Computer Vision
Defects are checked by eye, and some slip through.
- Semiconductor & electronics
- Medical & healthcare
- Biotech & life sciences
- Manufacturing
- Security
- Agriculture & food
Updated
Problem & solution
The problem
Inspection by eye
Results vary by inspector and shift, with no record.
Defects found too late
Caught at end of line or by the customer, when it costs most.
Hours of footage
Finding one event means scrubbing video by hand.
How Collit solves it
Consistent checks
The same rules on every image, every result saved.
Rules plus AI
Classical vision where rules work, trained models where they don't.
Built for your samples
Tuned to your materials, imaging and defect definitions.
Informed decisions
Measurements and pass/fail results your team can act on.
Applications
Semiconductor & electronics
Wafer and PCB defect inspection
Microscope images are segmented to find, measure and classify defects.
Medical & healthcare
Image analysis to support clinical review
Regions of interest are highlighted for review; decisions stay with clinicians.
Biotech & life sciences
Cell and colony counting
Microscopy images are segmented and counted the same way every time.
Manufacturing
Surface defect detection and QA
Cameras at the station flag defects and log every part for traceability.
Security & surveillance
Finding people or vehicles in recordings
Detection runs over footage so you jump straight to the right moment.
Agriculture & food
Grading and sorting produce
Images grade items by size, colour and visible damage before packing.
How it works
- CaptureCamera, microscope or footage
- PrepareClean and normalise
- DetectFind regions of interest
- MeasureSize, count, classify
- ReviewResults your team acts on
Modules
Customizable and extensible: adapted to your workflow, with new modules added as you grow.
Reference alignment
Match reference patterns to align inspection regions.
Regions of interest
Define the areas of an image to inspect.
Thresholding
Separate features from the background.
Image processing
Apply thresholds, morphology, arithmetic and logic.
Blob detection
Find and count regions of interest.
Measurements
Size and position of what was found.
AI inference
Run trained models on your images.
Anomaly detection
Find deviations from learned reference images.
Edge detection
Locate edges for inspection and measurement.
Rules
Classify detections using configurable conditions.
Judgement
Turn inspection results into pass or fail decisions.
Saved workspaces
Reuse inspection settings.
Your module
Built around your workflow.
What you get
- Feasibility review on your sample images
- Inspection or footage-search software
- Reusable inspection settings
- Result export to your systems
- Operator training
What we validate with you
Proven on your own data during the pilot.
- Written defect definitions
- Image quality and lighting
- Performance on your own samples
- Pilot acceptance criteria
Q&As
Do you supply cameras, lighting and automation?
Our focus is the software. A complete machine vision system, with cameras, lighting and automation such as conveyors, robot arms or linear motors, is possible but needs a longer study and a larger budget.
How accurate will it be?
We measure accuracy on your own good and defective samples during the pilot, against criteria agreed with you.
How many images do you need?
A small set of good and defective examples is enough for a feasibility review.
Does our image data leave our premises?
It doesn't have to. Inspection can run on your own computers.
Related solutions & services
AI Agents & Automation
AI agents that automate repetitive workflows, compile reports and answer questions from your data.
- Operations
- Management
- Company knowledge
Custom Software
Web, mobile and desktop apps built around your workflow.
- Field services
- Bookings & appointments
- Customer service
Tell us what's slowing you down
Free 30-minute discovery call. No obligation.