Skip to content

Getting started

This walkthrough installs the motion-only application, checks a camera-day directory, and creates its first JSON, CSV, and text reports.

Prerequisites

Requirement Minimum version Check with
Python 3.11 python --version
Recording MKV, MP4, MOV, or AVI camreview inspect D:\Cameras\garage\2026-08-12

Your recordings must use this filename form:

<camera>_YYYY-MM-DD_HH-MM-SS.<extension>
garage_2026-08-12_08-00-00.mkv

Camera names may contain underscores. The filename timestamp is the authoritative local start time; filesystem timestamps are ignored.

Install

Create an isolated environment and install the package from PyPI:

py -3.12 -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install camreview
python3 -m venv .venv
source .venv/bin/activate
python -m pip install camreview

See Installation for the detection extra, development tools, upgrading, and removal.

First run

  1. Verify the installed command.

    camreview --version
    
    CamReview 1.0.1
    
  2. Inspect filenames and inferred gaps without decoding video. Replace the path with a real camera-day directory.

    camreview inspect "D:\Cameras\garage\2026-08-12"
    

    The output identifies one camera and date and ends with the detected extensions:

    Detected camera: garage
    Detected date: 2026-08-12
    Video files: 1
    First: 08:00:00
    Last: 08:00:00
    Apparent missing intervals: 0
    Unrecognized filenames: 0
    Extensions: MKV
    
  3. Scan the complete selected day.

    camreview scan "D:\Cameras\garage\2026-08-12" --time all
    

    A successful run prints Scan complete followed by absolute paths to these files:

    garage_2026-08-12_motion.json
    garage_2026-08-12_motion.csv
    garage_2026-08-12_motion.txt
    

What happened

CamReview derived the recording timeline from filenames, probed video durations, sampled frames through a persistent MOG2 background model, and grouped positive samples into events. JSON preserves every event, timeline gap, warning, effective setting, and performance value. CSV and text are review-oriented projections of the same run.

Next steps