As the cost of vision systems continues to decline, machine vision systems are becoming more practical for converters. But no system will operate at optimal levels unless you follow these six essential machine vision lighting techniques.
We’ve talked about how machine vision inspection systems can help you spot production mistakes and keep your line running efficiently.
While many converters are familiar with the systems, many have shied away from them in the past, partially due to a lack of understanding how the systems work. “Vision is almost like black magic to people who haven’t used it,” said Sam Liebo of Automation, Inc.
A major part of that involves lighting techniques. We’ll shed some light on how they impact the “black magic” with Liebo’s help. But first, let’s take a look at why vision systems have become more affordable.
Less custom coding = lower overall costs
For the last few decades, the idea of putting a machine vision system on an item like a food product would have been crazy. The systems were typically used only in the production of pharma or medical devices due to the price tag involved.
So why is it no longer far-fetched to think we’re using machine vision cameras to verify items like mac and cheese barcodes?
The root cause is the standardization of inspection software tools. With more tools now available, there’s less custom-coded software required for each inspection. These cameras inspect better and faster thanks to the improved software, resulting in lower overall costs.
A more economical price tag allows machine vision systems to enter the realm of possibility for most converters. The next step is to demystify how they work, and improve their performance in the process.
Lights, camera, success: Six lighting techniques that will improve performance
According to Liebo, three components impact a machine vision system’s performance:
- Lighting – 80%
- Lensing and camera selection – 15%
- Software – 5%
The importance of lighting may come as a surprise, but it’s logical when you consider its impact on the image. “It doesn’t matter how good the camera or the software is,” Sam said. “If you don’t have a good image coming in, you’re not getting a good image coming out.”
To improve the quality of the image, use these six techniques:
1. Account for light reflecting
Liebo often sees converters attempting to just put a light around the lens, and avoid more costly backlights. That’s a big mistake. “The material you’re using on a converting machine is reflective,” he said. “When the light hits the surface, it reflects right back.”
If you shoot straight down and put a ring around the lens, the light will reflect right back up and wash out anything you want to see. Instead, angle bar lights, or use backlighting.
2. Use the proper color light for the materials
The electromagnetic spectrum is a continuum of all electromagnetic waves according to frequency and wavelengths. (Source) Different wavelengths will have different effects on an object. TV and radio waves, for example, are high wavelengths that penetrate buildings.
Light occupies a small portion of the broad electromagnetic spectrum, and the colors they emit will impact how the image appears in front of a camera. Lower wavelengths, like blue and ultra-violet, can excite materials while higher wavelengths, like red and infrared, can penetrate materials.
Using the right types of light for the specific materials is critical.
3. Strive for consistent lighting
The goal for any type of lighting is consistency. Using ambient lighting, such as daylight, can be problematic, as images can be affected by time of day or cloud cover. Fluorescent lighting can cause problems, as they flicker at 60 hz, and once they’re on, they start dimming immediately.
Liebo recommends LED lighting, as it provides consistent lighting for an extended period of time. “If your lighting is inconsistent you are no longer measuring your product, you are measuring the change in lighting,” he said.
4. Be aware of the restrictions of different types of lens
There are two types of lenses – standard and telecentric lens.
Standard lenses, typically used in basic cameras, are lower cost and smaller than telecentric lenses. They have greater flexibility; allow you to shoot a greater distance; and have a larger field of view.
However, the standard lens will introduce distortion into the image because of the curvature of the lens. “It’s like you’re looking at the outside of a big ball,” Liebo said. You’ll have to calibrate accordingly because of the distortion, and there will be inaccuracies. Chase Campbell at Automation, Inc. discusses lens distortion and calibration in this post.
Telecentric lenses, on the other hand, look straight down, and the calibration factor is consistent throughout the entire image. However, the lens is bigger, more expensive, and only works at a certain distance. This limits your flexibility.
“You’re not going to get a field of view bigger than 4” with a telecentric,” Liebo said. Perhaps with the new generation of higher resolution cameras, there could be more of a push for it. But due to the amount of glass and the weight involved, this option hasn’t been as cost-effective.
5. Match the right image format from your lens to your camera
Liebo notes that a big mistake people make is incorrectly matching the image format on the lens to the camera.
If you look at the image format on a lens and it says ⅔”, then you can’t assume it can handle a 5MP ⅔” camera. “That is not taking the size of the pixels into account at all and may leave you with a less than ideal image,” Liebo said.
You need to find the lens resolution relative to the pixel size on the camera, and the max image format on the lens relative to the camera. Liebo has produced a great post on lens selection to optimize camera resolution.
6. Determine the correct exposure time to prevent motion blur
A camera acquires light for a certain amount of time to form an image — this is its exposure time. When you have a moving object, failure to light the acquisition parameters correctly can result in a blurry image.
“If the exposure time is too long, a discrete point on the object will move through multiple pixel locations in the image, causing the image to be blurred as shown below,” notes Alex Maslakow of Automation, Inc.
Maslakow details how to calculate the required exposure time in this post.
Test, test and test some more
Obviously there are a lot of lighting techniques to consider. But the surefire path to success can be summed up in three words: Test, test, test.
When Delta ModTech works with Automation, Inc. on projects using machine vision systems, we have them test every new application. “I hear a lot of converters saying ‘I’ve done it before like this,’ but even so, you still need to test it out,” he said.
Every application requires the right lighting, lens and camera. Understanding aspects of machine vision systems such as lighting techniques can help you learn the science behind the “black magic.” And that translates into lights out operational performance.