Selecting Lenses to Maximize IR Camera Performance

Peter Kornik, Janos Technology Inc.Peter Kornik is commercial product marketing manager for Janos Technology Inc. in Townshend, Vt.

Many factors must be considered when choosing an infrared lens for a thermal imaging camera. These include knowledge of the technology being used and of the intended application for the imager. It is also essential to have a good idea of what results must be achieved in light of the price range budgeted for a new IR camera lens (the entire lens assembly mounted to the camera). Knowing the critical aspects of IR lenses facilitates the selection process.

There are three main regions, or wavebands, of sensitivity most common with today's IR cameras. The first is the near- or short-wave IR, which spans approximately 0.9 to 2.5 µm. Next is midwave IR, roughly from 3 to 5 µm, and finally long-wave IR, from about 8 to 12 µm. Some IR cameras can work outside of these areas, but they are generally optimized for them.

These camera lenses are, in fact, designed to operate over a set waveband, whether it is one of the bands listed above or more than one band; e.g., 1.5 to 5 or 3 to 12 µm. When designed for a particular waveband, many factors influence their performance, including material selection, lens thickness, air spacing, surface curvatures and coatings.

This said, an IR camera, for example, may be optimized as a long-wave device but have some low level of sensitivity down to 3 µm. For a particular application, there may be a requirement to use this camera's full range of sensitivity. If a lens is designed for the long-wave IR but is merely coated differently to achieve maximum transmission over the complete range, the image quality would be very poor below 8 µm. A lens would have to be designed specifically to perform over the full range of 3 to 12 µm by taking the factors listed above into account to ensure image quality.

Image size

The lens will need to create an image that will fill the focal plane array, or detector, of the IR camera. These arrays are rectangular or square in shape, but a lens creates a circular image on the focal plane. This requires that the lens create an image that has a diameter equal to, or larger than, the diagonal of the array. If the image does not fill the detector area, the resulting effect is commonly referred to as vignetting. This will appear as fuzzy, gray or darkened corners and/or edges of the image, depending on the severity the lens vignettes.

The only exception to this rule is when dealing with fish-eye lenses, which create a hemispheric image within the dimensions of the focal plane array. If the diagonal of the array in the IR camera is not documented, it can be calculated using simple trigonometry if the number of pixels and the pixel pitch are known; e.g., 320 X 240 pixels and a 50-µm pitch equates to a 20-mm diagonal.

Back working distance

The distance from the back of the lens housing to the focal plane array is the back working distance. A lens must be properly mounted so that it images at the same location as the focal plane array in the IR camera. Generally, IR cameras that require something between the lens and the array will dictate the use of a lens with a longer back working distance. In some cases, it is possible to reduce the size of the lens by minimizing this distance, so it can be advantageous to be as close as possible to the array.

Back working distance is more a characteristic of the lens than of the IR camera. A camera with a short back working distance can accommodate a lens with a long one by using an adapter or spacer. For example, an IR imager requiring a minimum distance of 10 mm can be fitted with a lens that has a back working distance of 30 mm employing a 20-mm spacer. However, the reverse is not possible. A lens that images at a distance of 10 mm cannot be properly positioned on an IR camera that needs a minimum distance of 30 mm.

Focal length

Lenses are commonly identified by their focal length, sometimes referred to as effective focal length. As the focal length increases, the field of view for that lens will be narrower. Conversely, as the focal length decreases, the field of view widens. To calculate it, take the inverse tangent of half the image area divided by the focal length and multiplied by 2. For example, for a 50-mm lens on a 20-mm-diagonal focal plane array, the formula to determine the full field of view is as follows:

[1/tan{(0.5 X 20 mm)/50 mm}] X 2 = 22.6

It is important to know whether a lens is specified by the horizontal or vertical field of view. This is the most common form that an IR camera company will use to specify lenses for a particular model. The horizontal number will help identify the angle that spans the width of the focal plane array, while the vertical field of view gives the angle spanning the height of the array. The same formula can be used to calculate horizontal and vertical field of view by substituting the applicable array measurement.

Manufacturers also categorize lenses based on their focal lengths. Many longer focal length lenses are termed as follows:

Telephoto, when their physical size is shorter than their effective focal length.
Normal, when lenses create an image close to what can be viewed by the human eye.
Wide angle, when lenses produce a scene more spacious than normal vision. Lenses with a field of view greater than 150° are commonly called fish-eye.
Two of the most common specialty lenses are multifield and continuous zoom lenses. The former is designed to switch between two or more focal lengths, while the latter can remain in focus anywhere between two boundary focal lengths. These types of lenses can survey a scene and magnify an object within that scene switched or zoomed. For example, a 50/250 dual-field-of-view lens can be switched from a 50-mm focal length to a 250-mm one to yield a 5X magnification of the object being viewed.

The f number

The f number of a lens dictates the amount of light -- or more accurately, in the case of IR imaging, the amount of energy -- transferred onto the focal plane array. This number can signal how significantly the lens influences image sharpness. The lower the f number of a lens, the larger the optics will be, which means more energy is transferred to the array. The f number is also referred to as the speed of the lens. For example, an f/2.3 lens may be considered faster than an f/4.0 lens.

Certain IR cameras are configured with an aperture between the lens and the focal plane array. This is common in cooled imagers, as well as in a few uncooled microbolometers in the form of a diameter opening through the middle of a physical shield. The function of this shield is to block unwanted radiation. For these cameras, the size of this aperture determines the required f number of the lenses used.

This aperture (also referred to as an iris or a warm, cold or radiation d shield) will be positioned at the pupil location. The optimal lens used on these types of IR cameras should match the required f number at the proper aperture or pupil location, although a faster lens could be used.

When cameras have no aperture, there is much more flexibility in the f number chosen for the lens. However, these are typically uncooled long-wave IR devices and are generally less sensitive to energy. A lower f number would be desirable to create the sharpest image; or, depending upon the application, this could be an area to make trade-offs to minimize lens size and weight.
Depth of field

Depth of field is determined by using the closest and the farthest objects that appear to be in focus. When a lens is focused in such a way that objects out toward the horizon are in focus, it is referred to as the infinity focus position. At this position, the closest object that is in focus is termed the hyperfocal distance.

For a particular application, it may be necessary to get in close to an object for maximum magnification. The minimum object distance will determine how close someone can view an object by adjusting focus. In the case of a lens that has a hyperfocal distance of 50 feet, it is possible to adjust focus to view an object as close as 10 feet, which would be considered the minimum object distance. The user can then expect everything closer than 10 feet to be out of focus.

There will be a small range beyond 10 feet that will appear in focus -- the depth of field -- but this is a minimal range and very subjective. In visible photography, slower f number optics are commonly used to increase depth of field. However, as reviewed earlier, IR camera lenses are typically fast optics, so depth of field is sacrificed for image quality.

Performance

Because most people have little knowledge of lens design, it can be difficult to judge lens performance outside of a subjective visual evaluation. The most common method is by means of a modulation transfer function measurement, which determines the ability of a lens to resolve detail. A simple way to understand this is to imagine looking at a brick wall and then backing away from that wall un-til the individual bricks can no longer be distinguished.

A more obvious performance characteristic to be considered is distortion, which is most significant in wide-angle and fish-eye lenses. These types exhibit a barrel-type distortion, where the corners of the image are pulled in toward its center. Some IR camera companies compensate for distortion by means of the electronics and software used in a full imaging system. In fact, most commercially available digital image editing software can correct for distortion in a captured image.

If a wide-angle lens with little distortion is required in real time for an application, it may be necessary to have a lens custom-designed.

Another performance characteristic of importance is spot size. A typical focal plane array has tens of thousands of pixels, and the IR energy is imaged as a spot into each pixel. If a lens is diffraction-limited, the spot that it makes within each pixel exceeds the IR camera's ability to resolve the detail made by the lens.

A diffraction-limited lens is a rarity, though, because a lens will resolve differently when used with different focal plane array configurations. The spot size may also perform differently with the center pixel (on-axis) than with the corner pixel (off-axis). So a lens may exhibit diffraction-limited performance on-axis but fall short of this definition off-axis.
Although a diffraction-limited lens may seem like the perfect lens, a visual inspection is too subjective to distinguish this performance characteristic when comparing one lens with another. Hence, it is most important to know if the lens is compatible with the IR camera, so ask the lens manufacturer to verify compatibility or to recommend another solution.
Transmission

While the f number relates to image sharpness, the lens transmission can relate to image brightness. The transmission value given for an infrared camera lens is the level of energy that passes through the lens over the designed waveband. If a lens is said to have a transmission of 94 percent, this means that an average of 94 percent of the energy in the designed waveband entering the lens will exit the lens. The other 6 percent is either reflected or absorbed. The transmission is considered to be an average value because it will vary across the designed waveband.

Factors that contribute to transmission value are the optical materials, antireflection coatings, diffractive surfaces and the number of optical components in a multielement lens assembly. The details of these contributing factors are typically not available to the consumer, so the value provided must be weighted along with performance to determine if the lens will work for a specific application.

If a lens is designed to remain in focus over a wide temperature range, it is considered to be athermal. If an application requires the IR camera to be in an environment where the temperature fluctuates an appreciable amount, the image will need to be refocused. The extent of defocus will vary from lens to lens, and longer focal lengths are more likely to be affected by temperature change.

Athermalization is passive when achieved by mechanical means using different optical and mechanical materials with compensating thermal expansion coefficients. It is active when achieved by electrical means with the use of a sensor that enables a motorized focus movement with temperature change.

The last thing to consider is how to mount the lens onto the camera. Proper mounting is critical to ensure that the lens is in position for optimal performance. The three most common methods for interfacing the lens with the camera are the flange, threaded and bayonet styles (also referred to as twist lock). A flange-mounted lens is mated to the camera by matching a bolt-hole pattern on both the lens flange and the camera housing. It is held in place with screws. This is generally the method used if a tight seal is required, or if the lens is to be permanently affixed to the camera.

With the other two styles, various lenses can be interchanged. The threaded interface is a low-cost approach, and the bayonet style is the most convenient method, locking the lens into place with a simple twist and removing it by releasing a pin. The latter has a more elaborate configuration, however, which makes it a more costly option than the flange or threaded style.

In conclusion, many variants differentiate one IR camera from another, and the lens can be the main discriminator.