going to be very hard making a living in that field... a separate topic in and of itself!), then studied Imaging Science in college, followed by many years working in the digital image processing industry.
Real Images vs. Synthetic Images
I've always worked with images of one sort or another: satellite images, aerial images, photogrammetric images, medical images, my own photography (still a prime hobby), and most recently microscopic particle images. The vast majority of the images I have worked with are what I refer to as "real images": they are optical images formed by focusing light [or other electromagnetic waves in the range from infrared (10-5 m) to ultraviolet (10-8 m) wavelengths] onto some sort of 2D sensor array through an optical system.
Along the way, I have also been exposed to different types of what I call "synthetic images". These are 2D arrays of data where the value(s) for each array element (pixel) are converted into intensity or color for display. Both types of 2D arrays (real and synthetic) are referred to and displayed as images, but the difference is that the array value of the real image corresponds to a signal produced by the sensor being struck by the electromagnetic waves from the optical system, and by contrast, the array value of the synthetic image is computed based upon some other type of signal.
As mentioned before, this distinction is truly analogous to the discussion of direct and indirect particle measurements referenced above. A real image is based upon a direct measurement technique, whereas the synthetic image generated by an indirect measurement technique.
Some examples of what I refer to as synthetic images include: CAT scans (the C in CAT stands for Computed!), radar images (including SAR, where the S stands for Synthetic!) and electron microscopy [where the illumination consists of a beam of electrons at a wavelength 100,000 times shorter than visible light, and where the lenses are electrostatic and electromagnetic devices used to control the electron beam (wikipedia reference source)].
Imaging and Particle Analysis
I bring this up because in the particle analysis arena, I frequently hear people discuss imaging techniques which are truly synthetic in nature as though they are working with real images. For instance, many people view the Nanosight instrument as an imaging system and wonder why an imaging particle analyzer like the FlowCam© is not able to also image particles smaller than 1µm in size. This is not to pick on Nanosight here; quite the opposite, their instrument is very powerful and even they are very upfront in saying "we are not imaging the particles."
Part of the confusion is caused by the fact that most people in the particle analysis field do not have the background and education I do in the field of imaging. Most of these people do not understand some of the basic limitations of real imaging caused by diffraction and sampling. A detailed discussion of this can be found in Imaging Particle Analysis: Resolution and Sampling Considerations (requires registration).
However, an additional part of the confusion, in my opinion, is caused by hype marketing; in particular the use of the term super resolution is a pet peeve of mine! Vendors talk about 'super resolution microscopy' as though they have somehow ingeniously managed to beat out the diffraction limit as defined by Abbe back in 1873! In reality, these super resolution systems are using very exotic techniques such as STED, STORM, SIM, and various other alphabet soups. These techniques are extremely complex, expensive and constructive in nature. In my opinion, these are synthetic images. It should also be noted that these techniques are similar to electron microscopy, in that they can only produce a small number of images in a long period of time, and thus are not really appropriate for particle analysis where statistically significant populations need to be studied.
Next time someone is talking about an image, remember that images can be created in many different ways. It's important to understand how they are made, just as it is important to understand how measurements are made with different particle analysis instruments.
Download our free eBook to learn about the importance of utilizing real images when analyzing subvisible particles. Click on the link below.