Selecting the optimal optical filter for your shot improves the contrast in images, reducing the processing times required to extract the relevant data from the image. Achieving the highest possible image contrast is the single-most critical factor when designing a machine-based vision system.
Your choice of aperture size, illumination level, and the quality of your lens all play a significant role in determining the performance of your system. It’s tempting for designers to enhance performance by upgrading lenses or lighting units. However, these additions can add considerably to your costs.
Fortunately, there’s a way to enhance your system performance and image quality through a more affordable option. Filters offer you the opportunity to improve your image, provided you carefully evaluate any spectral components of your target object. Filters improve performance while providing a minimal impact on the other elements of your image design.
Understanding the Different Types of Filters
There is a range of filters available for designers. The filters receive more definition according to the structure of the transmission curve.
- A long-pass filter blocks short wavelengths while allowing long wavelengths to pass through.
- A short pass filter works oppositely, blocking longer wavelengths while allowing shorter ones to pass through.
- A band pass filter transmits central wavelengths, blocking both shorter and longer wavelengths.
- A notch filter is the opposite of a bandpass filter, passing the shorter and longer wavelengths while blocking the wavelength band.
Within each of these filter groups, types of filters available depending on the technology solutions used in its creation. For instance, the colored glass filter is unavailable in notch varieties.
Designers have a vast array of filters available for use on projects. Some of the more common filters include the following.
- Conventional long pass and short pass transmission curve
- Typical notch and bandpass transmission curve
The use of a colored glass filter is an affordable and pragmatic solution for enhancing the contrast in applications. However, this practice has limitations on images where broad spectral characteristics distinguish objects, such as in the separation of purple and orange objects.
An interference filter transmits the specific range of wavelengths, and they offer more precision in use over colored glass filters. An interference filter provides the designer with a nanometer-level control over the transmission of all wavelengths. The same level of accuracy isn’t possible with a colored lens.
Polarization and neutral density filters may also assist with improving performance in specific imaging situations. Properly incorporating filters into your system requires designers to understand and comprehend the limitations and potential of each of the types of filters available to you.
Colored Glass Filter
Spectral discriminations caused by the use of a colored glass filter occurs due to the dopants present in the glass. The concentration and selection of dopants determine the transmission wavelengths and the filter attenuation.
A colored glass filter offers the designer an affordable solution for many design applications that have relaxed requirements on performance and are angle-independent. The optical transmission never shifts, even with the use of wide-angle lenses, or when tilting on the system’s optical axis.
It’s important to note that a colored glass filter features a slow transition between the transmission and blocking wavebands, with transmission curves appearing less steep than with using a coated interference filter.
There are plenty of types of color filters, including a daylight blue filter for balancing colors during the use of color sensors and polychromatic light sources.
Infrared (IR) Filters
IR filters are suitable for use in machine vision applications in color and monochrome cameras. Most machine vision cameras feature silicon image sensors that respond to infrared wavelengths. Near-infra-red wavelengths commonly occur due to overhead fluorescent lighting systems, creating inaccuracies in the camera sensors.
It’s for this reason that most color imaging cameras come with IR-cut filters as standardized equipment mounted over the sensor. Monochrome camera systems will experience massive degradation of the contrast in the image due to the presence of IR light.
For identifying small shifts in color, spectral discrimination of interference filters is a necessity due to the filter’s ability to create sharp transitions between wavelengths they block and transmit. The wavelength-selective interference filter consists of an alternating dielectric layer of low and high refraction indices depositing on a specific surface.
The uniformity and quality of the surface create a baseline optical quality for the interference filter while defining the wavelength limitations based on the transmission characteristics of the surface.
Dielectric layering produces detailed spectral characteristics from the filter, creating a destructive interference between the wavelengths that aren’t within the transmission band. As a result, it blocks the wavelengths from transmission through the interference filter.
Neutral Density Filters
Gain control over the image brightness without altering the settings for your exposure time of f/#. Both reflecting and absorbing types of neutral density filters can help you lower light transmitted to the sensor in the lens.
These filters are excellent for use in situations like capturing an image during welding. The neutral density filter reduces the intensity of the light, without compromising any of the other colors or contrast in the image.
These special filters decrease optical density away from the radial distance of the center of the image. These filters are excellent for handling image hotspots caused by reflections.
Limitations on Your Filters
Hard-coated filters get their performance from the specialized coating on the filter, but this same technology also creates limitations on the use of these filters in imagery. Interference characteristics depend primarily on the relationship between the length that light waves travel through a specific medium and the given light of the wavelength.
When traveling through interference coatings at an unfavorable angle, the light path changes through each layer of the lens, resulting in the modification of the filter’s wavelength selectivity. An interference filter functions and performs based on the distance that the light travels upon the filter.
With the proper angle of incidence, light waves will incident the filter, and destructively interfere, blocking them from passing through the filter. All interference filters feature a specific Angle Of Incidence (AOI), with most manufacturers setting it at 0°.
The angular field of view defines the acceptance angle when placing a filter in front of the lens. With a short focal length lens, light transmitted through a filter displays an undesirable effect where the slope decreases due to passband shifts down in the wavelength.
The common moniker for this effect is “Blue Shift.” For instance, a 4.5mm focal length wide-angle lens has a more significant blue shift than narrow-angled 50mm focal lenses. Designers will find that the filter behaves differently at various field points due to changes in the wavelength ranges: the further out, the more noticeable the blue shift.