How to Choose the Right Machine Vision Light Source?
As we know, machine vision is part of Industry 4.0 and artificial intelligence. It can effectively help enterprises reduce production costs and improve production efficiency. Among the components of machine vision, the light source occupies an important position — choosing the right machine vision light source is half the battle.
How to Evaluate the Quality of a Light Source?
01 Contrast
Contrast is very important for machine vision. The most important task of lighting in machine vision applications is to create maximum contrast between the features to be observed and the image features to be ignored, making feature differentiation easy. Contrast is defined as a sufficient gray-level difference between a feature and its surrounding area. Good lighting should ensure that the features to be detected stand out from the background.
02 Robustness
Robustness means good adaptation to the environment. A good light source needs to have the same effect in actual work as it does in the laboratory.
03 Brightness
When choosing between two light sources, the best choice is the brighter one. If the brightness of the light source is insufficient, the aperture must be increased, which reduces the depth of field.
04 Uniformity
Uniformity is a very important technical parameter of a light source. A light source with good uniformity makes the system work stably.
05 Maintainability
Maintainability mainly refers to the light source being easy to install and easy to replace.
05 Lifetime and Heat Generation
The brightness of the light source should not decay too quickly, as this affects system stability and increases maintenance costs. Lamps with high heat generation decay faster in brightness, and the lifetime of the light source is also greatly affected.
How to Enhance Image Contrast with the Light Source?
Illuminating with light of the same color or a similar color can brighten the illuminated part;
Illuminating with light of the opposite color or a similar color can darken the illuminated part.
Different wavelengths have different penetration capabilities (penetration rates) for materials. The longer the wavelength, the stronger the penetration; the shorter the wavelength, the greater the diffusion rate on the material surface.
Tips for Choosing a Light Source
Ø If greater contrast between foreground and background is needed, consider a black-and-white camera with a color light source
Ø For ambient light issues, try a monochromatic light source with a filter
Ø For glossy curved surfaces, consider diffuse dome light
Ø For glossy, flat but rough surfaces, try coaxial diffuse light
Ø To see surface shape, consider dark field (low angle) illumination
Ø When inspecting plastics, try UV or infrared light
Ø To see features through reflective surfaces, try a low-angle line light source (dark field)
Ø When a single light source cannot effectively solve the problem, consider combined light sources
Ø Strobing can produce light 20 times stronger than constant illumination
Choosing the Angle of the Light Source
According to the expected image effect, choose light sources with different incidence angles. High-angle illumination makes the overall image brighter, suitable for non-reflective objects; low-angle illumination makes the background black and features white, highlighting the contour and surface irregularities of the measured object; multi-angle illumination gives a softer overall image, suitable for curved object detection; backlight illumination produces a black-and-white contour of the measured object, often used for dimension measurement; coaxial illumination produces black features on a bright background, used for detecting highly reflective flat objects. A diagram of light sources at different angles is shown below:
Choosing the Color of the Light Source
Consider the color of the light source and the background. Using light of the same color family as the measured object brightens the image (e.g., red light makes red objects brighter); using light of the opposite color family darkens the image (e.g., red light makes blue objects darker).
The longer the wavelength, the stronger the penetration; the shorter the wavelength, the stronger the diffusion. Infrared has strong penetration, suitable for detecting objects with poor light transmission, such as impurities in brown oral liquids. Ultraviolet is sensitive to subtle surface features, suitable for detecting places with insufficient contrast, such as text detection on edible oil bottles.
Choosing the Shape and Size of the Light Source
Mainly divided into circular, square and bar types. Normally, choose a light source with the same shape as the measured object; the final shape should be based on test results. The size of the light source should ensure uniform light across the entire field of view, slightly larger than the field of view is best.
Choosing Whether to Use a Diffuse Light Source
If the surface of the measured object is reflective, a diffuse reflection light source is preferred. Multi-angle diffuse illumination makes the overall brightness of the measured object uniform, softens the image background, and keeps detection features free from background interference.
