Lighting Is the Foundation of Every Vision System

Ask any vision engineer what makes or breaks an inspection application, and the answer is almost always the same: lighting. You can have the best camera, the fastest processor, and the most sophisticated algorithm on the market, but if your lighting is wrong, none of it matters. Poor illumination creates inconsistent contrast, introduces shadows that confuse edge-detection routines, and generates false rejects that erode operator trust in the system.

In our experience integrating machine vision solutions across hundreds of manufacturing applications, we have found that getting the lighting right early in the design process eliminates the majority of downstream headaches. This post walks through the core lighting techniques used in industrial vision and explains when and why each one works.

Why Lighting Matters More Than the Camera

A camera sensor records reflected (or transmitted) light. It has no concept of what it is looking at. The algorithm running on the processor interprets pixel intensity values to find edges, measure features, read codes, or detect defects. If those intensity values are ambiguous—because of glare, uneven illumination, or insufficient contrast—the algorithm struggles.

Increasing camera resolution or switching to a more expensive sensor rarely fixes a lighting problem. A 20-megapixel image of a poorly lit part is still a poorly lit part. The correct approach is to engineer the lighting geometry and wavelength so that the features of interest produce strong, repeatable contrast against the background. Once that contrast exists, even a modest camera and a straightforward algorithm can deliver reliable results at production speeds.

Core Lighting Techniques

Backlighting

Backlighting places the light source directly behind the part, opposite the camera. The part appears as a dark silhouette against a bright, uniform background. This technique excels at dimensional measurement, gap and flush inspection, and detecting holes, slots, or edge profiles.

Because backlighting produces a high-contrast binary image, edge-detection algorithms run quickly and with high repeatability. It is one of the most reliable techniques available when the application allows it. The limitation is that backlighting only reveals the outer profile of a part—it cannot show surface detail, color, or texture.

Bright Field (Direct) Lighting

Bright field illumination positions lights at a relatively shallow angle to the part surface, illuminating it from the same side as the camera. Light reflects off the surface and returns toward the lens. Smooth, flat surfaces appear bright; angled surfaces or edges scatter light away and appear darker.

This is the most common general-purpose lighting geometry. It works well for reading printed text and barcodes, inspecting labels, and verifying assembly presence. The key challenge with bright field lighting is controlling specular reflections from glossy or metallic surfaces, which can wash out features or create hot spots.

Dark Field Lighting

Dark field lighting positions the illuminator at a very low angle—nearly parallel to the surface. On a smooth, flat surface, light skims across and reflects away from the camera, so the background appears dark. Any raised feature, scratch, edge, or surface defect scatters light upward into the lens and appears bright against the dark background.

This technique is highly effective for detecting surface scratches, cracks, burrs, engraving, and embossed features. It is widely used in metal parts inspection where surface defects must be found on machined or polished surfaces. Dark field is also the go-to approach for inspecting transparent materials like glass or film, where conventional bright field illumination passes straight through.

Diffuse Dome (Cloudy Day) Lighting

A dome light surrounds the part with uniform, diffuse illumination from all directions—similar to the lighting on an overcast day. This eliminates shadows and minimizes specular reflections from curved or shiny surfaces. The result is even, consistent illumination that reveals surface features without glare.

Dome lighting is the standard choice for inspecting curved or irregularly shaped parts, highly reflective components like chrome or polished metal, and applications where color accuracy matters. It is also commonly used for color vision inspection where consistent illumination is essential for accurate color measurement.

Structured Light

Structured light projects a known pattern—typically a line, grid, or series of stripes—onto the part surface. A camera views the pattern from an offset angle. When the pattern falls on a 3D surface, the lines deform according to the surface topology, and triangulation algorithms calculate the height profile.

This technique is essential for 3D surface inspection, measuring flatness, detecting dents or warpage, and inspecting solder joints or weld beads. Structured light has become increasingly accessible as laser line generators and smart cameras with built-in 3D processing have dropped in cost and complexity.

Coaxial Lighting

Coaxial illumination uses a beam splitter to direct light along the same axis as the camera lens. Light travels straight down onto the surface and reflects straight back into the lens. Flat, mirror-like surfaces appear bright, while angled or textured surfaces scatter light and appear dark.

This is the primary technique for inspecting highly reflective flat surfaces—wafer inspection, polished metal surfaces, and flat glass. It is also used for reading laser-etched markings on shiny surfaces where other techniques produce too much glare.

Choosing the Right Wavelength

Beyond geometry, the wavelength of the illumination matters. Most industrial vision applications use one of three options:

  • White light provides natural-looking images and is required for color inspection applications.
  • Red or infrared LEDs are commonly used for monochrome applications because silicon-based sensors have high sensitivity at these wavelengths, and red light is less affected by ambient lighting interference.
  • Blue or UV light enhances contrast for specific materials. Blue light improves surface defect detection on certain metals, and UV illumination causes some materials to fluoresce, which is useful for detecting adhesive presence or contamination.

Filters paired with specific wavelengths can further enhance contrast. A common example is using a red LED with a red bandpass filter to eliminate ambient factory lighting from the image.

Controlling the Environment

Even the best engineered lighting solution will degrade if the environment works against it. Key environmental factors to manage include:

  • Ambient light from overhead factory lighting, windows, and welding arcs. Shielding the inspection zone with an enclosure or shroud is the most reliable mitigation. Strobed illumination synchronized with the camera exposure can also help by overwhelming ambient light during the short capture window.
  • Part variability in surface finish, color, and contamination. The lighting must produce adequate contrast across the full range of acceptable parts, not just the ideal sample.
  • Mechanical stability ensuring the part-to-camera and part-to-light distances remain consistent. Even small variations in working distance can shift illumination intensity and angle enough to affect results.

Integration With the Overall Vision System

Lighting is one element in a system that includes the camera, lens, processor, and algorithm. Decisions about lighting geometry and wavelength directly influence camera selection because the lighting determines how much light reaches the sensor, which in turn affects exposure time, gain settings, and the required frame rate capability.

During system design, we recommend evaluating lighting and camera together using actual production parts—not just ideal samples. Testing with worst-case parts under candidate lighting configurations before finalizing the design prevents costly rework after installation.

Practical Recommendations

Based on years of deploying vision systems in production environments, here are guidelines we follow:

  1. Start with the defect or feature, not the camera. Understand what you need to see, then select lighting that makes it visible.
  2. Test multiple geometries. Bench-test parts under backlight, bright field, dark field, and dome illumination before committing. The results are often surprising.
  3. Use LED illumination. LEDs offer long life, consistent output, fast strobing capability, and wavelength selectivity. Halogen and fluorescent sources have largely been replaced in industrial vision.
  4. Strobe whenever possible. Strobing the light in sync with the camera eliminates motion blur and reduces sensitivity to ambient light changes.
  5. Plan for degradation. LED output decreases over time. Design the system with sufficient margin so that gradual intensity changes do not cause failures.

Partner With AMD Machines

Lighting selection is one of the most consequential decisions in a vision system project, and it is best made early with input from experienced integrators. AMD Machines has designed and deployed vision-based inspection systems across automotive, medical device, electronics, and consumer products manufacturing. Contact us to discuss your inspection challenge and learn how the right lighting approach can deliver the reliable, repeatable results your production line demands.