Shoe Upper and Flyknit Upper Inspection
Shoe upper and flyknit upper inspection: for yarn breakage, dropped stitches, stains, holes, color differences and pattern misalignment on shoe uppers/flyknit uppers, AI segmentation identifies low-contrast defects against patterned backgrounds and outputs OK/NG.
Application Overview
Low-contrast defect detection against flyknit upper pattern backgrounds
It inspects yarn breakage, dropped stitches, stains, holes, color difference and pattern misalignment on uppers (including flyknit / knitted uppers); against a patterned background, AI segmentation is used to identify low-contrast defects.
Flyknit / knitted uppers are patterned, textured fabric, and yarn breakage and dropped stitches are often the same color as the knit structure with very low contrast, which makes these difficult projects; manual visual inspection against a patterned background easily leads to fatigue and missed judgement.
This station covers the appearance of shoe upper cut pieces / semi-finished parts, focusing on weaving defects (yarn breakage, dropped stitches, pattern misalignment) plus stains and holes; color difference and pattern matching are also included in the judgement.
Inspection is placed after cutting or before stitching, imaging per piece / per region; defects against a patterned background rely on AI segmentation rather than a fixed threshold, trained specifically for the upper's pattern.
inspection Targets
What This Station Actually Has to Judge
| inspection item | Description |
|---|---|
| Yarn Breakage / Fuzz | Yarn breakage, fuzz |
| Dropped stitches | Dropped needles in weaving, missing pattern positions |
| stain | Oil Stains, Discoloration |
| hole | Knit Holes, Damage |
| color difference | Differs from the standard sample |
| Pattern Registration Misalignment | Pattern position offset |
| skipped stitch | Skipped Stitch at Seam |
Why Inspect
Problems with manual visual inspection at this step
Flyknit uppers have a patterned background, and yarn breakage and dropped stitches are the same color as the weave and low in contrast, so rule-based algorithms are hard to write; AI segmentation is mainly relied on to distinguish defects from normal texture within the pattern.
Manual visual inspection against a patterned background is highly fatiguing, and the escape rate is high in mass production; moreover, the upper is an appearance part, so pattern misalignment also affects finished product consistency.
- Yarn breakage and dropped stitches are the same color as the weave, so contrast is low
- Pattern backgrounds defeat rule-based algorithms
- Manual visual inspection causes fatigue and missed judgement
- Pattern registration errors affect finished product consistency
How Inspect
Stations and Inspection Chain
- 01 Loading and flattening shoe uppers
- 02 Area-scan camera images the pattern
- 03 Use the standard pattern as reference
- 04 AI segmentation of defect regions
- 05 Dropped stitch / yarn breakage classification
- 06 Print registration template matching
- 07 Color difference comparison
- 08 OK/NG output
- 09 Image and data archiving
How to Inspect
An area scan camera images against the patterned background, using the standard pattern as a reference; AI segmentation distinguishes low-contrast defects such as yarn breakage and dropped stitches from the normal weave; this is texture segmentation type imaging.
inspection defect
| Defect Types | Typical Manifestations | Inspection Focus Points |
|---|---|---|
| yarn breakage | Broken yarn, low contrast | AI segmentation |
| Dropped stitches | Missing Pattern Position | Pattern Comparison |
| stain | Oil Stains, Discoloration | Diffused Light |
| hole | Knit holes | Backlight |
| color difference | Local Discoloration | Color Patch Comparison |
| Pattern Registration Misalignment | Pattern Offset | template matching |
| skipped stitch | Skipped Stitch at Seam | Line Feature Recognition |
Applicable equipment
Application Industry
Common Question
What Makes Shoe Upper Inspection Difficult?
Can It Replace Manual Visual Inspection?
What Inspection Accuracy Can Be Achieved?
Can It Be Used Universally Across Different Patterns?
How are dropped stitches identified?
Are results traceable?
Does Switching Production Lines Require Retuning?
Submit sample testing
Typical OK/NG shoe upper / flyknit upper samples can be sent for measured testing: the segmentation model is trained and the imaging solution determined according to the pattern.
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 →
Can This Application Be Done? Sending a Sample for Testing Is the Most Direct Answer
Incoming material, part orientation and cycle time vary widely from factory to factory. Send us real samples and we will run imaging and judgement validation against your production line conditions, then give you a configuration proposal you can actually implement.