Why Is Frigate Person Detection Missing at Night?
- Poor night image quality (grainy IR)
- Detect resolution/stream poor at night
- IR/night vision washing out detail
Problem Description
Frigate detects people reliably during the day but misses them at night — person detections drop off after sunset even though the camera shows people moving in the IR view. Night images are grayscale with lower contrast, causing the detection model to return lower confidence scores that fall below the default threshold.
Why This Happens in Real Homes
Frigate detecting people reliably by day but missing them at night is usually an image-quality problem — the object model is trained largely on well-lit images, and grainy IR night footage, IR washout, or low detail gives it far less to work with, so confident detections drop. The detector is fine; the input at night is degraded.
Improve the night image: tune the camera's night vision and exposure so people are clearly visible without IR washout, and add supplemental lighting where practical, which dramatically improves detection. Make sure the detect stream's resolution is adequate at night, and consider lowering the person score threshold slightly for night conditions (balancing against false positives). A camera with strong low-light performance and good night image quality is what makes nighttime detection reliable.
Symptoms
- No person detection at night
- Detects by day, misses at night
- Night detection unreliable
- Misses people in the dark
- Person detection fails at night
- Nighttime false negatives
- Poor night detection
- Works in daylight only
Recognize these? Here's what usually causes it.
Common Causes
- Poor night image quality (grainy IR)
- Detect resolution/stream poor at night
- IR/night vision washing out detail
- Model less confident on IR/night images
- Detect fps too low at night
- Camera exposure/night settings
- Threshold too high for night confidence
- Lighting insufficient
Most fixes happen in the first 3 steps.
Do not over-lower thresholds globally; tune per camera/scene to limit false positives.
Tools & Requirements
Step-by-Step Solution
Check camera IR illumination and night image quality
Most IP cameras switch to infrared mode at night, producing a grayscale image. If the IR LEDs are weak or the camera's night mode is not activating: the image is too dark for the detection model to identify objects. Open the Frigate live view at night and check the image quality. People should be clearly visible as light figures against a darker background. If the image is very dark: check camera settings for IR LED intensity (set to high/auto), IR cut filter (should be on/auto at night), and exposure settings (avoid manual exposure locked to daytime values).
Adjust detection sensitivity for nighttime conditions
At night, people appear as low-contrast grayscale shapes — the detection model has lower confidence scores compared to well-lit daytime scenes. Lower the minimum detection score: cameras: your_camera: objects: filters: person: min_score: 0.4, threshold: 0.55 (defaults are 0.5 and 0.7). This accepts lower-confidence detections that the model is less sure about — appropriate for night images. Monitor for false positives: if lowering the score causes trees or shadows to trigger person detections, use motion masks on those areas instead.
Add supplemental lighting to the detection area
The most reliable way to improve night detection: add a motion-activated floodlight or always-on low-level LED light to the camera's field of view. Consistent lighting lets the camera stay in color mode (not IR), which dramatically improves detection accuracy because the model was primarily trained on color images. Even a 10W LED floodlight at the camera's mounting height significantly improves detection confidence at night. Smart floodlights (Ring Floodlight Cam, similar) combine camera + lighting in one unit.
Verify the detect stream resolution is adequate
At night with IR: fine details are lost. If the detect stream resolution is too low (320x240, 640x360): people at medium-to-far range are just a few pixels tall and the model cannot distinguish them from other objects. Increase the detect resolution for cameras that need night detection: detect: width: 1280, height: 720 — this gives the model more pixels to work with. The tradeoff is higher CPU/TPU usage per camera, but night detection accuracy improves substantially.
Use zones to focus detection on high-probability areas
Instead of detecting everywhere (which increases false positives at night): create zones over paths, doorways, driveways, and other areas where people are likely to walk. Set zone-specific object filters: zones: walkway: objects: [person], filters: person: min_score: 0.35. Lower scores inside high-probability zones are acceptable because the zone context already narrows what the object could be. Outside these zones: keep the default higher threshold to prevent shadow/animal false positives in less-controlled areas.
Quick Solutions
Still having issues? This is usually the deeper cause below.
If the sensor still misses events after repositioning, check whether a scheduled 'home' or 'away' mode is overriding the sensitivity setting silently.
Night detection quality depends heavily on camera image quality, not only model settings.
This issue almost always looks more complex than it is — the majority of cases trace back to a single setting, a stale credential, or a default that shipped wrong.
- Poor night image quality (grainy IR)
- Detect resolution/stream poor at night
- IR/night vision washing out detail
- Model less confident on IR/night images
- Detect fps too low at night
Before you go — try one of these (they fix most cases).
Official Manufacturer Manual
Frigate provides official product documentation through their online manual rather than downloadable PDF. Access setup guides, troubleshooting steps, and product specifications for your Frigate Night Detection.
Source: docs.frigate.video
Need More Help? Frigate Support
Note: The contact information below connects you directly to Frigate's official customer support team, not Trunetto. They can help with warranty claims, device replacements, and advanced technical issues.
Accessories owners commonly pair with Frigate Night Detection.
SANDISK 64GB Extreme microSDXC UHS-I Memory Card with Ada...
SANDISK 128GB Ultra microSDXC UHS-I Memory Card with Adap...
SANDISK 64GB Ultra microSDXC UHS-I Memory Card with Adapt...
As an Amazon Associate we earn from qualifying purchases.