3D visionGuide

3D profile sensors for groove measurement in rubber inspection

Laser-line 3D profile sensor groove measurement in rubber: resolution budget, scan-rate check, worked example and the decision per case.

LumaSpec Vision11 min read Share on LinkedIn

On this page 9 sections
  1. 01Why a profile sensor is the default for groove geometry
  2. 02How triangulation turns one laser line into width, depth and centerline
    1. Getting a clean line off black rubber
  3. 03Worked example: an 8 mm groove, 5 mm deep, to ±0.05 mm
  4. 04Choosing a profiler: what actually differs
  5. 05On-sensor recipes: changeover without a PC
  6. 06When a profiler is the wrong tool
  7. 07Checks before you order
  8. 08The decision, per case
  9. 09Questions

Groove inspection on rubber is a cross-section problem. Whether the part is a tread blank, an extruded seal or a hose cover, the questions — is the groove present, how wide, how deep, how far off-center — are answered by one slice of geometry, and that slice keeps moving. A 3D profile sensor answers them from a single calibrated laser-line scan, on the line, without touching the rubber. This guide is for the process or quality engineer sizing that sensor: it gives the numbers that decide whether one profiler covers the groove — point pitch against groove width, Z repeatability against depth tolerance, profile rate against line speed — works an 8 mm groove held to ±0.05 mm end to end, and ends with the sensor decision for each case.

Why a profile sensor is the default for groove geometry

A groove lives in a cross-section. Width and depth are distances between edges in one profile; centerline position is the groove's midpoint against a datum in the same profile. A triangulation profile sensor projects a laser line across the part, views that line from a calibrated angle, and returns true height for every point along it. Width, depth and centerline come out as measured geometry, not as greyscale gradients.

A 2D camera cannot do this reliably. Contrast between a groove edge and rubber shifts with shade, dust and glare, and the image shows the opening of the groove, never its depth. A contact gauge measures one point at a time, wears against the rubber and stops the line. A full-surface area 3D camera would work, but it is the slow way to collect one slice.

The decision rule: if the feature you judge is cross-section geometry and the part moves, start with 3D laser profile sensors. The wider system choice is laid out in 3D vs 2D machine vision system: when to choose 3D.

How triangulation turns one laser line into width, depth and centerline

The sensor projects a laser line across the part and photographs that line from an angle. Calibration converts the line's displacement in the image, point by point, into height, so every scan returns a profile of X across the part and Z in height; the part's motion supplies the third axis. Stacking profiles builds a height map of the surface as it passes.

Current smart profilers close the loop on the device. A groove tool locates U-shaped, V-shaped and open grooves, measures width, depth and centerline position, handles several grooves across one profile, and compensates a tilted or crooked part so the numbers do not wander with presentation. You set reference edges and limit bands per product; the sensor judges and sends pass or fail with the measured values.

Getting a clean line off black rubber

Carbon-black rubber absorbs most of the light and scatters the rest, so the line returns dim and wider than on a white target. Three habits fix most cases: keep exposure as short as the profile rate allows and buy brightness with laser power, not gain; tilt the sensor 10–30° so the camera sees diffuse scatter instead of the specular sheet; and on translucent stock prefer a shorter-wavelength blue line, which tunnels less into the bulk than red. The extraction failure modes are worked through in laser line extraction noise: get a clean height line.

Worked example: an 8 mm groove, 5 mm deep, to ±0.05 mm

The case: a tread shoulder groove, 8 mm nominal width (a typical shoulder-groove value; your drawing governs), 5 mm nominal depth, depth tolerance ±0.05 mm. The part runs on an extrusion line at 0.5 m/s (30 m/min, a mid-range rubber line speed), and one cross-section per millimetre of travel is enough to track groove continuity. The candidate is a mid-range triangulation profiler spanning 64 mm across 1,280 points at roughly 100 mm standoff; mid-range triangulation heads commonly quote Z repeatability in the 10–30 µm band at rated standoff. Repeat the four checks below with any candidate's own span, points, repeatability and standoff.

point pitch = profile span ÷ points per profile smallest robust width = 5 × point pitch Z repeatability budget = (2 × depth tolerance) ÷ 3 required profile rate = line speed ÷ profile pitch along travel

Point pitch. 64 mm ÷ 1,280 = 0.05 mm per point across the profile.

Width check. The 8 mm groove spans 8 ÷ 0.05 = 160 points. A groove measured over at least 5 points holds up before any sub-pixel edge interpolation, so the smallest groove this sensor measures well is 5 × 0.05 = 0.25 mm wide. The 8 mm target clears that by a factor of 32; width is not the binding constraint here.

Depth check. A ±0.05 mm tolerance is a 0.10 mm band. Keeping sensor repeatability to a third of the band leaves most of the band to the process: 0.10 ÷ 3 = 0.033 mm, a 33 µm budget. The candidate class quotes 10–30 µm, so it passes — typically near a 2× margin, worst case with about 10% to spare. Inverted, the budget becomes a purchase rule: at a ±0.015 mm tolerance the band is 0.03 mm and the budget 10 µm, which leaves this class behind and calls for a short-standoff precision profiler.

Rate check. 0.5 m/s is 500 mm/s; at a 1 mm profile pitch the line needs 500 ÷ 1 = 500 profiles per second. That caps exposure at 1 ÷ 500 = 2 ms per profile, which is why the laser-power habit above matters on black rubber. Run your own line's numbers:

Line-scan camera line rate
Units
Pixel on the part
– mm
Line rate needed
– kHz
Longest exposure
– µs
Data rate
– MB/s
Lines per metre
–
Encoder setting
–

Square pixels: line rate = speed / pixel on the part; the exposure must end within one line period. Encoder: lines per pulse = (π × roller diameter / pulses per revolution) / pixel on the part.

Find line scan cameras Open the full calculator

Result. One mid-range profiler at roughly 100 mm standoff passes all three checks: 160 points across the groove, repeatability inside a 33 µm budget, and class profile rates that run from a few hundred to several thousand per second, above the 500 required.

Decision. Order the 64 mm / 1,280-point class sensor for this groove. Step up to a short-standoff precision model if the tolerance tightens past ±0.015 mm, and rescale the rate linearly when speed or pitch changes: at 1 m/s and a 0.5 mm pitch, the requirement is 1,000 ÷ 0.5 = 2,000 profiles per second.

Choosing a profiler: what actually differs

Four differences separate the models that look alike in a thumbnail.

Span and points. Pitch is span ÷ points, so compare pitch, not span alone. Cover your widest variant with about 10% margin on each side so a wandering strand still lands inside the profile.

Standoff. Quoted repeatability belongs to the rated standoff; backing the head off degrades both lateral and height resolution. Shorter standoffs read finer but leave less room for clamps and deflect less when something bumps the mount.

Line optics and viewpoint. Red versus blue line, matched to your surface. Deep V-grooves hide wall sections from a single viewpoint; binocular single-line heads such as the MV-DP3062-01D V2.0 observe each profile from paired viewpoints and recover walls a single-camera head loses to occlusion.

Rate and output. Profile rate at your exposure, and an Ethernet link your PLC already speaks. When width exceeds any profiler's span, two options exist: mount profilers side by side and merge them into one synchronized profile, or move to a wide line-scan 2.5D system such as the MV-HLVS-16K100-L60, which covers a 600 mm emitting area at a 120 kHz sampling rate and 16K resolution.

The heads below are Hikrobot 3D laser profile sensors; the comparison table lists what their data sheets print, and the worked-example arithmetic sizes whichever you shortlist at the standoff you will actually mount.

3D sensorsHikrobot MIS-DP3850-01P 3D laser profile sensorMIS-DP3850-01P
Type
Laser profiler
$12,379.00 · quotedView and quote
3D sensorsHikrobot MV-DP4180-03P 3D Laser Profile SensorMV-DP4180-03P
Type
Laser profiler
$12,389.00 · quotedView and quote
Specification MIS-DP3850-01P MV-DP4180-03P
Data points/profile 3200 4080
Reference distance 126 mm 175 mm
Measurement range (Z-axis) 86 mm 100 mm
Measurement range (X-axis) 65 mm @ near end; 85 mm @ reference distance; 105 mm @ far end 75 mm @ near field of view; 100 mm @ reference distance; 125 mm @ far field of view
Resolution (Z-axis) 4.38 ~ 8.33 μm 3.72 μm to 8.40 μm
Repeatability (Z-axis) 1.68 μm @ data that sensor tests gauge block on optical platform1 1.5 μm @ data that sensor tests gauge block on optical platform
Profile data interval 20.4 ~ 35.7 μm 18.5 μm to 32.2 μm
Scan frame rate 3.4 kHz to 46.7 kHz (high frame rate mode) 2.5 kHz (within max. measurement range), max. 49 kHz (in ROI mode)
Data output Original image, profile data, depth image, brightness image Profile data, depth image, brightness image
Trigger mode Software trigger, hardware trigger (differential encoder) Software trigger, hardware trigger (differential encoder)
Laser wavelength 405 nm 650 nm
Laser safety class Class3R Class 3R
Price $12,379.00 · quoted $12,389.00 · quoted

All 2 share: Linearity Z-axis (±% of MR) 0.01; Power supply 24 VDC; Ingress protection IP67; Humidity 20% RH to 85% RH (no condensation). Values as printed in each product's data sheet; open a part for its drawing and full specifications.

On-sensor recipes: changeover without a PC

The integration burden around the sensor is what makes inspection lines complex, so keep the measurement on the device. A smart profiler — the on-device class the smart cameras and vision sensors page covers — runs acquisition, groove tools and the pass/fail decision internally, and hands results to the machine over standard PLC protocols. Recipes are built in a browser: groove type, reference edges, limit bands per product. The sensor stores a job per product, so changeover is a PLC command that switches the active job — seconds, with nothing repositioned or re-taught.

Two habits keep it that simple. Keep every recipe's widest part inside one sensor's span, so no product needs its own mounting position. And develop each recipe offline: record a batch of real profiles once, including the ugly ones, and tune against the recording — the workflow is develop 3D inspection without hardware: the record-once workflow.

When a profiler is the wrong tool

Full-surface shape on a stationary part — a rubber grommet, boot or diaphragm where many features matter at once — belongs to a projected structured-light camera such as the MV-DPS200P-02, which captures the whole part in a single shot.

3D sensorsHikrobot MV-DPS200P-02 Projected structured-light 3D cameraMV-DPS200P-02
Type
Stereo / structured light
$8,829.00 · quotedView and quote

Translucent rubber and silicone let the laser tunnel into the bulk and smear the apparent surface; line confocal sensors measure through such layers instead of bouncing off the first surface, a different sensing principle covered in 3D line confocal sensors: measurement through transparent layers.

If the real question is how much space packed goods occupy, a logistics volume camera answers it, at a depth resolution far too coarse for groove work — a separate buying decision from this one.

Checks before you order

  • Pitch against the narrowest groove. Span ÷ points ≤ narrowest width ÷ 5; the worked example's 0.05 mm pitch makes 0.25 mm the measurable floor.
  • Repeatability at your standoff. Budget is 2 × tolerance ÷ 3 — 33 µm at ±0.05 mm — and a figure that holds only at rated standoff does not transfer to a longer mount.
  • Rate at your exposure. Line speed ÷ pitch, plus margin; 0.5 m/s at 1 mm needs 500 profiles per second before any margin.
  • Mount stiffness. Vibration during the 2 ms exposure smears profiles like real geometry; vibration correction in 3D height scans: a laser profiler guide shows how to tell the two apart.
  • Trueness, not just repeatability. Scan a calibrated step or groove standard on install and compare reported depth; the procedure is in 3D camera infield correction: verify dimension trueness on the line.
  • Dust plan. Carbon-black dust fogs the window and dims the line; put window cleaning or an air purge into the daily round from day one.
  • Laser safety in the cell. Class 2 as delivered, still interlocked with the guarding so an open station means a dark line.

The decision, per case

  • Moving part, cross-section geometry (tread, seal, hose, bead): the mid-range profiler class of the worked example, sized by pitch, repeatability budget and profile rate.
  • Depth tolerance tighter than ±0.015 mm: the 33 µm budget no longer holds — step to a short-standoff precision profiler, since the budget math forces repeatability at or below 10 µm.
  • Full surface, stationary part: projected structured-light area camera.
  • Translucent stock: line confocal, not triangulation.
  • Wider than one span: synchronized side-by-side profilers, or a wide line-scan 2.5D system.
  • Dust, washdown or heat in the same cell: settle the housing and mounting question before the sensor question.

Size your groove against the three checks, then pick the class in the camera finder or step through the configurator below; send the groove drawing and line speed through the contact page if you want the arithmetic checked. The neighboring guides on calibration, recording and line quality live on the 3D vision topic page.

Build the whole station

The configurator matches a camera, lens, light, light controller and cables to each other, step by step, and lists only parts that fit what you already picked.

Start with the camera

Questions

How accurate is a 3D profile sensor on black rubber?

Expect repeatability somewhat worse than the clean-target figure, because carbon-black rubber dims and widens the laser line; the geometry itself stays valid. Short exposure with higher laser power, low gain and a 10–30° sensor tilt to avoid the specular sheet recover most of it, and the tilt matters more on glossy cured surfaces than on matte extrusion stock.

What profile rate does my line need?

Divide line speed by the profile pitch you need along travel. A belt at 1.2 m/s sampled every 2 mm needs 1,200 ÷ 2 = 600 profiles per second; at 0.5 m/s and 1 mm pitch it is 500. Choose a head rated above that figure at your exposure, and trigger profiles from the line encoder so the pitch stays constant when speed drifts.

What does the sensor send to the PLC?

A pass or fail verdict per groove tool, the measured width, depth and centerline values, and a data-valid flag so the machine ignores scans taken across gaps or splices. Transport is a standard industrial Ethernet protocol straight from the sensor link, with no frame grabber or PC in the cell; raw profiles can also stream out for offline analysis.

Profile sensor or area 3D camera for rubber parts?

Profiler when you judge cross-section geometry — groove width, depth, centerline, bead height — especially on a moving part: calibrated height along one line, at high rate. A projected structured-light camera fits full-surface shape of a stationary part in one shot. For packed-goods volume, a logistics volume camera is its own class; none of these matches 0.05 mm depth work except the profiler.

What does groove centerline position tell me?

It is the lateral position of the groove's midpoint relative to a reference edge in the same profile — the part edge, an adjacent groove or a taught datum. Tracking it catches strand wander, off-center extrusion and progressive die wear long before depth drifts out of band, and it comes from the same scan, so it adds no exposure, hardware or cycle time.