SubscribeThe Night Vision Thermal Imaging Systems Performance Model, or NVTherm, is an extremely useful tool that is commonly used to calculate range performance, which is the distance upon which a given thermal imager can produce an image of a given target.
This image produced is then classified as detection, recognition or identification depending on the quantity and quality of data collected about the target. This tool was developed by the U.S. Army Night Vision and Electronic Sensors Directorate and can be used to predict range performance for a variety of mid-wave infrared (MWIR) and longwave infrared (LWIR) sensor types.i It considers the influence of many parameters including, but not limited to, atmospheric attenuation, system noise, detector pitch and array size, optical blur, lens attenuation, focal length, aperture size, even the type of monitor being used. The model then uses these parameters to calculate estimates for expected range performance at detection, recognition and identification levels.
Night Vision Thermal Imaging Systems
As noted above range performance estimates are typically described in terms of target detection, recognition or identification. These terms
are clearly defined but often misunderstood. Target detection suggests that a target may be seen in the image as a small, but detectable, blur. It simply means that the target is visible with minimal number of pixels, as few as one pixel element, and that there is a reasonable probability that what is visible is something of interest.ii It is very useful for knowing when an area needs further investigation because some unknown intruder is present.
Classical recognition, or class discriminationiii, is a commonly misunderstood range performance category. Recognition of a target is being able to discern it with sufficient clarity that its specific class can be differentiated.iv This requires a small cluster of pixels on the target, and suggests that the operator could distinguish between a person and a car, or between a truck and a tank, etc. Often an operator will want something more than the standard definition for recognition when evaluating a potential threat in the image making it important to choose a camera system that offers a range performance comfortably greater than what your expected requirements really are.
Identification, or object discrimination, of a target implies that there is an ability to discriminate between objects, or to discriminate the correct vehicle, not just the vehicle type.vi Neglecting atmospheric effects, the recognition range would be 25% of the detection range, and identification range would be 12.5% of the detection range. In the real world we cannot neglect atmospheric effects so the recognition and identification ranges are generally cut even more than these percentages - the extent to which is determined by the atmospheric attenuations assumptions that are
used (e.g. clear weather is obviously going to yield greater range than a hazy day).
