Pro Lite Technology

Tech Brief: Spectral Imaging

Spectral Imaging

A guide to spectral imaging technologies

A spectral imager is a hybrid camera that is one-part camera and one-part spectrometer. As with a digital camera, you obtain a 2D image, but the spectrometer channel adds a third dimension. The “image” that you record with a spectral imager is a 3D data-cube, where the third dimension contains spectral information. In other words, each pixel in the image contains the spectrum of light emitted or reflected from the object or light source.

Hyperspectral Cube (Photon etc)To understand spectral imaging, first consider how a colour camera works. A standard colour camera typically employs a Bayer pattern of red, green and blue bandpass filters applied directly onto the image sensor to form a colour image. Simplistically, you can regard one third of the sensor’s pixels as filtered green, one third blue and one third red. Software interpolation of the separate RGB pixels is used to recreate a full frame colour image. The spectral information contained in a colour image is limited to the amount of visible light which passes through the nominal 100nm bandpass RGB filters. In other words, this is not spectral data per se, rather an image which correlates to some extent with the colour of the object that we would perceive looking directly at it.

Within spectral imaging, there are a number of technologies used to collect spectral information. This article will review the attributes of those technologies, considering the mode of operation and their relative merits.

Recent advances in sensor design and processing speed have cleared the path for a wide range of applications employing pectral imaging, ranging from satellite based/airborne remote sensing and military target detection to industrial quality control and lab applications in medicine and biophysics. Due to the rich information contained in multispectral and hyperspectral images, they are uniquely well suited for automated image processing, whether it is for online industrial monitoring or for remote sensing.

Multispectral versus Hyperspectral Imaging

So what is a multispectral camera and what distinguishes it from a hyperspectral imager? There is no universally accepted definition, although the more spectral channels the camera features, the more likely it is for the vendor to market it as being hyperspectral rather than multispectral.

Consider a standard RGB colour camera. This represents a multispectral device in its simplest form, there being three spectral channels. Each of the three bandpass filters collects light in an approximate 100nm wide channel. As you add more filters, each with a narrower spectral bandpass, so you are able to discern more about the shape of the spectrum of light emitted or reflected from an object. In extreme, a true spectrometer can resolve spectral power distributions every nm or so, but the data contains no spatial component. The value of a spectral imager is that you obtain spectral information (of some kind) for all points on the light source or illuminated surface. So back to the exact definition of a multispectral instrument – let’s say that if you have a camera with at least 50 spectral channels, then you can legitimately describe it as being hyperspectral.

Filter-on-Chip Multispectral Imager (Pixelteq)Multispectral Imaging

Having defined a multispectral camera as having fewer than 50 spectral channels, let’s now turn our attention to how a multispectral imager works. Real-time multispectral imagers are based on 2D image sensors overlaid with an array of discrete, bandpass filters. There are two types of multispectral imager: those employing "filter-on-chip" technology, in which a series of narrow bandpass filters are applied directly to the 2D sensor array; and those employing an interchangeable filter array together with plenoptic imaging. A multispectral imager provides a more limited spectral dataset than a hyperspectral camera, but provides a distinct advantage in terms of measurement speed. A multispectral camera can generally be regarded as a "real time" spectral imager, allowing for the capture of rapidly changing objects.

►Filter-on-Chip or Mosaic Imagers

Bayer Filter Pattern (Courtesy of C Burnett, Wikipedia)Lorem ipsum dolor sit amet, consectetur adipiscing elit. Cras feugiat dolor a enim vehicula, in viverra nulla volutpat. Donec scelerisque semper nulla at ullamcorper. Sed maximus, nibh ut dignissim accumsan, massa enim cursus urna, eu maximus quam ante quis urna. Vestibulum ante ipsum primis in faucibus orci luctus et ultrices posuere cubilia Curae; Nunc in egestas lorem, sit amet sodales lectus. Sed iaculis orci in pulvinar tincidunt. Curabitur vitae mi suscipit diam condimentum suscipit.

►Plenoptic Imagers

Multispectral imagers based on the filter-on-chip technology sample in fixed spectral bands. They provide fast spectral imaging and tend to be offered at lower prices, but there is no easy way of adjusting the filters for specific applications. Multispectral imagers (such as the LightShift™ from Surface Optics Corporation) employ a 4 x 4 filter array in an interchangeable "tray" placed at the entrance to the camera, and a series of micro-lenses (so-called "plenoptic" imager) which provides certain advantages.

As with filter-on-chip cameras, plenoptic systems are fast (giving real-time, video-rate capture) but in addition they allow every pixel in the 2D array to see all (12-16) wavebands (spectral channels), and the tray can be swapped to allow the selection of application-specific filters or polarisers.

Hyperspectral Imaging

For each pixel in an image, a hyperspectral camera acquires the light intensity (radiance) for a large number (typically a high tens to several hundred) of contiguous spectral bands. Every pixel in the image thus contains a continuous spectrum (in radiance or reflectance) and can be used to characterise the objects in the scene with great precision and detail. Hyperspectral images obviously provide much more detailed information about the scene than a normal colour camera, which only acquires three different spectral channels corresponding to the visual primary colours red, green and blue. Hence, hyperspectral imaging leads to a vastly improved ability to classify the objects in the scene based on their spectral properties.

►Push-Broom Hyperspectral Imaging

In a “push-broom” hyperspectral imager, the camera images the scene line by line using the so-called "push-broom" scanning mode. One narrow spatial line in the scene is imaged at a time, and this line is split into its spectral components before reaching the sensor array. On the 2D sensor array, one dimension is used for spectral separation and the second dimension is used for imaging in one spatial direction (along a row of sensor pixels). The second spatial dimension in the scene arises from scanning the camera over the scene (for example, by means of aircraft movement). The result can be seen as one 2D image for each spectral channel, or alternatively every pixel in the image contains one full spectrum.

Briefly, the camera operates internally as follows: the camera fore-optic images the scene onto a slit which only passes light from a narrow line in the scene or object. After collimation, a dispersive element (in the case of NEO HySpex cameras, a transmission diffraction grating) separates the different wavelengths, and the light is then focused onto the 2D detector array. The net effect of the optics is that for each pixel interval along the line defined by the slit, a corresponding spectrum is projected on a column of detectors on the array. The data read out from the array thus contains a slice of a hyperspectral image, with spectral information in one direction and spatial (image) information in the other. By scanning over the scene, the hyperspectral camera collects slices from adjacent lines, forming a hyperspectral image or "cube", with two spatial dimensions and one spectral dimension. Note that the scanning is often intrinsic to the application: in remote sensing, scanning is provided by aircraft or satellite movement. Also, in many industrial quality control applications, products conveniently pass the sensor on their conveyor belt.

►Multi-Point Spectrometer (MPS) Hyperspectal Imagers

The FireflEYE hyperspectral camera from Cubert employs a novel multi-point spectrometer (MPS) technology, which strikes a fair balance between areal resolution and spectral resolution. The result is an imaging spectrometer with no need for scanning (unlike push-broom imagers) or image combination after fast filter shifts. The technology provides clean hyperspectral images out of the box without any moving artefacts. [awaiting further content]

Spectral Range

Hyperspectral and multispectral cameras are capable of acquiring data well beyond the spectral range of the human eye, which is limited to a maximum wavelength of 780nm. Depending on the type of 2D image sensor deployed, spectral imagers can be configured for imaging from 400nm in the visible blue part of the spectrum, out to 2500nm in the near infrared, For many applications, the reflection or absorption properties in the IR region are essential to characterise, quantify or classify the objects in a scene. Sensors based on silicon (e.g. CMOS or CCD) typically provide access to the 350-1100nm (VIS-VNIR) band. Sensors based on InGaAs (Indium Gallium Arsenide) or MCT (Mercury Cadmium Telluride) instead provide access to the 900-2500nm (VNIR-SWIR) band. The absolute sensitivity and dynamic range of 2D image sensors is improved by cooling the array, which can be achieved by a variety of means, such as a Peltier (thermoelectric) cooler.

[14/18: this page is under construction]

Related Products
Case Studies
Knowledge Base
Have a Question?

Why not use our convenient enquiry form?

> Click Here

Buy Online

This product can be ordered via the Pro-Lite on-line shop.

> Click Here

Need a measurement?

Contract measurement services are available from Pro-Lite Photometrics using this product

> Click Here