Category: Lighting

Machine vision – indeed human vision too – relies heavily on contrast. The feature(s) being identified need to stand out against any competing candidate features. Otherwise confusion is present. Which means vision either doesn’t work, or isn’t efficient.

While it’s easy to achieve good contrast in certain machine vision applications, sometimes we’re presented with the special challenges of glare. The target object is the same, in the two images below. If the goal is to read the alphanumeric information, which image would you rather pass to your machine vision software?

Techniques to eliminate glare

Glare can arise due to highly reflective surfaces, especially in combination with the direction of the light source relative to the lens and sensor capturing the image. Thankfully there are a number of techniques to eliminate or substantially reduce glare.

Off-axis lighting

For the 1-D bar code reading application illustrated below, moving the light to an off-axis position creates dark field orientation, eliminating the glare. Identical materials but different geometry does the trick!

Diverse lights and geometries

Consider another example. In this case we have a titration tray, with multiple wells. Each well has a laser-etched 2-D code.

• (a) the titration tray at low resolution, marked up with red outline around 6 wells isolated in high resolution images (b) – (f)
• (b) High-angle ring light
• (c) Coaxial light
• (d) Dark field ring light
• (e) Diffuse dome light
• (f) Flat diffuse light

In this scenario, both (d) the dark field ring light, and (f) the flat diffuse light, are far superior to the other options, and the flat diffuse light is the winner.

NOTE: The example above is NOT meant to suggest that flat diffuse light is always the winner. It’s important to understand the characteristics of the surface you are inspecting, and each candidate type of lighting, and the geometric options. And to test!

Guideline

The following diagram is a general guideline, based upon the two most prevalent surface characteristics: (1) Surface flatness and texture, and (2) Surface reflectivity profile

Send your samples

Lighting is a complex topic, so we’re happy to help. If you are uncertain which light type to choose and how to configure, we may be able to do some testing for you. Contact us to arrange sending samples to test in our lab, and we can recommend sensor, lensing, lighting, and configuration options.

For comprehensive coverage on glare reduction – and more – download A Practical Guide to Machine Vision lighting from our knowledge base.

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.

#glare

#contrast

#machinevisionlighting

#diffuser

#polariser

While the endless parade of new CMOS sensors get plenty of attention, each bringing new efficiency or features, lighting and lensing are too often overlooked. The classic three-legged stool metaphor is an apt reminder that each of sensor, lighting, and lensing are critical to achieving optimal outcomes.

Lighting matters

If you haven’t investigated the importance of lighting, or want a refresher, see our Knowledge Base resources on lighting. In those illustrated articles, we review the importance of contrast for machine vision, and how lighting is so critical. By choosing the best type of light, the optimal wavelength, and the right orientation, the difference in outcomes can be remarkable. In fact, sometimes with the right lighting design one can utilize less expensive sensors and lenses, achieving great results by letting the lighting do the work.

Pictures worth a thousand words

Before digging into product details on CCS LFXV flat dome lights, let’s take a look at examples achieved without… and with… the selected models.

Consider an example from electronics parts identification:

Here’s an example reading 2-D codes from contact lens packages:

Consider identifying foreign materials in food products, for either automated removal or quality control logging:

More about wavelength

In the images above, you may have noticed various wavelengths were used – with better or worse outcomes. Above we showed “just” white light, red light, and infrared, but blue, green, and UV are also candidates, not to mention SWIR and LWIR. Light wavelength choice affects contrast – not just when using dome lights – see wavelengths overview in our knowledge base.

Key concepts

By way of contrast, let’s first look at the way a traditional dome light works:

Notice the camera is mounted to the top of a traditional dome light. The reflective diffusion panel coats all the inside surfaces of the dome – except where the camera is mounted. The diffusion pattern created is pretty good in general – but not perfect at hiding the camera hole entirely. If the target object is highly reflective and tends towards flat, one gets a dark spot in the center of the image…. and the application underperforms the surface inspection one hoped to achieve.

So who needs newfangled flat dome lights?

There’s nothing wrong with conventional dome lights per se, if you’ve got the space for them, and they do the job.

Three downsides to traditional dome lights

1. A traditional dome light may leave a dark spot – if the target is flat and highly reflective

2. A traditional dome light takes up a lot of space

Notice how much space the conventional dome light takes up, compared to a “see through” LED flat dome light. But space-savings aren’t the only benefit to flat dome lights….

3. Working distance is “fixed” by a traditional dome light

Most imaging professionals know all about camera working distance (WD) and how to set up the optics for the camera sensor, a matching lens, and the object to be imaged, to get the optical geometry right.

Now let’s take a look at light working distance (LWD). Consider the following can-top inspection scenarios:

Wondering how to light your application?

Send us your sample(s)! If you can ship it, we can set up lighting in our labs to do the work for you.

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.

Black and white vs. color sensor? Monochrome or polychrome light frequencies? Visible or non-visible frequencies? Machine vision systems builders have a lot of choices – and options!

Let’s suppose you are working in the visible spectrum. You recall the rule of thumb to favor monochrome over color sensors when doing measurement applications – for same sized sensors.

So you’ve got a monochrome sensor that’s responsive in the range 380 – 700 nm. You put a suitable lens on your camera matched to the resolution requirements and figure “How easy, I can just use white light!”. You might have sufficient ambient light. Or you need supplemental LED lighting and choose white, since your target and sensor appear fine in white light – why overthink it? – you think.

Think again – monochrome may be better

Polychromatic (white) light is comprised of all the colors of the ROYGBIV visible spectrum – red, orange, yellow, green, blue, indigo, and violet – including all the hues within each of those segments of the visible spectrum. We humans perceive it as simple white light, but glass lenses and CMOS sensor pixels see things a bit differently.

Chromatic aberration is not your friend

Unless you are building prisms intended to separate white light into its constituent color groups, you’d prefer a lens that performs “perfectly” to focus light from the image onto the sensor, without introducing any loss or distortion.

Lens performance in all its aspects is a worthwhile topic in its own right, but for purposes of this short article, let’s discuss chromatic aberration. The key point is that when light passes through a lens, it refracts (bends) differently in correlation with the wavelength. For “coarse” applications it may not be noticeable; but trace amounts of arsenic in one’s coffee might go unnoticed too – inquiring minds want to understand when it starts to matter.

Take a look at the following two-part illustration and subsequent remarks.

In the illustrations above:

• C denotes red light at 656 nm
• d denotes yellow light at 587 nm
• F denotes blue light at 486 nm

Figure 1, showing transverse chromatic aberration, shows us that differing refraction patterns by wavelength shift the focal point(s). If a given point on your imaged object reflect or emits light in two more more of the wavelengths, the focal point of one might land in a different sensor pixel than the other, creating blur and confusion on how to resolve the point. One wants the optical system to honor the real world geometry as closely as possible – we don’t want a scatter plot generated if a single point could be attained.

Figure 2 shows longitudinal chromatic aberration, which is another way of telling the same story. The minimum blur spot is the span between whatever outermost rays correspond to wavelengths occurring in a given imaging instance.

We could go deeper, beyond single lenses to compound lenses; dig into advanced optics and how lens designers try to mitigate for chromatic aberration (since some users indeed want or need polychromatic light). But that’s for another day. The point here is that chromatic aberration exists, and it’s best avoided if one can.

So what’s the solution?

The good news is that a very easy way to completely overcome chromatic aberration is to use a single monochromatic wavelength! If your target object reflects or emits a given wavelength, to which your sensor is responsive, the lens will refract the light from a given point very precisely, with no wavelength-induced shifts.

Making it real

The illustration below shows that certain materials reflect certain wavelengths. Utilize such known properties to generate contrast essential for machine vision applications.

In the illustration we see that blue light reflects well from silver (Ag) but not from copper (Cu) nor gold (Ag). Whereas red light reflects well from all three elements. The moral of the story is to use a wavelength that’s matched to what your application is looking for.

Takeaway – in a nutshell

Per the carpenter’s guidance to “measure twice – cut once”, approach each new application thoughtfully to optimize outcomes:

• Is the application best suited to visible or non-visible portions of the spectrum?
• If visible, is it best solved with a color sensor or a monochrome sensor (possibly using a specific color frequency)?
• Avoid chromatic aberration if possible by using monochromatic light
• Consider light sources and filters, in addition to the characteristics of your application’s substances and scenes

Additional resources you may find helpful from 1stVision’s knowledge base and blog articles: (in no particular order)

• CCS Illumination Wavelength Guide
• A Practical Guide to Machine Vision Lighting
• How machine vision filters create contrast

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.