JetTrack - MA_Intersection_Calculation

MA_Intersection_Calculation is JetTrack's main algorithmus for real time object position calculation.

Overview:

• Input-Data: Kafka Data Streams form the JetTrack - MA_DeepStream_Pipeline (running on each camera)
• Output-Data: Position of all objects in scene, which are detected by the deepstream pipeline

Development

1. Start KafkaBroker on local Network @ 10.42.0.1:9092 or wherever you want
2. Start MA_DeepStream_Pipeline on JetsonNano Devices
3. Adust to your CameraParameters in src/config.json
   • Resolution of detection stream
   • Translation vector of camera (eg. see OpenCV camera calibration)
   • Rotation vector of camera (eg. see OpenCV camera calibration)
   • intrinsicMatrix
   • pathToCalibrationFile (for realtime use of camera calibration files - cam mode)
   • distCoeefs (k1, k2, p1, p2,[,k3,[,k4,k5,k6]]) (eg. see OpenCV camera calibration )
4. Run main.py on localDevice (connected to CamNetwork or on JetsonDevice itself)

For running Realtime Object Positioning:

• source: local -> src/config.json, cams -> settings on cams webserver will be used
• method: aruco -> realign aruco marker to world coordinate system

$ python3 main.py source method

Todos

• Add more calibrationModes for realingment of different CalibrationObjects