Record3k

aim: real time bug tracker with GPU support

Parts

• image processing lib based on opencv (cpp)
• mini client app (html;js)
• mini tool to parameterizse the processing pipeline

Summary

features:

• moving object tracking on cameras with 10 FPS and more
• configurable data processing pipeline
• GPU support of Processing steps
• API Endpoints for controlling

Build

do a local build (deps needed)

/localBuild.sh --cuda -vimbax --api  -j 8    

do a docker build:

step 1: build config/DockerCustomBase

step 2: build config/DockerBuildOpenCV

step 3: build config/Dockerfile

Run

cd build/bin
# - copy default data folder if not exists
# cp -r ../../data . 

#  - mabye add lib to path
# export LD_LIBRARY_PATH=../lib:$LD_LIBRARY_PATH
# - run api server
# ./apiserver

View

cd client 
caddy run --config Caddyfile
# open via
firefox http://127.0.0.1:3000/index.html

License

Licensed under the MIT License. See the LICENSE file for the full text.

Copyright (c) 2025 record3k contributors