98% Accuracy Explained: Why ANPR Sometimes Gets It Wrong

Executive Summary

ANPR (Automatic Number Plate Recognition) cameras achieve 90 to 98% accuracy under optimal conditions, but that 2 to 10% error rate translates to millions of misreads annually across the UK’s 100+ million daily reads. While modern ANPR technology is highly sophisticated, it is not infallible. Environmental conditions, plate quality, camera angles, vehicle speed, and technical limitations all contribute to read failures and misidentifications that can result in incorrect fines, unnecessary police stops, or missed enforcement.

The 98% accuracy figure represents ideal laboratory conditions with perfectly clean, BS AU 145e:2018 compliant plates, optimal camera positioning, and moderate speeds. Real-world performance drops to 85 to 93% when you factor in rain, dirt, fading, illegal fonts, extreme angles, and high-speed scenarios. Understanding these limitations helps drivers keep their plates readable while providing context for challenging erroneous penalties.

As a DVLA-registered manufacturer (RNPS ID: 73132), we produce plates specifically engineered to maximise ANPR readability through BS AU 145e certified materials, precise Charles Wright font spacing, and NIR-compatible reflectivity. This guide examines why ANPR errors occur, what factors reduce accuracy, how to protect yourself from misreads, and when to appeal penalties caused by technology failures.

Understanding the 98% Accuracy Claim

What the Statistics Actually Mean

When ANPR manufacturers and police forces cite “98% accuracy”, this figure requires careful interpretation.

Laboratory vs Real-World Performance:

Controlled Testing Conditions (95 to 98% accuracy):

• Clean, new BS AU 145e compliant plates
• Optimal camera angle (0 to 15 degrees)
• Perfect lighting conditions
• Vehicle speeds under 40 mph
• Dry weather, clear visibility
• Professional camera calibration
• Single plate per vehicle clearly visible

Real-World Operating Conditions (85 to 93% accuracy):

• Mixed plate conditions (clean to heavily soiled)
• Variable camera angles (0 to 45 degrees)
• Changing light (dawn, dusk, night, glare)
• Speeds from 0 to over 100 mph
• All weather conditions (rain, snow, fog)
• Multiple vehicles in frame
• Partial plate obscurement

For a complete explanation of the technology, see our guide on how ANPR cameras work.

The Mathematics of Error Rates

UK ANPR Scale:

• 100+ million reads per day
• 98% accuracy = 2 million errors daily
• 95% accuracy = 5 million errors daily
• 90% accuracy = 10 million errors daily

Annual Impact:

At 95% accuracy (a realistic average):

• Correct reads: 34.675 billion annually
• Incorrect reads: 1.825 billion annually
• Misreads requiring human review: Hundreds of millions

Error Types:

False Negatives (Missed Reads), 60 to 70% of errors:

• Plate not detected at all
• System logs “no read”
• Vehicle passes undetected
• Lower risk, as the violation is not caught

False Positives (Misreads), 30 to 40% of errors:

• Plate read incorrectly
• AB51 CDE read as AB5I CDE
• Higher risk, as the wrong person is penalised
• Requires human verification

Partial Reads, 5 to 10% of errors:

• Some characters detected, others unclear
• System flags for manual review
• Confidence threshold not met

Confidence Thresholds and Decision-Making

How ANPR Systems Decide:

ANPR software assigns a confidence score (0 to 100%) to each read.

High Confidence (90 to 100%):

• Automated enforcement proceeds
• Database checks initiated
• Alerts triggered if a match is found
• Minimal human oversight

Medium Confidence (70 to 89%):

• Read logged but flagged
• May require human verification
• Used for intelligence gathering
• Not typically used for penalties

Low Confidence (Below 70%):

• Read rejected or heavily flagged
• Requires mandatory human review
• Often discarded as unreliable
• System may request additional frames

Threshold Settings:

Different applications use different thresholds:

• Speed enforcement: 90%+ minimum
• Stolen vehicle alerts: 85%+ (balancing speed vs accuracy)
• Congestion charge: 80%+ (human review layer)
• Intelligence gathering: 70%+ (pattern analysis)
• Parking enforcement: 85%+ (automated plus appeals process)

Technical Limitations: Why Cameras Fail

Optical Character Recognition Challenges

The Core Problem:

OCR (Optical Character Recognition) must convert pixel patterns into alphanumeric characters, but several factors complicate this.

1. Character Ambiguity:

Certain characters look similar even to human eyes:

Character Pair	Confusion Rate	Reason
8 / B	3 to 5%	Similar loop structure
0 / O	4 to 6%	Identical oval shape
1 / I / l	5 to 8%	Vertical line variation
2 / Z	2 to 3%	Angular similarity
5 / S	2 to 4%	Curved shape overlap
Q / O / 0	3 to 5%	Circular confusion

Mitigation Strategies:

• Context validation (is this a real UK registration?)
• Format checking (does it match DVLA patterns?)
• Multiple frame comparison
• Dictionary lookup against known registrations

2. Font Variations:

Legal Font (Charles Wright):

• Standardised since 2001
• Optimised for ANPR readability
• Specific character widths and spacing
• 95%+ accuracy when clean and compliant

Illegal Fonts:

• “VIP” style (elongated, condensed)
• “Carbon” style (italicised, slanted)
• Custom designs (decorative elements)
• Accuracy drops to 40 to 70%
• Character spacing violations
• Non-standard stroke thickness

Our guide on the Charles Wright font and why it’s mandatory covers the font requirements in full.

3. Resolution and Pixel Density:

Minimum Requirements:

• Plate region needs 100 to 200 pixels width
• Each character needs 15 to 25 pixels height
• Below this threshold, accuracy collapses

Real-World Issues:

• Distant cameras (plates appear small)
• Wide-angle lenses (plates occupy few pixels)
• Low-resolution cameras (older systems)
• Digital zoom degradation

Camera Hardware Limitations

Shutter Speed Constraints:

Motion Blur:

• At 60 mph, a vehicle travels 27 metres per second
• Shutter speed must be 1/1000s or faster
• Slower shutter = motion blur
• Blur reduces character edge definition
• Accuracy drops 10 to 20% with blur

Dynamic Range Issues:

High Contrast Scenarios:

• Bright plate, dark vehicle
• Headlights creating glare
• Shadows obscuring characters
• The camera must capture detail in both bright and dark areas
• Limited dynamic range loses detail
• HDR (High Dynamic Range) helps but adds processing time

Infrared Illumination Problems:

NIR (Near-Infrared) Challenges:

• LED arrays degrade over time
• Uneven illumination creates hotspots
• Range limitations (effective to around 30 metres)
• Weather absorption (rain/fog scatter IR)
• Reflective plate angle dependency
• Too much IR = overexposure
• Too little IR = underexposure

Processing Speed vs Accuracy Trade-offs

Real-Time Constraints:

ANPR systems must process images in 300 to 500 milliseconds, which forces compromises.

Speed Optimisations:

• Reduced image resolution for faster processing
• Simplified algorithms (less thorough)
• Limited character segmentation attempts
• Fewer validation checks
• Single-frame processing (no averaging)

Accuracy Optimisations:

• Higher resolution processing
• Multiple algorithm passes
• Extensive character validation
• Multi-frame comparison
• Takes 1 to 2 seconds (not suitable for high-speed)

The Balance:

Most systems prioritise speed for enforcement (catching violations) over perfect accuracy, relying on human review to catch errors before penalties are issued.

Environmental Factors That Reduce Accuracy

Weather Conditions

Rain:

Impact on Accuracy:

• Light rain: 90 to 94% accuracy (5 to 8% reduction)
• Moderate rain: 85 to 90% accuracy (10 to 13% reduction)
• Heavy rain: 75 to 85% accuracy (15 to 23% reduction)

Mechanisms:

• Water droplets on the lens scatter light
• Rain on the plate surface creates reflections
• Reduced infrared penetration
• Spray from tyres obscures plates
• Windshield wipers create motion blur

Mitigation:

• Hydrophobic lens coatings
• Higher IR intensity
• Faster shutter speeds
• Multiple camera angles

Snow and Ice:

Impact on Accuracy:

• Light snow: 80 to 88% accuracy
• Heavy snow: 60 to 75% accuracy
• Ice-covered plates: 40 to 60% accuracy

Problems:

• Snow accumulation on the plate
• Ice obscuring characters
• Reduced reflectivity
• White-on-white contrast loss
• Camera lens icing

Seasonal Challenge:

Winter months see 10 to 15% higher error rates across the UK ANPR network.

Fog and Mist:

Impact on Accuracy:

• Light fog: 85 to 92% accuracy
• Dense fog: 70 to 80% accuracy

Mechanisms:

• Water droplets scatter infrared light
• Reduced effective range
• Lower contrast overall
• IR absorption by moisture

Fog Penetration:

• 850nm IR penetrates better than visible light
• 940nm IR penetrates fog even better
• But reflectivity is also reduced

Lighting Conditions

Direct Sunlight:

Glare Problems:

• Reflective plates create intense glare
• Characters lost in overexposure
• Accuracy drops to 85 to 90%
• Polarising filters help but aren’t perfect

Solutions:

• HDR imaging
• Multiple exposures
• Polarising filters
• Optimal camera positioning

Night-Time Operation:

Challenges:

• Reliance on IR illumination only
• Headlight interference
• Street lighting variations
• Plate angle critical for IR reflection

Performance:

• Dry night: 93 to 96% accuracy (only 2 to 3% reduction)
• Wet night: 85 to 90% accuracy (8 to 10% reduction)
• IR technology actually performs well at night

Dawn and Dusk:

Transitional Lighting:

• Mixed visible and IR light
• Auto-exposure confusion
• Rapidly changing conditions
• Accuracy: 88 to 93%

Worst Conditions:

• Low sun angle creating glare
• Backlit vehicles
• Silhouette effect

Road and Traffic Conditions

Vehicle Speed:

Speed vs Accuracy:

Speed	Accuracy	Reason
0 to 20 mph	96 to 98%	Minimal motion blur
20 to 40 mph	94 to 96%	Optimal range
40 to 60 mph	91 to 94%	Standard operation
60 to 80 mph	87 to 91%	High speed challenges
80 to 100 mph	80 to 87%	Extreme speed
100+ mph	70 to 80%	Severe motion blur

Motorway ANPR:

• Average speed cameras accept lower accuracy
• Multiple reads over distance compensate
• Human verification essential

Multiple Vehicles:

Occlusion Problems:

• Vehicles passing side-by-side
• Plates partially hidden
• System confusion (which plate?)
• Accuracy drops 5 to 10%

Solutions:

• Multiple cameras
• Lane separation
• Temporal separation (time-based)

Road Debris and Dirt:

Plate Contamination:

• Mud splatter (common in winter)
• Road salt buildup
• Insect impacts
• Dust accumulation
• Accuracy reduction: 10 to 25%

Prevention:

• Regular plate cleaning (the driver’s responsibility)
• BS AU 145e smooth surface (easier to clean)
• Hydrophobic coatings (experimental)

Plate-Related Issues: The Driver’s Responsibility

Non-Compliant Plates

Illegal Modifications:

Tinted Covers:

• Acrylic or plastic covers with tint
• Reduces reflectivity by 40 to 60%
• ANPR accuracy: 50 to 70%
• Illegal under the Road Vehicles Regulations 2001
• Fine: Up to £1,000
• MOT failure guaranteed

Reflective Sprays:

• “Ghost plate” sprays
• Claims of “ANPR-proof” plates
• Actually reduce readability to 40 to 60%
• Illegal and easily detected
• Police stop likely

For the full legal picture, see our guide on whether ghost plates are legal.

Incorrect Fonts:

Common Illegal Fonts:

• VIP style (condensed, elongated)
• Carbon fibre style (decorative)
• Italicised/slanted fonts
• 3D characters obscuring spacing

Impact:

• Character spacing violations
• Non-standard stroke thickness
• ANPR accuracy: 40 to 70%
• Increased misread rate
• Police attention

Plate Condition and Maintenance

Fading and Wear:

UV Degradation:

• Reflective sheeting fades over time
• Typical lifespan: 5 to 7 years
• After 7+ years: 20 to 30% reflectivity loss
• ANPR accuracy drops to 80 to 88%

Signs of Fading:

• Yellowing of the white background
• Reduced contrast
• Characters appear washed out
• Gloss finish lost

Solution:

• Replace plates every 5 to 7 years
• Check reflectivity annually
• Compare to a new plate if unsure

Physical Damage:

Common Damage Types:

• Cracks in the acrylic substrate
• Peeling reflective sheeting
• Bent or warped plates
• Missing characters
• Scratch marks

Impact on ANPR:

• Cracks: 10 to 20% accuracy reduction
• Peeling: 20 to 40% accuracy reduction
• Warping: Angle issues, 15 to 25% reduction
• Missing characters: Complete read failure

Legal Requirement:

Damaged plates must be replaced immediately, as driving with illegible plates is an offence. If you’re not sure what tips a plate over into non-compliance, our guide on what makes a number plate illegal explains where the line sits.

Dirt and Obscurement:

Common Problems:

• Mud buildup (especially rear plates)
• Snow and ice coverage
• Road salt accumulation
• Grime and dust layers
• Insect debris

ANPR Impact:

• Light dirt: 5 to 10% accuracy reduction
• Moderate dirt: 15 to 25% reduction
• Heavy dirt: 40 to 70% reduction
• Complete obscurement: 0% (no read)

Driver Responsibility:

• Clean plates weekly in winter
• Clean monthly in summer
• Check before long journeys
• Remove snow/ice before driving

Legal Position:

Regulation 11 of the Road Vehicles (Display of Registration Marks) Regulations 2001 requires plates to be “clear and legible”, so dirt is not a defence.

Plate Positioning and Mounting

Angle Issues:

Optimal Positioning:

• Perpendicular to the ground (0 degrees)
• Height: 400 to 1000mm from the road
• Facing directly forward/backward

Common Problems:

• Lowered vehicles (plate angles upward)
• Raised 4x4s (plate angles downward)
• Tow bars obscuring plates
• Bike racks blocking plates
• Damaged mounting (plate swings)

ANPR Impact:

• 0 to 15 degrees: 95 to 98% accuracy
• 15 to 30 degrees: 90 to 95% accuracy
• 30 to 45 degrees: 80 to 90% accuracy
• 45+ degrees: 60 to 80% accuracy

Solutions:

• Adjust mounting brackets
• Use adjustable plate holders
• Ensure secure fixing
• Regular position checks

System Design and Infrastructure Issues

Camera Placement and Coverage

Suboptimal Positioning:

Common Problems:

• Camera too high (steep angle)
• Camera too far (small plate image)
• Wrong lane coverage
• Blind spots
• Sun glare at certain times

Impact:

• Reduced accuracy 5 to 15%
• Inconsistent reads
• Coverage gaps

Budget Constraints:

• Ideal placement costs more
• Compromises made for cost savings
• Legacy installations not optimised

Calibration and Maintenance

Camera Calibration:

Requirements:

• Focus adjustment
• Exposure settings
• IR intensity calibration
• Angle verification
• Regular testing

Reality:

• Calibration drifts over time
• Maintenance schedules vary
• Budget limitations
• Weather damage requires recalibration

Impact of Poor Calibration:

• Focus issues: 10 to 20% accuracy loss
• Exposure problems: 15 to 25% loss
• IR misalignment: 20 to 30% loss

Maintenance Challenges:

• 10,000+ cameras across the UK
• Access difficulties (motorway gantries)
• Weather windows for work
• Cost of specialist technicians

Software and Algorithm Limitations

Algorithm Updates:

Challenge:

• ANPR software evolves
• Different cameras run different versions
• Inconsistent performance
• Update rollout takes months or years

Legacy Systems:

• Older cameras (10+ years)
• Limited processing power
• Outdated algorithms
• Accuracy 5 to 10% lower than modern systems

Integration Issues:

• Multiple vendors
• Different database systems
• Compatibility problems
• Data transfer delays

The Human Verification Layer

Why Human Review Matters

Error Detection:

Despite 90 to 98% automated accuracy, human verification catches many errors.

Human Review Catches:

• Obvious misreads (clearly wrong characters)
• Partial plates (missing characters)
• Poor image quality
• Context errors (impossible registrations)
• Low confidence reads

Human Accuracy:

• Trained operators: 99%+ accuracy
• Takes 3 to 5 seconds per image
• Can use enhanced tools (zoom, contrast)
• Context understanding (UK registration formats)

When Human Review Occurs

Mandatory Review:

Speed Enforcement:

• All average speed camera violations
• Before a NIP (Notice of Intended Prosecution) is issued
• Operator confirms plate readability
• Verifies vehicle identity

Police Alerts:

• Stolen vehicle matches
• Wanted person vehicles
• Officer reviews before action
• Assesses risk level

Low Confidence Reads:

• System flags below threshold
• Automatic routing to review queue
• Operator makes the final determination

Automated Processing:

No Human Review:

• Routine reads (tax/MOT/insurance checks)
• Congestion charge (high confidence only)
• Parking enforcement (automated plus appeals)
• ULEZ compliance (automated)

Appeal Process:

• Driver can challenge the penalty
• Human review triggered by the appeal
• Evidence examined
• Decision may be overturned

Limitations of Human Review

Volume Challenges:

Scale:

• 100+ million reads daily
• Even a 5% review rate = 5 million images/day
• Requires hundreds of operators
• Cost constraints limit review percentage

Reality:

• Only high-stakes enforcement gets full review
• Low-value penalties are often fully automated
• The appeals process is the safety net

Operator Fatigue:

Human Factors:

• Reviewing hundreds of images per shift
• Attention lapses inevitable
• Error rate increases with fatigue
• Quality varies by operator

Safeguards:

• Regular breaks
• Quality audits
• Performance monitoring
• Double-check for serious cases

Real-World Error Examples

Case Study 1: Character Confusion

Scenario: Vehicle AB51 CDE passes an ANPR camera.

What Happened:

• Camera captured the image in heavy rain
• Water droplets on the lens
• “8” character partially obscured
• OCR read it as “B” instead of “8”
• System read: AB5I CDE (also misread “1” as “I”)

Consequence:

• Database check performed on AB5I CDE
• A different vehicle registered to that number
• No alerts triggered
• The actual vehicle (stolen) was not detected
• Missed enforcement opportunity

Root Cause:

• Weather conditions
• Character ambiguity (8/B similarity)
• No human review (routine read)
• Confidence score: 87% (above threshold but wrong)

Prevention:

• Better lens protection
• Higher confidence threshold
• Multi-frame comparison
• Context validation

Case Study 2: Dirt-Induced Error

Scenario: Parking enforcement in London.

What Happened:

• Vehicle AB12 CDE parked in a restricted zone
• Rear plate covered in mud (winter conditions)
• ANPR camera captured the image
• System read: AB12 C0E (the final “D” obscured by dirt, read as “0”)
• Automated PCN issued to AB12 C0E

Consequence:

• The wrong vehicle owner received the penalty
• The registered keeper of AB12 C0E appealed
• Provided evidence (V5C, plate photos)
• PCN cancelled after a 3-week investigation
• The actual violator was never penalised

Root Cause:

• Plate not cleaned (driver negligence)
• No human verification (fully automated)
• Dirt obscured the character
• Similar character confusion (D/0)

Prevention:

• Driver responsibility (clean plates)
• Human review threshold adjustment
• Image quality checks
• Appeal process (it worked but was inefficient)

Case Study 3: Angle-Related Misread

Scenario: Motorway speed enforcement.

What Happened:

• Vehicle on the M6 at 75 mph
• Camera positioned at a 35-degree angle
• Plate: XY12 ABC
• Perspective distortion from the angle
• System read: XYI2 ABC (“1” read as “I”)
• Confidence: 82%

Consequence:

• Speed violation detected (82 mph in a 70 zone)
• NIP issued to XYI2 ABC
• The wrong vehicle owner received the notice
• Appeal submitted with evidence
• Investigation revealed the angle issue
• Penalty cancelled

Root Cause:

• Camera positioning (not optimal)
• High speed (motion blur)
• Character ambiguity (1/I)
• Angle exceeded the recommended 30 degrees

Prevention:

• Better camera positioning
• Angle limits enforced
• Higher confidence threshold for high speed
• Mandatory human review

Practical Takeaways

How to Maximise ANPR Readability

✅ DO:

Maintain BS AU 145e Compliance:

• Use only DVLA-registered suppliers (RNPS ID displayed)
• Ensure plates meet the current standard
• Verify the Charles Wright font
• Check character spacing (11mm)

Keep Plates Clean:

• Wash plates weekly in winter
• Monthly cleaning in summer
• Remove snow/ice before driving
• Check for mud after off-road use
• Use a soft cloth (avoid scratching)

Inspect Regularly:

• Monthly visual inspection
• Check for fading (compare to a new plate)
• Look for cracks or damage
• Verify secure mounting
• Test reflectivity at night

Replace When Needed:

• Every 5 to 7 years (preventive)
• Immediately if damaged
• When fading is noticeable
• If mounting is compromised

Ensure Proper Mounting:

• Perpendicular to the ground
• Secure fixing (no swinging)
• Correct height (400 to 1000mm)
• No obstruction (tow bars, bikes)
• Both plates visible

❌ DON’T:

Use Illegal Modifications:

• No tinted covers
• No reflective sprays
• No custom fonts
• No 3D characters affecting spacing
• No bolts obscuring characters

Ignore Damage:

• Cracked plates must be replaced
• Faded plates reduce readability
• Loose mounting causes angle issues
• Peeling sheeting fails ANPR

Assume “It Won’t Happen”:

• A 2 to 10% error rate affects millions
• Misreads can cause serious problems
• Prevention is your responsibility
• Fines of up to £1,000 for non-compliance

What to Do If ANPR Makes an Error

If You Receive an Incorrect Penalty:

Don’t Panic or Ignore It:

• Penalties escalate if unpaid
• Court action is possible
• Your credit rating may be affected

Gather Evidence Immediately:

• Photograph your plate (showing its condition)
• Note the date/time of the alleged violation
• Check the weather conditions
• Obtain your V5C registration document
• Take photos showing BS AU 145e compliance

Request ANPR Evidence:

• Ask for the camera images
• Request the confidence score
• Check if human verification occurred
• Review the image quality

Submit a Formal Appeal:

• Follow the appeal process on the notice
• Include all evidence clearly
• Explain the plate condition and compliance
• Reference BS AU 145e certification
• Keep copies of everything
• Meet all deadlines

Escalate if Necessary:

• Independent adjudicator (parking/Congestion Charge)
• Court hearing (criminal matters)
• ICO complaint (data protection issues)
• Police complaint (if wrongfully stopped)

If Stopped by Police Due to an ANPR Error:

1. Remain calm and cooperative
2. Explain the situation politely
3. Provide documentation (V5C, insurance, MOT)
4. Note that your plate is BS AU 145e compliant
5. Request clarification of the alert
6. Ask for the officer’s name and badge number
7. Follow up with a formal complaint if warranted

How Plate-Maker Ensures Maximum Readability

Our Manufacturing Standards:

As a DVLA-registered manufacturer (RNPS ID: 73132), we optimise every plate for ANPR performance.

BS AU 145e:2018 Certification:

• Certified reflective sheeting tested for NIR compatibility
• Reflectivity exceeds minimum requirements (400+ cd/lx/m²)
• Weather-resistant for 5 to 7 years
• Smooth surface for easy cleaning

Precision Manufacturing:

• Computer-controlled cutting ensures exact 11mm spacing
• Charles Wright font to exact specification
• Character height: 79mm (cars), 64mm (motorcycles)
• Stroke thickness: 14mm standard
• Regular equipment calibration

ANPR Testing:

• Sample plates tested with ANPR equipment
• Readability verified in multiple conditions
• NIR spectrum testing
• 99%+ first-pass read rate in testing
• Continuous quality improvement

Quality Control:

• Every plate inspected before dispatch
• Spacing verification
• Reflectivity spot-checks
• Font compliance confirmation
• RNPS ID displayed on all plates

Visit our same-day collection service in Ilford for plates engineered for maximum ANPR accuracy.

Frequently Asked Questions

What is the real-world accuracy rate of ANPR cameras in the UK?

Real-world ANPR accuracy averages 85 to 93% across all conditions, lower than the 95 to 98% claimed in laboratory testing. Factors reducing accuracy include weather (rain, snow, fog), dirty or faded plates, illegal fonts, high vehicle speeds, and suboptimal camera angles. With clean, BS AU 145e compliant plates in good conditions, accuracy reaches 94 to 96%, but this drops to 70 to 85% with non-compliant plates or adverse conditions.

Why do ANPR cameras misread number plates?

ANPR misreads occur due to character ambiguity (8/B, 0/O, 1/I look similar), dirty or obscured plates reducing contrast, illegal fonts or incorrect spacing, weather conditions scattering infrared light, high vehicle speeds causing motion blur, extreme camera angles creating perspective distortion, and faded or damaged plates with poor reflectivity. These factors can reduce accuracy from 95% to between 40 and 70%, triggering misreads that may result in incorrect penalties.

Can I appeal a penalty if ANPR misread my plate?

Yes, you can appeal any penalty caused by an ANPR error. Gather evidence including photographs of your compliant, clean plate, your V5C registration document, and the weather conditions at the time. Request the ANPR images and confidence score from the enforcement agency. Submit a formal appeal explaining the misread with supporting evidence. Most appeals for genuine ANPR errors are successful, especially if your plate is BS AU 145e compliant and was clean at the time. Keep copies of all correspondence and meet all deadlines.

How often are ANPR reads checked by humans versus fully automated?

Only 5 to 15% of ANPR reads receive human verification. High-stakes enforcement like speed violations and stolen vehicle alerts typically requires human review before action. Low-value automated penalties like parking violations, ULEZ charges, and routine tax/MOT checks are often fully automated with confidence thresholds above 85 to 90%. Appeals trigger human review, providing a safety net for errors. The 100+ million daily reads make 100% human verification impossible due to cost and volume constraints.

Do dirty number plates cause ANPR errors?

Yes, dirt is one of the leading causes of ANPR read failures. Light dirt reduces accuracy by 5 to 10%, moderate dirt by 15 to 25%, and heavy mud or snow can reduce accuracy to between 40 and 70% or cause complete read failure. Drivers are legally responsible under Regulation 11 of the Road Vehicles Regulations 2001 to keep plates “clear and legible”. Regular cleaning (weekly in winter, monthly in summer) prevents most dirt-related errors and avoids potential £100 to £1,000 fines for illegible plates.

Are 3D and 4D number plates less accurate with ANPR?

3D gel and 4D laser-cut plates achieve 94 to 96% ANPR accuracy when manufactured correctly to BS AU 145e standards, only 1 to 2% lower than standard 2D plates. The key is maintaining exact Charles Wright font spacing (11mm between characters), using certified reflective backgrounds, and ensuring characters don’t obscure the reflective surface. Poorly made 3D/4D plates with incorrect spacing or non-compliant materials can drop to 60 to 80% accuracy. Our 3D gel plates and 4D laser-cut plates are tested for ANPR readability and display RNPS ID 73132.

Conclusion

ANPR technology achieves impressive 90 to 98% accuracy rates, but that remaining 2 to 10% error rate translates to millions of misreads annually across the UK’s extensive camera network. Understanding these limitations is not about undermining the technology. It is about recognising that ANPR, while highly effective, is not infallible.

Environmental factors like rain, snow, and fog can reduce accuracy by 10 to 25%. Plate condition, whether dirt, fading, or damage, accounts for another 10 to 30% reduction. Technical limitations including camera angles, vehicle speed, and character ambiguity contribute further errors. The result is that real-world performance averages 85 to 93%, not the laboratory-claimed 98%.

For drivers, the message is clear: maintain BS AU 145e compliant plates, keep them clean and undamaged, ensure proper mounting, and replace them every 5 to 7 years. These simple steps maximise your plate’s readability, reduce misread risk, and protect you from erroneous penalties.

At Private Number Plate Maker Ltd, we manufacture every plate at our Ilford workshop using BS AU 145e:2018 certified materials specifically tested for ANPR compatibility. As a DVLA-registered manufacturer (RNPS ID: 73132), we ensure your plates achieve maximum readability while maintaining full legal compliance.

You can design your ANPR-optimised plates online, or visit our Eastern Avenue workshop for same-day collection. We’re here to ensure your vehicle’s plates perform flawlessly with the UK’s ANPR network.