1stVision Inc. - Machine Vision Articles - Page 3 of 56

April 14, 2025April 14, 2025

Six new JAI line scan cameras across two product families

JAI has over 60 years experience in the vision industry, and is likewise a longstanding part of the 1stVision portfolio. JAI produces a diverse range of cameras, including area scan and line scan; across multiple interfaces; and spanning visible, NIR, SWIR, and UV portions of the spectrum.

In this article we announce six new line scan camera models: 4 in the Sweep series at 4k each; and 2 in the Sweep+ series at 4k and 8k respectively. There’s a niche for every model.

Two new Sweep+ line scan cameras – one 4k one 8k – Courtesy JAI Imaging

Four new 4k line scan Sweep models; Two new Sweep+ models incl. 8k

At the risk of information overload, we’ve bundled this product release update of six new camera models across two product families into a single article. We hope the reader appreciates the compare-and-contrast approach as a way to understand the value proposition for each camera/family.

When you need line scan it’s nice to have options

While area scan sensors with big 2D pixel array outputs are right for some applications, line scan is a better fit for others. In particular, if the object being imaged is essentially continuous, line scan can be an effective way to take narrow image slices with a lower cost 1D sensor that’s long in just the dimension that maps to the object width. Movement of the object (or sensor) together with successive slices then creates the second dimension.

The 1stVision Knowledge Base offers illustrated articles on line scan topics, including How to Calculate Line Rate Based on Conveyor Speed and Polarization Techniques with Line Scan Cameras. Now on to the new camera models!

Example line scan applications – Courtesy JAI Imaging

JAI’s Sweep line scan cameras vs. Sweep+

JAI’s Sweep line scan cameras include a wide offering of monochrome, bilinear, and trilinear models, at 2K, 4K, 8K, and 16K. Interface options include 5GigE, 10GigE and CoaXPress. If you need color sensing, the bilinear approach is an affordable approach, and often plenty good enough. For true color capture, the trilinear sensors are superior – if needed.

Sweep+ cameras, as suggested by the + suffix, are JAI’s top of the line color and multispectral line scan products. The beam-splitter prism optics, together with three (color only models) or four (color plus NIR) sensors, deliver high-precision color resolution. Due to the data volumes generated, the Sweep+ cameras use CoaXPress, 10GigE, and CameraLink interfaces to achieve line rates suitable to the respective sensor sizes and anticipated applications.

Click to contact — Give us some brief idea of your application and we will contact you to discuss camera options.

Mini-tutorial on single sensor vs. multi-sensor color imagin

While Bayer filters or equivalent provide single sensor color, that’s good enough for some applications, there is some loss of color resolution via that inherent physical interpolation. So demanding applications take advantage of separate sensors for each segment of the spectrum, using a prism to separate wavelengths. The resulting camera is physically larger to accommodate the the prism, extra sensors, and electronics – but the best possible color performance is achieved.

A deeper dive on the new Sweep cameras

At-a-glance overview of the Sweep 4k cameras – Courtesy JAI Imaging

Monochrome line rates are 2x that of the bilinear color models

… because there is a single monochrome pixel mapped to the target object’s corresponding area.

But if you need color the bilinear approach is an efficient way to do it, since there are only two rows of data transmitted for every real-space row imaged….

A peak at the bilinear approach to color line scan imaging

Other JAI Sweep camera features

Keystone correction, aka tilted view correction, onboard the camera

Trigger-width dynamic exposure control

Resume-scanning control for stop-start motion

Trigger delay

… and more

Drill in to JAI Sweep camera models and datasheets. Or let us guide you.

More details on JAI’s new Sweep+ line scan cameras:

A 4k and and 8k model, both with CoaXPress interfaces – Courtesy JAI Imaging

The two new models round out the Sweep+ lineup – Courtesy JAI Imaging

Precise prism optical design yields superior images – Courtesy JAI Imaging

Some of the key benefits of CoaXPress – Courtesy JAI Imaging

And yet more features…

You get the idea… Sweep+ is a remarkable camera family. There are even more features to find in the Sweep+ product details and datasheets. Or just call us at 878-571-5683 for expert assistance. Tell us about your application requirements and we’ll guide you to an optimal camera. Lighting, lenses, filters, cables, and software too.

1st Vision’s sales engineers have over 100 years of combined experience to assist in your camera and components selection. With a large portfolio of cameras, lenses, cables, NIC cards and industrial computers, we can provide a full vision solution!

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.

March 26, 2025March 26, 2025

IDS Imaging μEye XCP-E Event-based cameras

The μEye XCP-E event-based camera utilizes Sony’s Prophesee IMX636 sensor. So, by design, it captures only relevant image changes. Event-based imaging can be a game changer for certain applications. Unlike area scan or line scan imaging – which capture every pixel and render a “full image” – event imaging only senses and delivers changes.

Event-based imaging captures the changes:

Left: uEye XCP-E image vs. Right: Area scan image – Courtesy IDS Imaging

Less is more

Playing on the “less is more” adage reveals key insights into event-based imaging.

The human eye is adept at delivering and entire scene, of course, and that’s how most of us imagine we see the world around us. But our overall vision perception also builds upon the eye’s ability to sense brightness changes within small segments of the overall scene..

Consider a baseball batter awaiting a pitched ball. The overall scene is relatively static: the outfield fence, bases, and foul lines aren’t moving. And the infielders are almost static – relative to the motion of the ball. But the pitched ball approaching at 80 – 90 miles per hour can be identified by a good batter, to gauge “strike or ball” and “swing or take”.

The batter’s visual processing does NOT have time to capture the full scene at each instant of “ball release”, “just released”, “mid-way”, and “arriving soon”. Rather, the ball’s trajectory is discerned as successive changes against a static background. So too with an event-based camera.

Less data -> More speed: In other circumstances, less data might seem like a handicap. For area scan applications it often would be. Finding defects on a static surface requires ingesting a lot of detail – all the pixels – in order to do edge detection, blob analysis, or other algorithmic processing. But by detecting “just the brightness changes”, transmitting less data is exactly what delivers the increased speed!

Applications example: motion detection and analysis

Airport security application – Courtesy IDS Imaging

What is delivered are pixel motion coordinates and timestamps – NOT pixel brightness values. So you get useable results rather than having to algorithmically compute the results from a traditional area scan image. Track moving objects easily.

How much?

Already intrigued? The housed model, UE-39B0XCP-E, is available now, as this blog releases in early March 2025. Board-level models to be released soon.

Temporal resolution better than 100 μsec

Detect rapid changes – a conventional camera would need > 10,000 fps to capture this – Courtesy IDS Imaging

Efficient data processing

Courtesy IDS Imaging

IDS Imaging uEye XCP-E event-based cameras can be directly integrated with the sensor manufacturers’ software tools, called Metavision. That’s all thanks to Sony’s partnership with Prophesee. Since event-based imaging is a paradigm shift away from conventional machine vision approaches, the visualization tools, API, and training videos help you get up to speed quickly.

March 15, 2025March 26, 2025

IDS Imaging Sony IMX662 cameras with STARVIS sensors

We’ve previously written about Sony STARVIS sensors, and all that still holds true, of course. If you don’t feel like chasing that last link, a two-word summary would be “high sensitivity”. But this is our first piece on the Sony IMX662 STARVIS sensor in particular. And the corresponding 10 camera models into which IDS Imaging has embedded this remarkable new sensor.

Sony IMX662 sensor – what’s so special?

Before reviewing STARVIS in general, let’s cut to the chase on the IMX662. Three specifics jump out.

1. Wider dynamic range:

Dynamic range characterizes the expressive power of the sensor. It’s the ratio between the smallest and largest values the sensor can capture. Per the side by side images below, the more performant sensor (in this case the IMX662 of course), gets the saturated segments bright, the darker segments dark, and a lot more nuance in the middle. Which translates into actionable imaging data for your machine vision algorithms.

2. No chromatic aberration in HDR mode:

Chromatic aberration is the introduction of color artifacts not present in the original scene, due to the physics of light passing through a lens. Note that the “lens” might be the user-added camera lens, or the micro-lens inherent to every pixel on the sensor. Either way, it’s an undesirable phenomenon, since if your application uses color, it can be a source of “confusion” in your image processing.

So it’s a nice benefit that Sony’s “Clear HDR” feature overcomes chromatic aberration when the IMX662 is used in HDR mode. As shown below:

3. Low cost

The Sony IMX662 sensor is very attractively priced. Since it doesn’t cost camera manufacturers very much to buy the sensors wholesale, they can design them into their value-added cameras, and price the overall package attractively for you, the customer. Whether in small volumes or large.

Think “go where no camera has ever gone before.” Or if you already have an application with another sensor, consider a Gen 2 application with a higher return on investment.

Why 10 different camera models for just this one sensor?

The Sony IMX662 sensor is so compelling that IDS Imaging designed it into 10 different camera packages, providing form factors for diverse customer requirements. Per the snapshot below from 1stVision’s camera selector, with sensor dropdown Sony IMX662 selected, we see all 10 models. In fact it’s 20 models, as each is available in a monochrome or color version.

The top 6 rows are GigE models, with framerates to 59fps. The 4 bottom rows utilize the USB3 interface, delivering up to 93fps.

1stVision carries all 10 IDS Imaging camera models using the Sony IMX662 sensor

GigE Vision models:

The GV prefix in the model name denotes the GigE Vision interface. M/C indicates both Monochrome and Color offerings. MB stands for MotherBoard, which is an especially small form factor, for applications with tight and/or angled spaces. For MB variants, and optional daughterboard and flex ribbon cable, if desired.

Top row: GigE no mount; GigE C-mount; GigE motherboard;
Bottom row: GigE MB C-mount; GigE MB S-mount; GigE S-mount

USB3 Vision models:

The models with the U3 prefix are offer board level and housed models, similarly ideal for tight spaces and embedded applications. And with frame rates to 93fps:

Top left: BL no lens mount; Top right: BL S-mount;
Bottom left: BL C-mount; Bottom right: Housed C-mount

Since both interface options, GigE Vision and USB3 Vision, are industry standards, you can use “IDS peak” SDK, or any other standards compliant software you like.

Sony STARVIS technology

Underlying the Sony IMX662 – and indeed all the Sony STARVIS sensors – is the innovative back-illuminated structure. This means more photons get into the pixel well, greatly enhancing low light performance.

Choose the sensor – and camera – that’s right for the application

Is the Sony IMX662 right for your application? For all cameras with the IMX662, go to our camera selector and choose IMX662 in the sensor pulldown. Or other members of the Sony STARVIS sensor collection? While lensing and lighting are also important, choosing the right sensor is at the heart of your application solution. We’re always happy to advise.

March 14, 2025March 26, 2025

Considerations in using an F-mount vs. a threaded mount

For machine vision purposes, we generally advise against using F-mount. We don’t have a dog in the hunt for leisure or artistic photography, but for reliability in many machine vision applications there are shortcomings with F-mounts.

EMERALD series F-mount lens – Courtesy Schneider Optics

Even our premier partner Schneider Optics, produces very fine F-mount lenses – the optics are first class! But just as a reliable auto dealership would counsel against buying a compact sedan to tow a heavy boat, we would hope to engage you in dialogue about your imaging application – and help you choose the best lens.

There are many lens mounts available

The wide range of lens mounts is, generally speaking, a very good thing. It permits camera designers and manufacturers to specialize in what they are good at; and likewise for lens designers and their own production facilities. The former is electronics while the latter is about optics.

By defining standards for lens mounts, we all benefit from market competition, which drives quality, feature, and price differentiation.

Some mounts are for large-format sensors. Some for small sensors. Some for “quick change”. Some for stability. Some can fit several niches. Some can in theory but shouldn’t be pushed in practice. Below we provide examples.

Lens families from a single manufacturer by sensor size and pixel size – Courtesy Schneider Optics

What should we consider with F-mounts?

If you are a sports photographer, F-mount may be an excellent choice. The bayonet mount design is ideal for quick lens changes, and the “money shot” may require a quick lens swap. If shooting at really short exposure times, inherent instability in the mount design may not have an impact on image quality.

While the F-mount name traces to the famous Nikon F camera series, one might also think of it as for “Fotographie” – the German work for photography – if you like mnemonic hints.

But for machine vision applications, one typically mates the lens to the camera “forever”, so quick-change benefits are of little value. And compared to screw mounts, for example, it’s clear that the bayonet mount, with just two securing points, might suffer in a high-vibration environment. Whether on the factory floor, a traffic camera gantry, or a moving vehicle, the lens and camera should act as one. Stability is key.

For the record, F-mount defines a standard Flange Distance of 46.5mm from the mounting surface of the camera to the sensor, as well as standardized mechanics on the size and position of the male pins on the lens and the female receptors on the camera body. For those cameras or lenses with electrical contacts as well, there are variances. But mechanically and optically and F-mount is an F-mount.

What about the M42 lens mount?

While the M42 lens mount came into existence for “photography”, when computers and then machine vision emerged, M42 has become one of many popular lens mounts. The threaded screw mount design provides a very secure connection that binds along several turns of the screw – so it’s very resistant to vibration.

But while the conceptual design is great, M42 isn’t really a “full standard” in two important ways…

M42 doesn’t define flange distance

Every lens must precisely focus light rays onto the sensor surface. Camera makers position sensors a precise distance from the shoulder of the mount, a distance referred to as the flange distance, or flange focal distance. Curiously, there isn’t universal agreement on what that distance should be – though there is some convergence on a couple of de facto conventions.

The M42 lenses may need to have an extension tube added so that they can image on a camera sensor. Thankfully an extension tube is an inexpensive but effective accessory – but it’s important to note focal distances and source components accordingly.

M42 doesn’t define thread pitch

The second caveat with M42 mounts is that there are at least two common thread pitches offered: both 0.75 and 1.0.

Annotated snapshot of 1stVision lens selector tool.

If we use the 1stVision lens-selector, the Lens Mount dropdown shows both M42x0.75 as well as M42x1 options (as well as many other lens mounts). So if you know the camera mount specifications, you can find lenses that will thread correctly – but one needs to pay attention!

All the mounts we offer

To the left is a snapshot from the 1stVision lens-selector, showing the dropdown for all lens mounts we offer. If one chooses “All” it yields some 200 lenses across more than 10 manufacturers.

By specifying the lens mount, for the camera you are considering, one need only select a couple of additional dropdowns like sensor format, and focal length, to home in on candidate lenses for your application.

Each mount has its role. While a camera designer sometimes has options relative to sensor size, intended market, and price : performance decisions, there are good reasons why each mount type exists. And while some cameras are available with two or more mount options, generally speaking the camera you choose will dictate the mount – and hence the lens options.

Don’t fixate on the mount

While we’ve discussed certain aspect of lens mounts in this blog, you can generally trust that camera designers have chosen a suitable one for any particular camera – after all they want to maximize the number of cameras they can sell. There are many other considerations in machine vision lens selection. You can read all about it, or just…

1stVision Inc. – Machine Vision Articles

Posts