Small housings and powerful sensors didn’t used to appear together. But thanks to ever more compact electronics, and good engineering, one can have both. IDS has expanded its portfolio by integrating the Sony Pregius IMX900 sensor into selected models of the compact USB3 uEye XCP, XLS and GigEuEye LE series.
uEye low cost cameras with Sony Pregius IMX900 sensor – Courtesy IDS
The photo above provides an at-a-glance overview of the various board-level, housing, and lens-mount options, each of which is classified as low cost and very compact.
If you prefer a tabular view like the following, with clickable links to specs and the option to request a quote, go to 1stVision’s cameras page, and scroll down to the selector page with drop-down filters, at https://www.1stvision.com/cameras/industrialCameras
The image below is an annotated screenshot, where the manufacturer (IDS) and the sensor (Sony IMX900) bring you to the 10 distinct camera models utilizing this sensor.
Sony’s Pregius sensor evolution
Sony is an industry leader with its Pregious sensor series, on which we’ve written previously with an overview. The Sony IMX900 is in the Pregious S 4th generation, extraordinarily sensitive thanks to Sony innovations including the back-illuminated stacked architecture.
Back-illumination improves sensor quantum efficiency – Courtesy IDS and Sony
VIS plus NIR sensitivity
We won’t bother showing VIS images in monochrome or color but of course they look great. But calling out a special capability of the Sony IMX900 – it’s also very sensitive to near infrared (NIR). Besides being part of the Pregious group, the IMX900 utilized STARVIS 2 technology, yielding NIR performance.
Yet another Sony IMX900 feature: Quad HDR
While high dynamic range (HDR) features aren’t new per se, Quad HDR on the Sony IMX900 takes HDR to another level. Getting the dark sections sufficiently saturated while not oversaturating the brighter regions is really evident with Quad HDR below.
Quad HDR generates a balanced image – Courtesy IDS
The feature list goes on and on
Call us at 978-474-4000 to learn more about this remarkable sensor and the range of IDS uEye cameras into which the sensor has been designed.
About you: We want to hear from you! We’ve built our brand on our know-how and like to educate the marketplace on imaging technology topics… What would you like to hear about?… Drop a line to info@1stvision.com with what topics you’d like to know more about.
While we humans can only see what we’ve named to be visible light, bees can see UV light! Some camera sensors register IR wavelengths! Some cameras can sense both visible light and on through NIR and SWIR.
In this piece we focus on applications that benefit from combined VIS-SWIR solutions, from 400 nm through 2.5 nm.
Deconstructing the electromagnetic spectrum into it’s commonly known constituent regions
Example applications
Just to whet the appetite, consider the 4 sets of image pairs below. In each case, the leftmost image was captured with visible wavelengths, while the righthand image utilized SWIR portions of the spectrum. These pairs were chosen to highlight the compelling power of SWIR to identify features that are not apparent in the visible portion of the spectrum.
VIS-SWIR image pairs – Courtesy Allied Vision – a TKH company
For certain applications, one wouldn’t need the human-visible images, of course, as with machine vision the whole point is to automate the image processing and corresponding actions. So for counterfeit banknote detection, bottle fill level monitoring, materials identification, or crop monitoring, one might just design for the SWIR portion of the spectrum and ignore the VIS.
Vein imaging application overlays SWIR image of veins into visible image of patient forearm –Image courtesy TAMRON
But some applications might benefit from both the VIS and the SWIR images. For example, the vein imaging application might require a VIS reference image as well as a SWIR-specific image, for patient education and/or medical records.
Monitor moisture levels in crops from airborne drone – Image courtesy TAMRON
For the crop monitoring application above, the VIS spectrum clearly orients trees, hills, buildings, and roadways. Meanwhile pseudo-color-mapping shows the varied moisture levels as sensed in the SWIR portion of the spectrum.
The range of potential applications combining VIS and SWIR is staggering. One can improved on one’s own or a competitor’s previous application. Or innovate something altogether new.
Sensors that register both VIS and SWIR wavelengths
Sony’s IMX992 and IMX993 sensors utilize Sony’s SenSWIR technology, such that a single sensor and camera may be deployed across the combined VIS and SWIR portions of the spectrum. Without such sensors, a VIS SWIR solution would require at least two separate cameras – one each for VIS and SWIR, respectively. That would add unnecessary expense, takes up more space, and require camera and image synchronization.
Now there are cameras, such as several in Allied Vision’s Alvium series, in which Sony’s SenSWIR sensors are embedded. With several interface options, including mipi, USB3 Vision, and 5GigE Vision:
Mipi, USB3 Vision, and 5GigE Vision interface options – Courtesy Allied Vision – a TKH Company
Lens manufacturers doing their part
One of the beauties of the free-market system, together with agreements on standards for interfaces and lens mounts, is that each innovator and manufacturer can focus on what he does best. Sensor manufacturers bring out new sensors. Camera designers embed those sensors and provide programming controls, communications interfaces, and lens mounts. And optics professionals design and produce lenses. The benefits from a range of choices, performance options, and price points.
Navitar VIS-SWIR lenses
Navitar’s ZOOM 7000-2 macro lens imaging system delivers superb optical performance and image quality for visible and SWIR imaging. Their robust design ensures reliability even in harsh environments. ZOOM 7000-2 macro lenses are ideal for applications, such as machine vision, scientific and medical imaging applications.
ZOOM 7000-2 VIS-SWIR lens – Courtesy Navitar
In fact there are three models in the series:
Each model has its application – but only the middle one is designed explicitly for VIS-SWIR – Courtesy Navitar
Kowa FC24M multispectral lenses
Kowa’s FC24M C-mount lens series are manufactured with wide-band multi-coating. That minimizes flare and ghosting from VIS through NIR. These lenses are also compelling for a number of other reasons, including wide working range (as close as 15 cm MOD), durable construction, and a unique close distance aberration compensation mechanism.
FC24M C-mount lens series – Courtesy Kowa
That “floating feature” creates stable optical performance at various working distances. Internal lens groups move independently of each other, which optimizes alignment compared to traditional lens design.
Tamron Wide-band SWIR lenses
Other lensing options include Tamron’s Wide-band SWIR lenses. While the name says SWIR, in fact they are VIS-SWIR. Designed for compatibility with Sony’s IMX990 and IMX991 SenSWIR sensors, you have even more lens choices. Call us at 978-474-0044 if you’d like us to help you navigate to best-fit components in cameras, lensing, and lighting, for your particular application.
About you: We want to hear from you! We’ve built our brand on our know-how and like to educate the marketplace on imaging technology topics… What would you like to hear about?… Drop a line to info@1stvision.com with what topics you’d like to know more about.
Gidel is a leading provider of high-performance FPGA-based imaging and vision solutions. Their product offerings, engineered for data intensive applications that demand real-time processing and minimal latency, include edge computers powered a Nvidia Jetson™ embedded computer, FPGA-based frame grabbers, recording & streaming systems, and a camera simulator for developing and testing imaging and vision applications. These solutions are available as out-of-the-box, open infrastructure, or fully tailored to your specific application requirements.
Products and features to help with High Dynamic Range
Below we show and describe useful features for applications where High Dynamic Range is needed, but typically are plagued by processing time and image degradation. We explain how Gidel can provide HDR in real time, compress and correct the image for a great image.
For High Dynamic Range (HDR) context, download our HDR whitepaper for an optional review. And/or let these images motivate the topic:
Courtesy Gidel
The aerial imaging application is just an example – the principle is widely applicable. They key context for HDR is that many imaging scenes are high-contrast, with deeply nuanced darker regions and equally nuanced brighter regions.
If one controlled for a single-exposure duration across all regions, whether by fixed timing or average pixel saturation, the one-exposure-fits-all image is likely to be a poor or unusable compromise – like the “original image” shown above left. One exposure can’t allow you to see the darks and brights. Either your exposure set to see the dark will saturate the brights… or if you optimize exposure to see the nuanced brights it will make the darks so dark you can’t distinguish them.
At the risk of sounding like an advertisement for laundry detergent about whiter whites and preserving colors, one doesn’t have to dwell on how HDR is achieved to recognize the HDR image shown above right is an improvement over the non-HDR original.
There are other helpful tools besides just HDR
HDR is an often powerful technique to effectively expand the dynamic range of the delivered image. But HDR isn’t always needed, and isn’t always the best tool, whether alone or in combination. Consider also tools and techniques like compression, gamma correction, and white balance – a bit more on each of these below – advantages Gidel can offer.
Compression
Another sometimes-useful technique is compression, reducing transmission volume (and time). Ideally one seeks lossless or low-loss compression, but that’s a tradeoff typically determined by assessing performance outcomes against any rigor or risk requirements for outlying cases. Often “good enough” and demonstrably reliable is the preferred choice.
If your application is NOT challenged by data transfer volume and time… whilst sending uncompressed images, then you don’t need compression. Conversely, might compression speed up your application for a competitive advantage?
Or thinking outside the box, might there be innovative applications doable with compression that aren’t possible without it? Let us help you do the math. Or try some applied testing.
Yet another digital imaging tool is gamma correction, an operation on luminance values. It was originally introduced to adapt digital images and map them better to human visual perception, due to the ways our eyes perceive brightness. BUT it also helps machine vision algorithms perform better by linearizing color and brightness data. That can improve accuracy in object tracking and color segmentation, for example.
Likewise for white balance. Here too the origin of white balance was to create a more pleasing image for human viewers of digitally rendered images, so that “white” objects appear white even if the light source skewed a bit blue or yellow, for example.
It turns out that white balance is also important for machine vision. By white balancing to the light source (if sunlight, it changes throughout the day), whether natural or artificial, color segmentation and mapping is improved. Whether doing inspection, object identification, or medical imaging, accurate color mapping is essential.
With HDR, Compression, white balance and Gamma correction
Throwing the whole kitchen sink at it, consider the following image pairs:
Right-hand image benefits from HDR, white-balance, gamma correction, and compression – Courtesy Gidel
Do it in real-time, on board the frame grabber from Gidel
The above illustrations are pretty compelling. But imagine if that had to happen on your host PC, operating on large raw data images, while trying to keep up with incoming data from the camera that might overwhelm the PC’s ability to achieve the desired machine vision decisions. That’s where Gidel’s FPGA frame grabbers excel. They receive the raw image stream from the camera, and do real-time value-added pre-processing – very fast – before passing the improved (and optionally compressed) image to the host software. So the workload on the host is reduced.
Better images – faster
With many engineering design challenges, it can be difficult or impossible to simultaneously satisfy “qualitatively better” as well as “faster performance” criteria. Indeed some conceptualizable applications may not be doable in practice due to such constraints. Or may not be affordable with common interfaces.
With Gidel’s FPGA frame grabbers, you get improved speed and high-quality actionable images – for high-performance applications that deliver.
About you: We want to hear from you! We’ve built our brand on our know-how and like to educate the marketplace on imaging technology topics… What would you like to hear about?… Drop a line to info@1stvision.com with what topics you’d like to know more about.
Recently we announced Allied Vision’s 5GigE SWIR Goldeye Pro cameras. Pretty cool. Pun intended. Click that link above for the 10,000 foot view of the camera series, features, and value proposition overall.
SWIR cameras come in both cooled and uncooled models
So you are already doing Short-Wave-InfraRed imaging, or think you might want to. For all the reasons and applications discuss in our SWIR cameras and applications knowledge-base article.
Cooled vs uncooled performance – what are the differences?
As with many engineering design choices and product selection options, for a given application one needs components that are good enough – perhaps with a little margin – to get the job done. But not overdesigned – as that would add cost, weight, and volume delivering no measurable benefit.
5GigE SWIR Goldeye Pro – Courtesy Allied Vision – a TKH Company
Thermoelectric cooling (TEC)
The InGaAs (indium gallium arsenide) sensors used for SWIR imaging deliver the best images when temperature-stabilized. That’s provided by the thermoelectric cooling (TEC). That helps reduce dark noise and thermal current.
With TEC, see key performance metrics
For the Goldeye Pro G5-320, we snapshot key metrics from the datasheet:
Goldeye Pro G5-320 imaging metrics – Courtesy Allied Vision – a TKH Company
Compare to uncooled Goldeye G-033 metrics:
Uncooled Goldeye G-033 imaging metrics – Courtesy Allied Vision – a TKH Company
Sensor performance comparison summary
For background on sensor performance testing, review our tutorial on EMVA 1288 attributes and standards. There we go into the meaning of key terms like dark noise, dark current, saturation capacity, and dynamic range.
While the G-033 and G5-320 sensors above are different in size and release date, they are both InGaAs sensors, so share essential basic characteristics. And we can be sure that Allied Vision engineered each camera for the best possible performance relative to housing design, electronics positioning, and so forth.
Camera
G5-320
G-033
Temporal dark noise
183 e~
390 e~
Dark Current
8.9 ke~/s
430 ke~/s
Dark current and temporal dark noise mitigated by cooling an InGaAs sensor
What performance requirements does your SWIR application require?
We love to learn about client applications, and to guide you to best-fit selections for sensors, cameras, lenses, lighting, and software. Whether you are new to SWIR or experienced, let us help!
About you: We want to hear from you! We’ve built our brand on our know-how and like to educate the marketplace on imaging technology topics… What would you like to hear about?… Drop a line to info@1stvision.com with what topics you’d like to know more about.
