v0.0.9

Add ffmpeg writer: store files with sound
Update precompiled archives: add mandatory json values
Minor fixes

v0.0.8

Add pause/resume commands
Update main menu
Update json
Update README
Minor refactoring

v0.0.7

Add HybridCodeprojectAI and HybridOpenCV object detect options
Add 'trigger' for simple motion detection - one frame is skipped to filter out image artifacts
Add decrease_detect_rate_while_writing option to decrease CPU/GPU load while writing video
Add option to set max frame buffer size
Some minor improvements and fixes

v0.0.6

Add simple motion detection engine (NB: changed engine selection option in settings json)
Add tweakable video writer codec (useful on low-end systems)
Uncomment some more objects to detect with OpenCV
Add uptime to status
Add notification on app start
Add detected objects classes to alarm description
Add preview_sampling_interval_ms option to settings.json to configure previews from config
Fixes related to linux
Bugfixes and tweaks

v0.0.5

Add some tweaks:

• Allow to define frame color and width in settings.json
• Option to send video file after it has been recorded
• Some minor fixes

v0.0.4

Add OpenCV DNN AI backend
AI backend can be selected in settings.json configuration file
Note that OpenCV DNN uses onnx model - download one (yolov5s recommended)
To enable CUDA with OpenCV DNN download OpenCV_CUDA_libs and unpack to application folder (replace existing)

v0.0.3 - Add command-line config parameter

Quick improvement allowing to run multiple instances with different configuration files. Could be useful when working with multiple cameras.
Usage: CameraAiDetector -c settings.json

v0.0.2

Add recorded video scale
Add /log command
Add translations
Some minor improvements and fixes

v0.0.1 - Initial release

Initial release