Lighting for plastic lid print inspection without glare

Reflective plastic lids can make readable printing disappear beneath highlights, even when the camera has enough resolution. The problem becomes harder when several lids share one image, their textures vary, or handling equipment needs access above the inspection area. Start by testing opposed, off-axis bar lights rather than assuming a light directly above the lid will work. This guide explains how to adjust their geometry, compare diffusion, and verify print contrast across the complete operating envelope before choosing hardware.

Choose the geometry by print contrast, not surface brightness

For an approximately flat lid viewed from above, opposed bar lights are a useful starting point. Place them on opposite sides of the inspection area, aimed inward. Move their reflections away from the lens while retaining enough scattered light to distinguish ink from plastic.

This is a test strategy, not a guarantee. A glossy lid with curved lettering, embossed features, or a raised rim can redirect reflections toward the camera. Opposed lights can also create two highlights instead of eliminating one.

Compare these arrangements using the same representative lids:

Choose the arrangement with the strongest minimum print contrast across acceptable lid positions. Do not choose from the appearance of one bright, centered sample.

Define the angles before moving the lights

Use a repeatable coordinate system. Keep the camera approximately normal to the nominal print plane. Define light elevation from that plane: a small elevation is grazing illumination; a large elevation approaches overhead illumination.

For a point on the lid, a useful mounting estimate is:

For example, a source center 120 mm above the print plane and 170 mm horizontally away gives:

This describes only the source center. A real bar illuminates from an extended area, and each lid position sees a different range of angles.

Start with trial elevations around 20, 35, and 50 degrees. These are experimental settings, not universal recommendations. Record source height, horizontal offset, aiming direction, and camera position so a promising arrangement can be reproduced.

At grazing incidence, the flat background may become dark while raised texture becomes conspicuous. That can help surface-defect inspection but leave ordinary printing poorly separated from its background.

The objectives differ from those in scratch detection with lighting and dynamic thresholding. For print inspection, preserve readable character strokes rather than maximizing every surface feature.

Measure local contrast and clipped pixels

Judge illumination in the actual print region. Whole-image average brightness can look stable while a small highlight erases a character.

For each representative character, measure the background beside the stroke and the stroke interior. One useful comparison metric is:

Use this only when the denominator is meaningfully above the noise floor. Keep exposure, gain, and image processing fixed while comparing lighting arrangements.

Suppose an illustrative image has background intensity 180 and print intensity 60:

Elsewhere, glare might raise the print reading to 150 while the background reads 190:

The second region is brighter but has much weaker separation. These figures illustrate the metric; they are not acceptance limits for every recognition algorithm.

Also record:

• Clipped pixels inside the print region, especially where they cross character strokes.
• The weakest local contrast, not just the average across the lid.
• Background gradients that require different thresholds across the image.
• Character recognition or verification performance on known acceptable and defective samples.

Disable automatic exposure during geometry comparisons. Otherwise, camera adjustments can conceal the effect of moving a light.

Work through a multiple-lid resolution example

Lighting cannot recover print detail that the image never resolves. Calculate sampling before evaluating whether illumination alone will solve a weak reading.

Assume two lids each occupy 80 mm across the inspection direction. Allow a 20 mm gap and 30 mm of margin at each outer edge:

Suppose the narrowest relevant printed stroke is 0.30 mm. As an initial design target, assign four pixels across that stroke:

Four pixels across a stroke is a starting assumption, not a recognition guarantee. Optical blur, printing variation, motion, and the selected algorithm may require more sampling. Check narrow gaps between strokes as well as stroke widths.

Keep this field width fixed during lighting trials. Cropping down to one centered lid makes illumination easier but no longer tests the proposed inspection.

For the final optical layout, allow for lid displacement and lens performance near the field edges. If the print is only readable after aggressive sharpening, investigate focus, sampling, and motion before accepting the lighting arrangement.

Compare standard and diffused bar-light candidates

Use the lighting finder to shortlist physical formats after establishing the approximate geometry. Source dimensions must fit both the illuminated field and the available mounting space.

The LS-BAR-1047 is a standard bar-light candidate. The LS-BAR-1276 is a diffused bar-light candidate. They provide different source formats for evaluation, not a controlled comparison of diffusion alone.

Values as printed in each product's data sheet; open a part for its drawing and full specifications.

A standard source may keep its reflection away from the print more effectively in some geometries. A diffused source may reduce abrupt brightness changes, but its larger apparent luminous area can broaden a reflection. Neither behavior establishes a winner without sample images.

When comparing candidates, hold the following constant:

• Camera position, lens settings, field of view, and image processing.
• Lid position and orientation.
• The application exposure limit.
• The criteria used to judge print readability.

First compare geometry at fixed capture settings. Then optimize each candidate within its documented operating limits and the application's exposure budget. Record both stages so brightness adjustments do not obscure the reason one arrangement works better.

If ink and substrate remain poorly separated under otherwise suitable geometry, test spectral contrast. The guide to choosing LED colour for image contrast explains why visible colour alone does not predict a monochrome image's separation.

Run a controlled angle, spacing, and diffusion trial

Use a small test matrix instead of moving everything until an image looks acceptable.

1. Establish a single-bar baseline. Capture the same lid at the center and outer inspection positions. Identify which areas contain direct reflections or deep shadows.
2. Sweep elevation. Test the selected low, medium, and high settings while keeping the field of view unchanged. Record the geometry rather than relying on bracket marks alone.
3. Add the opposite bar. Capture each side independently, then both together. This distinguishes helpful fill illumination from a second source of glare.
4. Change spacing. Move both sources outward or inward symmetrically, then restore the intended aiming angles. Check every lid position after each move.
5. Evaluate diffusion. Compare the size and position of reflections, not merely their peak brightness. Record any exposure increase needed for adequate separation.
6. Challenge the arrangement. Repeat with realistic lid tilt, translation, rotation, and surface variation.

For a compact initial study, three elevations and two spacing settings give six arrangements. Capturing left-only, right-only, and both-on images produces 18 images per representative sample.

That is enough to identify trends without attempting every possible adjustment. Expand the matrix around the strongest candidates rather than adding uncontrolled variables.

Set exposure and controller requirements together

Once a geometry works on stationary samples, verify that it supplies enough light at the required production exposure.

For constant object speed, approximate motion blur using:

Using the earlier sampling example, suppose the lids move at 500 mm/s and the exposure is 100 microseconds:

This estimate covers translation during exposure, not defocus, vibration, or rolling-shutter distortion. Use the moving-part exposure guide to relate the available exposure to the inspection's detail requirements.

Select light controllers by electrical compatibility as well as channel count. Confirm the required drive mode, load, connector wiring, adjustment method, and any timing requirements through the controller finder.

The LS-CTL-0290 is a digital constant-voltage controller candidate. Its listing alone does not establish compatibility with the selected lights or suitability for pulsed operation.

If pulsed illumination is necessary, specify it separately. Check pulse duration, repetition rate, trigger timing, and exposure overlap. The strobe timing and duty-cycle guide explains why a strobe feature does not automatically authorize overdrive.

Scale illumination across all lid positions

A longer source can help cover a wider inspection area, but length alone does not establish uniform illumination. End positions may see different distances, angles, and portions of the emitting surface.

Map the complete field using actual lids. Include the nearest and farthest positions relative to each bar, both outer corners, and any location where handling tolerances allow a lid to move.

Separate two questions:

• Is the background sufficiently uniform? Large gradients may complicate segmentation or produce position-dependent thresholds.
• Is print separation sufficiently consistent? A moderate brightness gradient can be acceptable when character strokes remain distinct everywhere.

Do not apply an inverse-square estimate as though an extended bar were a point source. Near-field behavior depends on source dimensions, directionality, and the inspected geometry. Measure the actual arrangement.

Inspect simultaneous images of multiple lids, not only one lid moved between positions. Neighboring lids can cast shadows or introduce reflections that are absent during isolated tests.

Lock the selected geometry into adjustable mounts with recorded reference dimensions. Save separate recipes when lid colour, finish, or print formulation changes enough to alter the required illumination.

Know when opposed bars are the wrong answer

Use dome lights as an alternative when curved surfaces send troublesome reflections toward the camera over many orientations. A dome supplies illumination from a broad range of directions, but it still needs validation with the actual print and handling arrangement.

The LS-DOM-0085 is a dome-light candidate for that comparison. Confirm its viewing geometry, clearances, and usable coverage before selecting it.

The dome-light selection guide explains the coverage and camera-opening considerations behind this choice. Do not infer usable inspection coverage from an outside dimension alone.

Other reasons to reconsider opposed bars include:

• A raised rim shadows important printing from both sides.
• Metallic or glossy ink changes brightness sharply with orientation.
• Transparent material reveals a variable background beneath the lid.
• Lid curvature or tilt moves a highlight through the print at normal production tolerances.
• Required light positions interfere with feeding, picking, or reject handling.

Polarization can be worth testing for suitable reflective materials, but it reduces available light and may not suppress every reflection. Compare print separation at the required exposure rather than assuming a filter will solve the geometry.

When appearance changes mainly with curvature and pose, the bottle recognition lighting guide provides a related explanation of why repeatable reflected appearance matters.

Check the evidence before ordering

Use this checklist to turn a promising bench image into a documented hardware requirement:

• Representative samples: Include acceptable print variation, known defects, surface textures, colours, and production lots. A single ideal lid is insufficient.
• Position envelope: Test every lane and the permitted translation, rotation, and tilt. Record the weakest readable position.
• Optical sampling: Verify stroke widths and gaps at the field edges, with focus and motion included.
• Glare limits: Check whether highlights cross character strokes or erase small defects. Average brightness is not an adequate acceptance metric.
• Mechanical access: Reserve space for feeding, gripping, adjustment, and cleaning without disturbing light alignment.
• Electrical compatibility: Confirm the selected lights' drive requirements against controller documentation. Do not infer compatibility from category membership.
• Production timing: Validate the final exposure and any pulse timing at operating speed, including trigger variation.
• Recognition performance: Use fixed acceptance criteria on a separate validation set, then monitor false accepts and false rejects.

Define the acceptable error rates from the production risk and inspection task; there is no universal contrast threshold that proves print verification will work.

Start with the lighting configurator once the geometry and exposure requirements are recorded. If the comparison remains unresolved, send representative sample images, field dimensions, mounting constraints, and timing requirements through application support before choosing the final combination.