Lighting shiny metal pins without glare: dome, diffuse and polarizer

Shiny cylindrical metal pins — the connector contacts on a Type-C or similar assembly — are difficult small parts to light. The classic sequence goes like this: a coaxial light gives a saturated stripe down every pin; a low-angle ring light moves the stripe but does not remove it; the edges you need to gauge disappear into the highlight. This troubleshooting guide works through the causes, practical fixes using diffuse lighting, polarization trials and geometry changes, and ways to verify the result before you order parts. You do not need exotic equipment to begin improving pin-edge contrast.

The symptom: hotspots that swallow the pin edges

The image looks usable until you zoom in: each pin carries a bright core running along its axis, and the true edge of the pin disappears into the surrounding blur. Thresholding fails in two directions at once — sometimes merging neighbouring pins, sometimes breaking one pin into segments.

Two clues suggest that illumination geometry or exposure needs attention before software tuning:

• The hotspot moves when you tilt the part or the light by a few degrees. A specular reflection can travel across the surface as the geometry changes.
• The camera histogram is clipped at the top end. If the brightest pixels are pinned at maximum, threshold tuning cannot recover the intensity differences lost to saturation.

If both are true, fix the light first, exactly as with glass bottle defect detection: fix lighting before thresholding and scratch defect detection with lighting and dynamic thresholding. The same principle applies to pins.

Cause 1: coaxial light on a cylinder behaves like a mirror

A coaxial light directs illumination along the viewing axis, which can be useful for flat, shiny or engraved surfaces — the applications covered by the coaxial lights category. A polished cylinder, however, has changing surface normals across its width. Only a narrow region reflects the source toward the camera, producing a bright stripe. Coaxial illumination is not necessarily collimated; its angular spread depends on the optical design.

A quick test: rotate the pin 90 degrees in the image plane under the coaxial light. A reflection stripe generally follows the cylinder's axis. Then tilt the part slightly and watch whether the stripe shifts across its surface. Use both observations rather than treating rotation alone as proof of a reflection.

The narrow angle coaxial light LS-CXL-0019 is an option for testing more directional coaxial illumination. On cylindrical surfaces, a narrower illumination angle can produce a more localized reflection rather than removing glare.

Cause 2: low-angle ring light moves the glare, not the glow

Switching to a low-angle ring light is a common next move, and it can help. Light arriving around 0–15 degrees above the part plane can leave upper surfaces dark while highlighting edges. But each polished pin can still reflect portions of the ring. You may trade one hotspot for several smaller ones rather than eliminate saturation.

A diffused ring softens individual LED reflections, although it does not necessarily remove the reflection of the ring itself. The LS-RNG-0143 is a 100 mm dia. ring light with a diffuser and a 30° angle. It provides a diffused-ring option for comparison, rather than the grazing geometry of a low-angle ring.

Fix A: surround the pin with diffuse light

Distinct highlights remain whenever the part reflects a small, bright source toward the lens. Broader, more uniform illumination can soften those reflections. Two product families are worth testing:

• Dome lights — the classic cloudy-day source, as covered in dome lights for shiny, curved and wrinkled parts. A dome presents a broad, diffuse hemisphere to the part, reducing abrupt brightness changes across curved surfaces.
• Spherical integral lights — a diffuse enclosure with a camera port. This geometry can provide broad illumination around polished round parts, but the result still depends on the access openings and fixture. The LS-DOM-0133 is a 134 mm dia. spherical integral light. Check the dimensional drawing for usable opening size and fixture clearance.

As an initial sizing trial, consider a diffuse opening roughly twice the part's largest dimension, then verify coverage across the complete position range. This is a starting point, not a guaranteed sizing rule. Working height, recess depth and camera-port reflections can matter more than the outside diameter.

The LS-DOM-0085 has a Ø116 mm emitting area. Check its geometry against the connector, fixture and required viewing angle rather than assuming that emitting-area diameter alone guarantees coverage.

A dome does not make polished metal stop reflecting. It changes what the metal reflects. The camera port and part-access opening can still appear as dark features, while excessive exposure can saturate even a uniform reflection.

Browse both families in the dome lights category, or use the lighting configurator to begin narrowing the options.

Fix B: crossed polarization plus a bandpass filter

When diffuse illumination still leaves unwanted contrast, test polarization on the actual pin finish. Crossed polarizers are not a guaranteed cure for bare-metal glare: attenuation depends on the material, incidence angle and changes in polarization during reflection.

1. Fit suitable linear polarizer film over the light's emitting area, within its temperature and irradiance limits.
2. Fit a linear analyzer on the lens and rotate it while watching both the highlight and the useful edge contrast.
3. Measure the light loss before increasing exposure, strobe duration or current. Stay within the documented ratings of the light and controller, and check motion blur. Review compatible lighting options through the lighting finder.
4. If ambient illumination contributes to the image, test a bandpass filter matched to the source wavelength. It rejects out-of-band ambient light, not ambient light within its passband.

For an exposure-budget example, a measured loss of two stops requires approximately four times the exposure at unchanged illumination and gain. Increasing pulse duration by that amount may create motion blur, so compare filtered and unfiltered images at the intended production speed.

Polarization is useful only if it improves the inspection signal relative to unwanted reflections. A darker image alone is not evidence of improvement. For further background on tuning light and glare settings, this glare reduction and lighting adjustment walkthrough discusses related controls from a vision-sensor perspective.

Fix C: geometry tricks when you must use the lights you have

If the lab cupboard has only coaxial and ring units, three bench trials can help recover usable edges:

• Try a tilt of 15–25 degrees. Angle the camera or part fixture and observe whether the troublesome reflection moves away from the measured edge. A cylinder can still reflect the source elsewhere, so tilt is not a guaranteed cure. Oblique viewing introduces foreshortening and potentially perspective distortion; recalibrate the measurement and verify it over the actual part-height range.
• Build a diffuse tent. Surround the pin area with suitable translucent material and illuminate it from outside. The tent wall becomes a broad source, approximating a dome. Keep material clear of hot components and moving equipment. Judge the result from edge contrast and saturation rather than assuming a fixed relationship between source area and hotspot brightness.
• Flip to dark field. Keep the low-angle ring and let the pin edges appear bright against a dark background. Gauge those outlines only if their positions remain stable across acceptable part poses. This is related to the edge-contrast trade-off discussed in collimated backlight or diffuse backlight for edge gauging.

How to verify the fix

Define the measurement scale before judging edge stability. For example, a 12 mm field of view across 1200 pixels gives 12 mm / 1200 pixels = 0.01 mm per pixel. A 0.50 mm pin then spans about 50 pixels, and an allowable width variation of 0.02 mm corresponds to 2 pixels. These are sampling calculations, not a guarantee of measurement accuracy.

Run these four checks before committing to a production set-up:

1. Histogram. Avoid clipping in the measurement region. As an initial exposure target, keep the brightest pin pixels below about 90% of full scale, then confirm that the edges retain adequate signal.
2. Rotation test. Rotate the part 90 degrees where the fixture permits and remeasure. Use a reference with known geometry, because real pin shape can also cause orientation-dependent readings. Compare the result with your allocated measurement-error limit.
3. Wiggle test. Shift the part 1 mm in X and Y, or through its actual production position range if different. After localization, measured width should remain within the agreed tolerance.
4. Repeatability over 100 frames. Record a fixed pin's measured width and calculate its standard deviation. A possible initial target is less than 10% of the allowed width variation; set the final acceptance criterion from the inspection requirements.

If checks fail, isolate the cause rather than assuming ambient light is responsible. Capture a frame with the inspection light off to estimate ambient contribution. Then check exposure, focus, fixture movement, source uniformity and edge-tool settings separately.

Repeat the successful test at production speed. A stationary bench image does not reveal motion blur, trigger variation or changes caused by part presentation.

Checklist before you order

• Part's largest dimension and full position range recorded; diffuse-opening coverage verified on representative parts
• Working distance, camera-port diameter and fixture clearance checked against the light's drawing
• Exposure and strobe headroom measured if polarization is planned, with motion blur included in the budget
• Bandpass-filter passband matched to the source spectrum, with remaining ambient contribution measured
• Fixture allows a tilt trial where practical, with measurement calibration checked afterward
• Light and controller ratings compatible with the required current, pulse duration and duty cycle

Start with the lighting finder, and use the system configurator to bring light, controller and camera choices together. If your pin geometry is unusual — very long pins or recessed contacts — use the contact page for an application-specific recommendation. Include representative images, field of view, working distance, line speed and measurement tolerance.