Skip to main content

← Back to Blue Iris

Blue Iris – Fix False Alerts


🔹 Overview

False alerts are the most common problem in Blue Iris systems.

They are typically caused by:

  • Poor motion detection
  • No zone control
  • Incorrect thresholds
  • Environmental noise (rain, bugs, shadows)
  • AI being used incorrectly

🎯 Objective

  • Reduce false alerts
  • Maintain real event detection
  • Improve AI accuracy
  • Create a stable alert workflow

🧠 System Insight

Motion → AI → Alerts

If motion is noisy, AI and alerts will also be noisy.


🛠️ Fix Workflow

Step 1 — Fix Motion Zones

Go to:
Camera → Motion/Trigger → Edit zone

  • Limit detection to important areas only
  • Exclude sky, trees, water, and roads
  • Avoid reflective surfaces

👉 Most false alerts originate here


Step 2 — Adjust Motion Thresholds

Recommended starting point:

  • Min object size: increase until noise disappears
  • Min contrast: ~25–40%
  • Min travel: small but not zero
  • Make time: ~1.0 sec

👉 Goal: filter out small or brief motion


Step 3 — Enable AI Confirmation

Go to:
Camera → AI tab

  • Confirm alerts with AI: Enabled
  • Minimum confidence: ~60%
  • Detect: Person, Vehicle

👉 AI filters false motion events


Step 4 — Use Main Stream for AI

Enable:

  • Use main stream images (if dual stream available)

👉 Better image = better AI decisions


Step 5 — Control Alert Behavior

Go to:
Camera → Alert tab

  • Trigger: New triggers only
  • Add to alerts list: Database only
  • Optional: require multiple alerts

👉 Prevents repeated or noisy alerts


📊 Common Causes of False Alerts

  • Bugs at night (IR reflection)
  • Rain / snow
  • Moving shadows
  • Trees / vegetation
  • Water movement (pools, lakes)

🧠 Real-World Strategy

Best results come from combining:

  1. Clean motion zones
  2. Proper thresholds
  3. AI confirmation

Not from increasing sensitivity


⚠️ Common Mistakes

  • Increasing sensitivity instead of reducing noise
  • Enabling AI before motion is tuned
  • Using too many object types
  • Ignoring environmental factors


✅ Result

A clean, stable alert system that captures real events without noise.