In the world of machine vision systems, illumination plays a crucial role. The right lighting components are key to ensuring consistent and reliable performance.
Why? Because improper illumination can result in irrecoverable loss of information that no software algorithm can fix.
Choosing quality lighting components involves considering various parameters. These include lighting geometry, light source type, wavelength, material properties (such as color and reflectivity), item shape, application speed, mechanical constraints, environmental factors, and cost.
With so many factors to consider, the decision process can be challenging. Feasibility studies with different light types can help uncover the best options. However, there are also simple rules and best practices that can guide the selection process and enhance image quality.
The main objectives of any application are as follows:
- Maximize contrast for the features that require inspection or measurement.
- Minimize contrast for irrelevant features.
- Eliminate unwanted variations caused by factors like ambient light and non-relevant differences between items.
This guide will help you master machine vision lighting, optimize your system’s performance, and ensure the utmost accuracy in your inspections and measurements.
Understanding light in machine vision
In machine vision, the characteristics of light play a crucial role. Light is measured by its wavelength, expressed in nm (nanometers). It falls within a specific portion of the electromagnetic spectrum and can be either quasi-monochromatic (with a narrow wavelength band) or white (containing all colors).
Exploring the Electromagnetic Spectrum
The human eye can perceive light with wavelengths ranging from 400-700 nm. This falls between infrared (longer wavelengths) and ultraviolet (shorter wavelengths). Some specialized applications may require the use of infrared or ultraviolet light instead of visible light.
Interacting with Materials: Reflection, Absorption, and Transmission
When light encounters materials, it interacts in different ways:
- Reflection: Light may be reflected off the surface.
- Transmission: Light can pass through the material.
- Absorption: Light can be absorbed by the material.
As light passes through different media, it undergoes refraction, or a change in direction. The amount of refraction depends on the wavelength of the light. Violet light is bent more than red light, indicating that shorter wavelengths scatter more easily when hitting a surface. This makes shorter wavelengths ideal for surface inspection applications.
Considering wavelength as the most important factor, blue light is recommended for scratch inspection, while longer wavelengths like red light are better for enhancing the silhouette of transparent materials.
Enhance your understanding of light in machine vision and optimize your applications with the right choice of light.
LED illumination
Light is a crucial factor in machine vision, impacting its effectiveness. By measuring light in nanometers (nm), we can determine its wavelength. Light falls within a specific range on the electromagnetic spectrum and can be either quasi-monochromatic or white.
The human eye can perceive light between 400-700 nm, which is between infrared and ultraviolet. However, certain applications may require the use of infrared or ultraviolet light instead of visible light.
LEDs, or Light-Emitting Diodes, are not natural sources of white light. In fact, they emit light that is almost exclusively monochromatic. However, there are two main methods used to produce white LED light.
The first method involves combining multiple monochrome LEDs, such as red, green, and blue, to create an RGB system. By blending these primary colors, white light is achieved.
The second and more popular method involves applying a yellow phosphorus coating over the die of an LED with a short wavelength, such as blue or UV light. This combination of blue and yellow light results in the perception of white light.
This second method offers several advantages over the RGB system, including greater efficiency and improved color rendering. Different color temperatures can be achieved by adjusting the wavelength of the LED and varying the composition and thickness of the phosphorus coating.
It’s important to note that the use of white light in imaging applications should be reserved for situations where a true chromatic response is crucial, and a color camera is used. In other cases, the utilization of white light can negatively impact the performance of the optical system. This is due to the inevitable chromatic aberrations that are generated and the loss of resolution in color sensors.
LED Power Supply and Output: Ensuring Stable and Consistent Light
Controlling an LED illuminator can be achieved through two methods: adjusting the circuit’s voltage (V) or supplying it with a precise electric current (I).
It is important to note that the luminous flux produced by an LED is directly influenced by the current, whereas voltage application does not have the same effect. A mere 1% uncertainty in the driving current can lead to a 1% uncertainty in luminance. However, a 1% uncertainty in input voltage can result in a significant variation of several percentage points (refer to the picture). Thus, it is highly recommended to regulate the current rather than the voltage. This ensures stable, tightly controlled, and highly repeatable light output.
In applications such as measurement, achieving consistent and reliable results is crucial. This requires a stable grey level background, free from light flickering. Precise control of the LED forward current in the telecentric light is key to achieving this.
Take note of the provided LED current, tension, and light output graphs for further reference.
LED pulsing and strobing
LEDs have the ability to provide a steady stream of light or be turned on and off in a sequence. Operating LEDs in a pulsed mode offers several advantages such as extending their lifespan, reducing power dissipation, and minimizing heat generation.
In pulsed mode, the LED is simply switched on and off by setting the driving current or voltage to the manufacturer’s specified value for a certain amount of time before resetting to zero. Alternatively, LEDs can be driven at higher intensities, known as strobed mode, to produce more light for a limited duration.
Strobing is particularly useful when applications require a higher amount of light to capture fast-moving objects, eliminate ambient light interference, preserve LED lifespan, and synchronize with camera acquisition time. To achieve proper strobing, several parameters such as maximum pulse width (ON time) and duty cycle need to be taken into consideration.
The duty cycle, expressed as a percentage, represents the fraction of time during the cycle when the LED can be switched on. The cycle period (T) can also be measured as the cycle frequency (f) in Hertz (Hz).
By understanding and optimizing these parameters, effective triggering and strobing of LED lights can be achieved.
How to determine the maximum ton for different strobing frequencies?
LEDs offer the flexibility to provide a continuous or pulsating stream of light. When operated in pulsed mode, LEDs offer numerous benefits such as extending their lifespan, reducing power usage, and minimizing heat generation.
In pulsed mode, the LED is simply switched on and off according to the manufacturer’s specified current or voltage for a specific duration before being reset. Alternatively, LEDs can be driven at higher intensities, known as strobed mode, to generate more light for a limited period of time.
Strobing is particularly useful in applications where a higher amount of light is needed to capture fast-moving objects, eliminate interference from ambient light, preserve the LED’s lifespan, and sync with camera acquisition time. To achieve effective strobing, it is important to consider parameters such as the maximum pulse width (ON time) and duty cycle.
The duty cycle is expressed as a percentage and represents the fraction of time during the cycle when the LED can be switched on. The cycle period can also be measured as the cycle frequency in Hertz (Hz).
By understanding and optimizing these parameters, it is possible to achieve effective triggering and strobing of LED lights.
In summary, the advantages and the disadvantages of strobing LED sources are the following:
| ADVANTAGES | DISADVANTAGES |
| A large amount of light can be obtained for a short period of time (mandatory for fast application) | A light controller is necessary in order to strobe the LED source properly |
| Increase the lifespan of the LED | The synchronism between illumination and camera acquisition must be guaranteed |
| Can reduce the power dissipation |
LED lifetime
The life of an LED is defined as the time that it takes for the LED luminance to decrease to 50% of its initial luminance at an ambient temperature of 25°C.
Line speed, strobing and exposure time
When working with online applications, there are crucial factors to consider. These include setting the camera exposure time to a minimum to prevent motion blur and ensuring visibility of black and opaque objects that do not reflect light well.
To illustrate, let’s say we need to inspect an object moving at speed vo using a lens with magnification m and a camera with pixel size p.
The object’s speed on the sensor will be m times vo: vi = mvo.
Therefore, the distance traveled by the object xi during the exposure time t is xi = vit. If this distance exceeds the pixel size, the object will appear blurred over multiple pixels. Let’s assume we can tolerate a blur of up to 3 pixels, meaning we require vit < 3p.
This means the camera exposure time t must be less than 3pmvo.
For example, using p = 5.5 µm, m = 0.66, vo = 300 mm/s (equivalent to a line speed of 10,800 samples/hr on a 100 mm FoV), we determine a maximum exposure time of t = 83 µs. At such high speed, the continuous LED illuminator may not emit enough light, so using a strobing illuminator for the same duration is the recommended solution.
Another parameter we can adjust to improve lighting is the lens F/#. Lowering it allows more light to enter, but this reduces the system’s depth of field. Additionally, it may decrease image quality due to lens aberrations, resulting in overall blurriness. Increasing camera gain is another option, but it introduces noise and compromises image clarity.
Therefore, it is best practice to choose sufficiently bright lighting components, combine them with lenses set at the optimum F/#, and avoid digital increase in camera gain. This ensures accurate visibility of the object’s features during inspection.
Illumination geometries and techniques
How to determine the best illumination for a specific machine vision task? There are in fact several aspects that must be taken into account to help you choose the right illumination for your vision system with a certain degree of confidence.
Application purpose
To effectively inspect the surface of an object and identify any defects or printed text, it is essential to have front illumination. This means that light should come from the camera side. The direction and angle of the light, as well as other optical properties like diffuse or direct light, will depend on the specific surface features that need to be emphasized.
On the other hand, when measuring the diameter or length of an object or locating a through-hole, back illumination is the ideal choice. This means that light is blocked by the object on its way to the camera, maximizing contrast at the edges. However, when dealing with more complex situations such as transparent materials, determining the best lighting solution can be less straightforward and may require a combination of approaches.
Illumination angle
Determining the most suitable type of illumination and setting the angle of light hitting the object surface is crucial. We will explore the two important subgroups of front and backlight illumination: bright field and dark field illumination. Let’s dive into the four combinations that follow in the sections below. Discover the W rule for illumination and directionality.
Bright field, front light illumination
When using bright field illumination, the optics collect the light reflected by a flat surface. This is the most common scenario, and it can highlight non-flat features like defects and scratches by scattering light outside the lens’s acceptance angle. The resulting image shows dark characteristics against a bright background.
To achieve bright field, front light illumination, you can use LED barlights or ringlights, depending on the symmetry of your system. These LED lights can be either direct or diffused by a medium. Sometimes diffusing the light is preferable to ensure even illumination on reflective surfaces.
Take a look at the provided images to see the effects of front bright field illumination using a ringlight on an engraved sample and a metal coin with embossed parts.
Dark field, front light illumination
The dark field, or front light, illumination technique captures only scattered light, highlighting the non-planar features of the surface against a dark background. This effect is commonly achieved using low angle ringlights. Take a look at the diagram (Fig. 11 and 13.a – 13.b) for a better understanding of the dark field setup.
For a more specific example of this method, an engraved sample illuminated with a front dark field ringlight is shown. Additionally, an image of a metal coin with embossed parts showcasing the front dark field illumination is also provided.
Enhance your imaging with the front dark field illumination scheme and the low angle ringlight geometry.
Bright field, backlight illumination
Using bright field or backlight illumination, light is either stopped or transmitted depending on whether the material is opaque or transparent (see Fig. 14).
When the material is opaque, we can see the outline of the object as a black object on a white background (see Fig. 16 and 18). However, when the material is transparent, the non-planar features appear dark against a white background, unless there are sharp curvatures like air bubbles in plastic.
To achieve these lighting effects, diffuse backlights (see Fig. 15a, 15b, and 16) or telecentric illuminators, designed for high accuracy applications (see Fig. 17 and 18), can be used.
Example of a plastic cap with backlight illumination.
Example of a precision mechanical component with telecentric backlight illumination.
Improved Bright Field Backlight Illumination Scheme.
Diffuse Backlight Geometry (Back Emitting).
Diffuse Backlight Geometry (Side-Emitting).
Telecentric Backlight Geometry.
Dark field, backlight illumination
Experience enhanced features in a dark background with dark field backlighting. The only light that is collected is transmitted through the sample and scattered by non-flat features. Achieve this effect by using ring lights or bar lights positioned behind a transparent sample. See Figure 19 for a visual representation of this dark field backlighting scheme.
Coaxial illumination
Enhance your inspection process with coaxial illumination. When light is directed onto the object surface at a perpendicular angle, coaxial illumination is achieved. This setup can be further optimized by using collimated coaxial boxes or telecentric lenses with built-in coaxial illumination. The level of collimation determines the contrast levels and the ability to detect defects on highly reflective surfaces. Take a look at Figures 21 and 22 for a visual reference of the coaxial illumination scheme and geometry. See the difference in the image quality of an engraved sample using coaxial illumination.
Dome lights and tunnel lights
If you need to inspect an object with a complex curved shape to find specific surface features, using front light illumination from various angles is the best option. This eliminates reflections that can cause uneven lighting. Dome lights are perfect for this type of application because they offer illumination from any direction. In fact, they are often called “cloudy day illuminators” because they provide uniform lighting similar to that on a cloudy day.
Another lighting method is tunnel illumination, which provides uniform lighting on long and thin cylindrical objects. These lights have a circular opening on top, similar to dome lights.
Image of a metal coin with embossed designs, illuminated by dome lights.
Combined and advanced illumination solutions
Enhanced Illumination Solutions: Achieving Optimal Inspection
When it comes to examining intricate object geometries, a combination of various lighting techniques is often essential to uncover surface flaws with utmost precision. Consider the power of combining a dome light with a low angle light, resulting in uniform illumination across the entire field of view.
Introducing the XXXX series – a prime example of combined lighting. Featuring all-in-one dome and low angle ring lights, these innovative solutions can be effortlessly used simultaneously or independently, providing unrivaled versatility in illuminating any inspection task.
Elevate your inspection game with our groundbreaking combined light (dome + low angle ring light) illumination geometry.
Telecentric illumination
Telecentric illumination is essential in a range of applications, including high speed inspection and sorting. When paired with a telecentric lens, it allows for incredibly short exposure times, ensuring efficient throughput.
It is also valuable for silhouette imaging, accurately detecting edges and analyzing defects.
Moreover, telecentric illuminators are ideal for measuring reflective cylindrical objects. Diffuse backlights can cause unwanted reflections, distorting the appearance and leading to inaccurate measurements. By using collimated rays, telecentric illuminators eliminate this issue, providing precise and consistent readings.
In any application that requires precision measurements, accuracy, repeatability, and high throughput are crucial factors.
Using a collimated light with a telecentric lens increases the lens’s natural depth of field by around 20/30%. This improvement depends on factors such as lens type, light wavelength, and pixel size.
Furthermore, the excellent light coupling allows for increased distance between the object and the light source without compromising image quality. This is because the illuminator’s numerical aperture (NA) is lower than the telecentric lens’s NA.
Therefore, the optical system behaves as if the lens had the same NA as the illuminator in terms of field depth, while maintaining the image resolution provided by the actual telecentric lens NA.
When inspecting objects with curved edges, collimated light is the optimal choice. This illumination technique is extensively used in measurement systems for shafts, tubes, screws, springs, o-rings, and similar samples.
Collimated light offers distinct advantages over diffuse backlight illumination.
Wavelength and optical performance
The performance of an optical system is influenced by the wavelength of the light used.
When designing an optical system, you have the option to optimize it for a single wavelength or for a range of wavelengths.
Optimizing for a wider range of wavelengths requires a more complex optical system to correct chromatic aberrations. It’s better to use monochromatic lighting when possible because it simplifies the optical system and improves stability and efficiency.
Choosing the right wavelength depends on factors such as the surface of the sample, desired resolution, system complexity limits, and availability of illuminators.
Using a wavelength outside the design range will significantly degrade the performance of the optical system.
In machine vision applications, selecting the proper light wavelength is crucial for highlighting specific features of an object.
Contrast can be maximized by selecting a light color opposite to the feature color.
It’s important to consider the difference in sensitivity between human eyes and image sensors, and to assess how the vision system perceives the object.
Monochromatic light can be achieved through optical filters or monochromatic sources.
Optical filters
The performance of an optical system is influenced by the wavelength of the light used. When designing an optical system, you can choose to optimize it for a single wavelength or a range of wavelengths. However, optimizing for a wider range of wavelengths requires a more complex optical system to correct chromatic aberrations. It is generally better to use monochromatic lighting when possible, as this simplifies the optical system and improves stability and efficiency.
The choice of wavelength will depend on factors such as the surface of the sample, desired resolution, system complexity limits, and availability of illuminators. Selecting a wavelength outside of the design range will significantly degrade the optical system’s performance.
In machine vision applications, selecting the right light wavelength is crucial for highlighting specific features of an object. Maximizing contrast can be achieved by choosing a light color that is opposite to the feature color. It is also important to consider the difference in sensitivity between human eyes and image sensors, and how the vision system perceives the object.
Monochromatic light can be achieved through the use of optical filters or monochromatic sources.
Structured illumination
Structured light, also known as structured illumination, involves projecting light with a predetermined pattern onto a scene. This technique is primarily used to detect and measure deformations in the projected pattern, contributing to 3D object reconstruction.
There are various forms of structured light projection, including projecting a grid of lines to reconstruct surface shape, using a single line (known as a sheet of light) to scan an object and recreate its surface, employing a pseudo-random cloud of dots for precise 3D reconstruction, and projecting sinusoidal patterns for advanced surface analysis.
While both LED and laser sources are used for pattern projection, lasers have some drawbacks. Laser light produces a Gaussian-shaped line profile, with the highest intensity at the center and lower intensity towards the edges. Laser illumination can also create a phenomenon called “speckle” on reflective surfaces, where coherent radiation scatters. LED lighting, which is non-coherent, avoids these issues.
By using LED lighting for structured illumination, we can overcome these problems. RODER LED pattern projectors offer advantages such as thinner lines, sharper edges, and more uniform illumination compared to lasers. LED light, being non-coherent, delivers consistent intensity across the pattern and eliminates visible speckle. Additionally, LED lighting allows for easy production and usage of white light in the projection process.
Illumination safety and class risks of LEDs according to EN 62471
IEC/EN 62471 provides guidelines for assessing the safety of lamps that emit optical radiation, such as LEDs (excluding lasers), in the wavelength range of 200 nm to 3000 nm. The standard categorizes light sources into different risk groups based on their potential to cause photobiological harm.
Risk Groups:
Exempt: No risk of causing photobiological harm.
Group Ia: No risk of photobiological harm under normal behavioral limitations.
Group II: No risk of harm due to aversion response to bright light or thermal discomfort.
Group III: Hazardous even for momentary exposure.
Edge diffraction
Edge diffraction is a mesmerizing optical phenomenon that occurs when a diverse range of objects is struck by a focused beam of light. This striking effect can be observed on a screen placed behind the object or on the image plane of a lens projecting the object onto a sensor. It manifests as a captivating white border along the edge of the object’s image.