Best Practices for Using Optical Filters
Optical filters ramp up the accuracy and effectiveness of Cognex machine vision cameras. This training video helps you understand the importance of choosing the proper optical filters for your In-Sight vision application.
The four-part video covers the following topics:
Introduction to Filters
Machine vision cameras expand on human vision by sensing ultraviolet and infrared light. Optical filters block ambient light in production environments and help machine vision developers optimize their applications to boost speed, accuracy, and throughput. Most filters transmit light from specific bands of the electromagnetic spectrum and block unwanted light. This is essential to counting items, reading codes, inspecting for defects, and other machine vision functions.
Types of Filters
Optical filters manipulate the light waves that hit the image sensors in machine vision cameras. Examples of this type of function include:
- Short pass. Light in the short-wavelength section of the visual spectrum passes through to the camera sensor while other waves are blocked.
- Long pass. Long-wavelength light passes through while other waves are blocked.
- Band pass. Specific long- and short-wavelength bands of light reach the camera sensor while other waves are blocked.
- Notch. Light in a specific range gets blocked but other waves pass through.
- Others. Polarizers, color correctors, and neutral density filters help with a range of ambient-light challenges.
Benefits of Filters
Optical filters optimize machine vision applications, saving up to 70% of post-processing work, and offer a host of beneficial features.
- Contrast. The gray values in a digital image run on a continuum from white to black. Optical filters can tweak these values to highlight certain areas of an image and downplay others. Adjusting the contrast in images boosts accuracy and reliability.
- Resolution. Some digital images have aberrations because a lens cannot converge all wavelengths into perfect focus. A bandpass filter boosts resolution to solve these problems.
- Repeatability. Filters ensure the production environment is equivalent to the lab environment where the machine vision application was set up. Filters remove ambient light on the production line that isn’t present in the lab setting.
- Glare reduction. Polarizing filters prevent glare from overexposing parts passing through the machine vision system.
- Color correction. Optical filters help industrial cameras reconcile with light in the near-infrared portion of the visual spectrum.
- Protection. Filters provide a thin layer of defense for camera lenses. Damage to a filter is far less expensive than damage to a lens.
The basic filtering function takes colored images, converts them to grayscale and blocks certain light bands while enhancing others. Another option is to use ultraviolet and infrared filters to highlight features on a part or package that are invisible to the human eye.
Infrared filters also are effective in production environments that have bright lights, strobes, and other variations in ambient light. Infrared filters detect objects on parts and components that ambient light might otherwise blot out.
A certain amount of experimentation is required to see how filters affect the accuracy of your machine vision application. For instance, you might start out with no filter and then add a polarizer and a band-pass filter to achieve the highest clarity image.
Manager, Global Content Marketing, Cognex
A self-proclaimed marketing maverick, Mary has experience in a variety of industries and marketing activities. Her career began in the world of defense and aerospace, then shifted to collaborative robotics and now machine vision. She drives brand awareness, nurtures customer relationships, and elevates the impact of sales through compelling content. Outside of the office, Mary enjoys baking (and eating) desserts, running outdoors, and exploring new places. She is also a self-published author of a children’s book about her dog, Doma.