Advanced Materials AI Visual Inspection Equipment
Advanced materials AI visual inspection equipment: imaging is configured to the material's optical properties and then AI acceptance criteria are configured, targeting surface defect inspection of advanced materials such as carbon fiber prepreg, optical film, foam and tape; internal properties require non-destructive testing.
Products Overview
An advanced materials inspection platform that configures imaging to the material's optical properties and then applies AI-based criteria
Advanced materials AI visual inspection equipment serves carbon fiber prepreg, fiberglass, aramid, various high-performance films, foam and tape, configuring imaging to each material's optical characteristics and then applying AI criteria to inspect surface defects; bulk properties such as internal delamination and porosity require ultrasonic / X-ray or physical-chemical testing.
"Advanced materials" is not one product but a collective term for a class of advanced materials — carbon fiber prepreg, fiberglass, aramid, various high-performance films, foam and tape. What they have in common is: New materials, many specifications, and defect standards are often still being established. A single fixed hardware setup cannot cover everything; inspection must start from "material optical properties".
The core logic of this type of inspection is "Configure imaging according to the material's optical properties, then set the AI acceptance criteria". Fabric and prepreg require suppressing fiber specular gloss; transparent film requires transmitted light to find holes; foam requires distinguishing foam texture from genuine defects; tape requires accounting for release paper pattern interference. Different materials require different imaging solutions — this is the essence of "advanced materials inspection equipment".
The boundary must be clearly defined: internal delamination, porosity, resin content, and similar properties fall under Bulk material properties, they cannot be judged by surface vision and require ultrasound, X-ray or physical and chemical testing. Making this boundary clear is what makes the solution defensible, and it avoids disputes at the acceptance stage about "why internal defects were not detected".
- Positioning: a class of inspection platform for advanced materials rather than a single product
- Core logic: imaging matched to the material's optical properties + AI acceptance criteria
- Material differences: fabric / film / foam / tape each require different imaging methods
- Boundary: internal defects require non-destructive / physical and chemical testing; surface vision only covers the surface and contour
Core Functions
What the Equipment Can Do and How Far It Can Go
Imaging Configured by Material
Different optical properties call for different illumination and camera setups, not fixed hardware.
Reflection Suppression for Fabric / Prepreg
Coaxial light / polarization suppresses the specular sheen of fibers so highlights do not drown out defects.
Transmissive Inspection of Transparent Film
For PET / PI / OCA, transmitted light checks for holes and break points.
Foam Texture Differentiation
The algorithm distinguishes normal foam texture from true defects to avoid false calls.
Handling Interference from Tape and Release Paper
Recognize the release paper texture and avoid treating the texture as a defect.
Well-Defined Boundaries
Internal defects are referred to non-destructive / physical and chemical testing, with no exaggerated promises.
inspection Object
Fiber and Fabric Category
Carbon fiber prepreg, fiberglass and aramid, with the focus on reflection suppression and surface defects
High-performance Films
PET / PI / OCA optical film, with the focus on transmitted-light checks for holes and break points
Foam and Cushioning Category
Foam / EVA / IXPE, with the focus on separating foaming texture from actual defects
Tape and Functional Materials
Die-cut tape and conductive fabric, with the focus on release paper texture interference and foreign matter
inspection defect
| Defect Types | Typical Manifestations | Inspection Focus Points |
|---|---|---|
| Fiber specular gloss interference | Specular highlights on fabric / prepreg surfaces mask defects | Requires reflection-suppressing illumination (coaxial light / polarization) |
| Transparent film holes / pinholes | Pinholes and holes in PET / PI / OCA film | Transmitted light gives the highest contrast for holes |
| Foam texture causes false calls | Normal pores are mistaken for defects | Must distinguish foam texture from real defects |
| Interference from tape and release paper patterns | Release paper texture misjudged as a defect | Requires a dedicated algorithm to distinguish |
| Surface foreign matter / inclusions | Fibers, particles and gel particles mixed in | Varied shapes, suited to deep learning |
| Color difference / uneven coating | Inconsistent color or coating within the same material batch | Requires stable light source and datum |
| Delamination / bubbles (visible on surface) | Interlayer separation, surface blistering | Requires multiple angles or specific illumination |
| Internal Delamination / Porosity / Resin Content | Bulk material property defects | Beyond the scope of surface vision; requires ultrasonic / X-ray / physical and chemical testing |
Working Principles
- 01 Material loading / unwinding
- 02 Determine the material type and recipe
- 03 Optional imaging solution
- 04 Dedicated illumination imaging
- 05 surface defect inspection
- 06 Separating texture from interference
- 07 Result Composition OK / NG
- 08 NG Marking and Recording
- 09 Data archiving / guidance for non-destructive testing
Relationship Between Imaging and Judgement
The success of advanced materials inspection depends on "whether the imaging solution matches the optical properties of the material". With the same hardware, if the lighting is not matched to the material, fabric will be drowned out by reflection, transparent film will be overexposed, and foam will have its pores treated as defects. The core of such equipment is therefore not fixed hardware but the ability to map "material -> optics -> acceptance criteria".
Vision System
How Cameras, Lenses, Light Sources and Controllers Are Configured
industrial camera
Select area-scan or line-scan by material area and minimum defect size; continuous web feeding usually uses line scan
- Area-scan camera: fixed-shot, single-piece, intermittent feeding
- Line Scan Camera: Continuous Feed, Wide Format
- Back-calculate the pixel size from the smallest resolvable defect
lens
Determines field of view, distortion and depth of field; use telecentric lenses for dimensional measurement
- Standard industrial lens: appearance inspection
- Low-distortion lens: edge judgement on large areas
- Telecentric lens: hole diameter, contour, dimensional measurement
Light Source and Illumination
Whether defects in advanced materials can be captured depends largely on whether the illumination matches the material's optical properties
- Diffused light: color difference, stains
- Low angle / stripe light: scratch, indentation
- Transmitted light: holes and short shots in transparent film
- Coaxial / polarization: suppresses reflections from fabric and prepreg
Controllers and Industrial PCs
The platform for running algorithms, outputting results and interlocking with the production line
- Industrial Computer or Vision Controller
- Light source controller dimming synchronized with strobing
- Interacts with the PLC to complete interlocking, alarming, and rejection
AI algorithm
Choose the Algorithm by Defect Form, Not by Complexity
Template Matching and Positioning
Locate first, then judge, to establish a stable coordinate system
- Shape / grayscale matching: suited to fixed stations
- Region segmentation after alignment avoids false calls caused by drift
Blob and Morphological Analysis
Defects with obvious gray-scale / color differences, such as stains, holes and foreign matter
- Threshold segmentation → connected-component statistics
- Depends on stable illumination
Edge and Contour Measurement
Judgments for dimensions, contours, hole positions, and spacing
- Sub-pixel edge extraction
- Line / circle / arc fitting and calculation
Deep Learning (Classification / Detection / Segmentation)
Defects with varied shapes that are hard to describe by rules, such as foam texture anomalies and foreign matter
- Requires OK / NG Sample Training
- Supports pixel-level segmentation to output defect size and position
- Sample coverage sets the upper limit
Automatic Alarm and rejection
How inspection results act on the production line
Inspection results are not only shown on the screen. Material judged NG must have its position marked, be rejected or sorted by interlocking, and have its images and judgement results archived for re-judgement and traceability.
The alarm method is set according to on-site practice: audible and visual alarm, pop-up window, PLC set bit, or all three at once. Critical defects and general defects can be handled differently — the former stops the machine and alarms, while the latter is only marked.
- 01 Defect judgement result generation
- 02 NG marking and alarm trigger
- 03 Image and data archiving
- 04 Production line interlocking rejection / sorting
data traceability
Records, Queries and Quality Closed Loop
Per-Piece Records
The judgement result, defect type, position, and timestamp of every part are stored in the database
- Search by time / batch / defect type
- NG image retention for re-judgement
Batch and Recipe
Different materials use different recipes, called up at changeover to reduce manual parameter tuning
- Recipes store the inspection region, thresholds and algorithm parameters
Production Line Data Integration
Exchange inspection data with the MES / host system
- Output pass rate and defect distribution
- Interface method is subject to site conditions
equipment configuration
Optional Configuration Items and Selection Logic
| Configuration Item | Description |
|---|---|
| Equipment Type | Customized to the material area and loading method (roll line / sheet station) |
| Inspection Method | Industrial camera + material-specific light source + AI algorithm |
| camera | Selected by field of view and the smallest resolvable defect (area scan / line scan) |
| algorithm | Deep Learning Classification / Detection / Segmentation + Rule-Based Algorithm Combination |
| Inspection Speed | To be added |
| inspection accuracy | To be added |
| communication method | I/O · TCP · RS485 · Modbus · S7 · Profinet |
| Power Supply | Subject to final equipment confirmation |
| Protection and Structure | Customized to site conditions |
Application Industry
Which Industry This Equipment Is Usually Installed In
Applicable Materials
Detectable Material Types
Common Question
What Is the Inspection Accuracy of the Advanced Materials Inspection Equipment?
Why can't one hardware setup inspect all advanced materials?
Can holes in transparent film be detected?
Will normal foam pores cause false calls?
Can internal delamination and porosity be detected?
What materials can be inspected?
What Information Do You Need to Provide?
Submit sample testing
Advanced materials vary widely, so illumination must be configured on actual material samples. Please provide OK / NG samples of each material; we will validate the imaging and judgement solution separately for each according to its optical characteristics, and define the coverage limits of surface vision.
Submitted successfully
We have received your sample testing request. A solution engineer will contact you within 1 business day via contact you.
Need to assess imaging conditions, defect criteria and cycle time item by item? Go to the Full Requirement Assessment →
Send Us Your Workpiece and We Will Show You the Measured Results
Equipment configuration varies with the inspection object, field of view and cycle time. Provide OK and NG samples and we will run actual imaging and judgement tests and recommend the corresponding model and configuration.