Leather defect inspection lighting for wrinkled surfaces

Wrinkled leather creates two competing inspection problems: surface relief produces changing shadows, while gloss can hide defects or printed lines. Increasing overall brightness rarely solves both. Broad illumination may reduce wrinkle shading but also weaken features that depend on directional reflection.

For inspection developers and equipment buyers, the useful question is which lighting geometry separates unacceptable defects from acceptable grain and wrinkles. Start with a diffuse reference image, compare directional illumination, and add polarization only where reflections remain troublesome. The evaluation below keeps defect visibility and printed-line readability as separate acceptance requirements.

Define what must remain visible

Separate the inspection requirements before choosing a light. A stain, a shallow scratch, a missing printed segment and a fold are not equivalent optical targets.

• Color or reflectance defects: discoloration, contamination and inconsistent coating often benefit from relatively even illumination.
• Surface-relief defects: cuts, raised edges and impressions may become clearer under directional light.
• Printed features: lines and symbols need sufficient contrast against the surrounding finish, regardless of wrinkle shading.
• Presentation changes: acceptable grain, local slopes and folds must not trigger excessive rejection.

Include representative acceptable material in the evaluation, not only obvious defects. Otherwise, a lighting arrangement that exaggerates every wrinkle can appear successful until normal production starts.

Define the mechanical presentation as well. Record how the material is supported, its permitted height variation and whether gentle flattening is allowed. Stretching leather during evaluation can change both genuine defects and normal texture. Use the same presentation method expected during inspection.

A useful initial sample set contains light and dark finishes, glossy and matte regions, weak printing and shallow defects. An exploratory set of 10–20 samples can expose obvious weaknesses, but it is not evidence of production detection performance.

Compare diffuse, directional and polarized illumination

Begin with broad diffuse illumination as a reference, not as an automatic final choice. Its purpose is to reduce strong changes caused by individual reflection directions. It does not remove the leather's physical wrinkles.

The closest starting point is the guide to dome lighting for shiny, curved and wrinkled parts. For leather, the additional requirement is to preserve printed features while deciding how much relief contrast the inspection needs.

Low-angle illumination can create a dark-field condition when direct reflection misses the camera and disturbed regions scatter light toward it. On wrinkled leather, local slopes can redirect reflection back into the lens, producing bright patches unrelated to defects.

Use the explanation of lighting angles for surface inspection to distinguish illumination geometry from brightness adjustment. Raising intensity cannot correct a reflection entering the lens from an unfavorable direction.

For directional trials, illuminate from opposing directions separately. Combining both directions immediately may fill the shadows that made a shallow defect visible.

Size the image before sizing the illumination

Specify the smallest feature and the required field of view before comparing light sizes. A feature that occupies too few pixels will remain unreliable even when the lighting looks attractive.

Consider an illustrative inspection field of 240 × 160 mm. Suppose the narrowest printed line to inspect is 0.4 mm wide. An initial sampling target of 4 pixels across that line gives:

This is a starting target, not a guarantee of detection. Lens sharpness, defocus, noise and line contrast all affect the usable sampling margin. A requirement to measure a line precisely may need more sampling than simple presence detection.

Now evaluate illumination over the entire field, including permitted height variation. The apparent angular size of a diffuse source matters, not simply whether its outside dimensions exceed the inspected region.

A source close to the leather may cover a broad range of directions at the center but behave differently near an edge. Increasing separation can improve access and coverage while reducing the angular extent seen by the surface. Check both effects with actual images.

For larger sheets, use the coverage principles in large-field illumination planning. Do not scale a successful small sample arrangement by length alone: edge coverage, camera clearance and reflection geometry also change.

Use polarization to solve a measured glare problem

Cross polarization places a polarizer in the illumination path and an analyzer in front of the lens. Their relative orientation reduces the directly reflected polarized component reaching the camera.

Where practical, place the illumination polarizer after the final diffuser. A diffuser placed after it can alter the polarization state and reduce the intended suppression. Confirm the arrangement with the actual light, diffuser, lens and sample.

Rotate the analyzer while observing three regions simultaneously:

1. A glare patch that obscures information.
2. A genuine defect near the acceptance boundary.
3. A printed line that must remain readable.

Choose an orientation based on all three, rather than the darkest glare patch alone. Some finishes produce useful contrast through gloss differences; removing those reflections may also remove the signal you wanted.

The guide to cross polarization for surface-glare reduction explains the arrangement in more detail. Treat its effectiveness as material-dependent, especially across different coatings and surface slopes.

Polarizers also reduce received light. After comparing geometry at fixed camera settings, establish whether the polarized arrangement can meet the required exposure without excessive gain. A clear stationary image is insufficient if production motion requires a much shorter exposure.

Run a repeatable lighting evaluation

Use a structured comparison rather than choosing the most visually pleasing image. Keep the lens, camera position, focus, support surface and image-processing settings unchanged during each comparison stage.

1. Capture a diffuse reference. Disable automatic exposure and gain. Keep highlights below clipping and record the acquisition settings.
2. Add directional illumination. Change the illumination direction while leaving the sample fixed, then repeat with rotated samples.
3. Evaluate polarization. Record the polarizer arrangement and analyzer orientation so the setup can be reproduced.
4. Measure separate targets. Track defect contrast, printed-line contrast and normal wrinkle variation independently.
5. Optimize each candidate fairly. After the fixed-setting comparison, adjust exposure or permitted light output within the same motion and operating constraints.
6. Repeat across positions. Include the center and corners of the inspection field, plus the permitted material-height extremes.

For an initial screen using 12 samples, 4 orientations and 3 lighting configurations, the capture count is:

That count excludes repeats and field-position checks. It is an evaluation workload, not a statistical acceptance plan.

For a local feature and adjacent background, a simple contrast measure is:

Use linear, black-level-corrected image values where available. Avoid clipped regions and use comparable sampling areas.

As an arithmetic example, background and feature readings of 150 and 120 give 30 / 270 = 0.111, or about 11% contrast. Another arrangement giving 100 and 65 produces 35 / 165 = 0.212, or about 21%.

These are illustrative readings, not product measurements. The second arrangement has higher normalized contrast, but its lower intensity may also increase sensitivity to noise. Compare repeat variation and printed-line visibility before declaring it better.

Choose candidate lights without assuming coverage

Use shadowless lights to shortlist broad diffuse arrangements. The name describes an illumination approach, not a promise that all wrinkle shading disappears.

The LS-FDL-0002-L8 is a flat dome candidate. Check its documented lit area against the inspection region, then evaluate the actual reflection coverage at the intended separation.

The LS-SDL-0006 is a square shadowless candidate. Confirm viewing clearance, mounting access and usable coverage rather than selecting from the outside dimensions alone.

Use bar lights for adjustable directional trials. The LS-BAR-1047 is a standard bar-light option, while the LS-BAR-1276 is a diffused bar-light option. Diffusion softens the source appearance, but a narrow diffused source does not automatically provide dome-like angular coverage.

The lighting finder helps narrow the shortlist by geometry, size and color. The final choice still requires representative samples at the intended camera and light positions.

Review lighting accessories and the accessory finder for supporting components. Do not assume that a diffuser fits a particular bar or that a cable matches its connector. Confirm compatibility, mounting method and electrical requirements before ordering. Polarizer availability and fit also need separate confirmation.

Check exposure and motion before choosing one image or two

If the leather moves during acquisition, calculate the exposure budget from acceptable image blur. For example, use the earlier sampling of 10 pixels/mm and allow 1 pixel of motion blur. That corresponds to 0.1 mm at the object.

At a material speed of 200 mm/s:

This estimate assumes constant motion perpendicular to the viewing direction. Include speed variation and timing uncertainty when setting the practical limit.

Check whether the selected illumination produces adequate signal within that exposure. Do not assume a light supports overdrive or a particular pulse setting. Confirm the permitted operating mode and controller compatibility from the relevant documentation. The guide to machine-vision lighting operating modes helps separate continuous, switched and strobed operation.

If no single image preserves both print and relief information, consider separate diffuse and directional captures. Keep their acceptance criteria separate rather than immediately averaging the images together.

Multiple captures introduce timing and alignment requirements. Material motion, wrinkle movement or vibration between exposures can make corresponding regions differ. Verify that the intended feature remains visible in each required channel and that the processing can associate the correct locations.

Verify the arrangement before ordering

Use the following checklist to turn an attractive trial image into a documented selection.

• Smallest defect: confirm that the weakest unacceptable feature remains visible, not merely the largest demonstration defect.
• Printed-line retention: check faint printing and lines crossing wrinkles. A clean background is not useful if required markings disappear.
• Normal texture: include acceptable grain and folds that resemble defects under directional light.
• Orientation: test directional features at 0°, 45°, 90° and 135° relative to the illumination direction.
• Field coverage: inspect the center and all corners using representative material, not only a flat reference target.
• Height variation: repeat at the permitted presentation extremes, checking focus as well as reflection changes.
• Exposure margin: confirm signal quality at the required motion-limited exposure without relying on automatic gain changes.
• Mechanical repeatability: record light position, angle, separation, analyzer orientation and support method.
• Electrical compatibility: confirm supply, controller, connector and cable requirements for the selected configuration.
• Acceptance evidence: record false rejections on acceptable material and missed detections on labeled defects using samples not used for setup adjustment.

Finally, repeat captures after the arrangement reaches its normal operating condition. The lighting stability test before deployment provides a useful framework for checking drift rather than relying on a single capture.

When the usual lighting choice is wrong

Broad diffuse illumination is the wrong final choice when the required defect is a shallow relief feature that becomes indistinguishable from its background. Keep the diffuse image for printing or color inspection, but evaluate a directional channel for that defect.

Low-angle illumination is the wrong choice when acceptable wrinkles dominate the image or when defects disappear with orientation. Try a different illumination direction, broader angular coverage or separate directional captures before adding complicated processing.

Cross polarization is the wrong default when useful printing depends on glossy reflection, or when its light loss makes the exposure requirement impractical. Compare polarized and unpolarized images using the same acceptance targets.

Also reconsider the presentation when folds repeatedly conceal inspection areas. No threshold adjustment can recover information that never reaches the camera. Establish a controlled view before selecting processing methods.

The practical recommendation is therefore conditional: start diffuse, add directionality for relief, and use polarization only when it improves the required signals. Select the arrangement with the best repeatable separation between acceptable material and unacceptable defects—not the smoothest-looking background.

Bring representative samples, the field of view, minimum defect size, line speed and presentation limits to the light configurator. If diffuse and directional requirements conflict, use application support to review a controlled evaluation before committing to the final arrangement.