Specify camera resolution, lens and lighting geometry for a defined defect and part geometry.
Quality & Productivity
Machine Vision Systems for Automated Visual Defect Detection
Shows quality and manufacturing engineers how to specify, deploy and tune machine vision inspection systems, from camera and lighting selection to defect model training and line integration.
🗓️Duration5 training days
📚Content4 modules · 8 sessions
🎓On completionAccredited attendance certificate
About the programme
Course Overview
Manual visual inspection is slow, inconsistent between shifts and vulnerable to fatigue, yet many machine vision projects fail because camera, lighting or model choices do not match the defect they are meant to catch. This course gives quality and manufacturing engineers a structured method for specifying, building and sustaining automated visual inspection. Participants learn to match defect types to camera and lighting technology, apply classical image processing and deep learning classifiers, and set false accept and false reject thresholds that protect both quality and line throughput. Sessions combine specification exercises, dataset labelling practice and integration case studies covering PLC and reject mechanism connections. By the end, participants can plan a pilot vision installation, validate it against human inspectors, monitor it for performance drift, and build the business case needed to fund a wider rollout across the plant.
Expected Learning Outcomes
Compare rule-based image processing with deep learning classifiers for different defect types.
Build and label a training image dataset that represents real production defect variation.
Set false accept and false reject thresholds that balance quality risk against line throughput.
Integrate a vision system with PLCs, robots and reject mechanisms on an existing production line.
Monitor vision system performance over time and detect drift caused by lighting or product changes.
Justify a machine vision investment with a business case covering cost, cycle time and defect escape rate.
Who Should Attend
Quality engineers specifying automated inspection equipment
Manufacturing engineers integrating vision systems into production lines
Process engineers responsible for reducing defect escapes
Automation and controls engineers supporting vision deployments
Quality managers evaluating machine vision investment cases
Continuous improvement leads replacing manual visual inspection stations
Course Modules
Select any module to see its sessions and points.
01Defect Types and Vision System Fundamentals
2 sessions · 8 points
Session 1Matching Inspection Tasks to Vision Technology
- Map common defect types such as scratches, dents, colour variation and missing components to suitable vision techniques.
- Explain the roles of area-scan and line-scan cameras in static and moving part inspection.
- Select lens focal length and working distance to achieve the resolution needed to resolve target defects.
- Choose lighting geometry, including backlighting, dark-field and diffuse illumination, for specific surface finishes.
Session 2Classical Image Processing Techniques
- Apply thresholding, edge detection and blob analysis to isolate defects from background noise.
- Use pattern matching and template subtraction to detect missing or misplaced components.
- Calibrate pixel-to-millimetre ratios for accurate dimensional measurement from images.
- Tune image processing parameters to reduce sensitivity to normal part-to-part variation.
02Deep Learning for Defect Classification
2 sessions · 8 points
Session 1Building a Defect Image Dataset
- Plan image capture campaigns that cover defect severity, orientation and lighting variation.
- Label defect and non-defect images consistently using an agreed severity taxonomy.
- Balance datasets across defect classes to avoid bias towards common defect types.
- Split data into training, validation and test sets that reflect real production ratios.
Session 2Training and Validating Classification Models
- Train convolutional neural network classifiers using transfer learning from pretrained models.
- Evaluate model performance using precision, recall and confusion matrices by defect class.
- Set confidence thresholds that separate automatic accept, automatic reject and human review.
- Retrain models periodically as new defect types or product variants appear on the line.
03Integration with Production Systems
2 sessions · 8 points
Session 1Connecting Vision Systems to Line Controls
- Interface vision system outputs with PLCs and SCADA systems for real-time reject decisions.
- Synchronise camera triggering with part presence sensors and conveyor encoders.
- Configure reject mechanisms and diverter logic to remove defective parts without stopping the line.
- Design fail-safe behaviour for camera faults, lighting failure or network interruption.
Session 2Line Trials and Throughput Validation
- Run pilot trials comparing vision system decisions against experienced human inspectors.
- Measure cycle time impact and adjust inspection station layout to protect line throughput.
- Set escalation rules for parts flagged as uncertain rather than automatically rejected.
- Sign off the system against agreed false accept and false reject rate targets before go-live.
04Sustaining Performance and Building the Business Case
2 sessions · 8 points
Session 1Monitoring Drift and Maintaining Accuracy
- Track false reject and false accept rates over time using a control chart by shift and product.
- Detect lighting degradation, lens contamination and camera drift through scheduled checks.
- Establish a change control process for retraining models after product or packaging changes.
- Maintain a defect image library that documents new failure modes as they are discovered.
Session 2Building the Investment Case and Rollout Plan
- Quantify the cost of current defect escapes, rework and manual inspection labour.
- Compare vision system total cost of ownership against manual inspection over a multi-year horizon.
- Sequence a rollout plan from a single-station pilot to multi-line deployment.
- Present a business case to plant leadership covering payback period and quality risk reduction.
What the participant receives
4 course modules
A structured syllabus
8 training sessions
across 5 days
32 detailed points
Applied, detailed content
Accredited attendance certificate
On completing the programme
Complete your registration
We will contact you within one business day to confirm.
Ready to start?
Reserve your seat and start building the skill.