TO DO's in this project.
-Object Definition -Water Bottle -++something
-Preparing a dataset
-Write a custom NN to make predictions and detections.
-Prepare a GUI for these. GUI will has 3 specs. -Live CAM and detection screen (Updating per frame) -Objects counts -Start-stop the production line
Production line is working => start detection and counting Production line is NOT working => stop detection and counting (Live Stream keep going but counting and detection will not going on) (Kivy)
-write a main python code for counting and combine it with NN and import main NN code to use.
STRUCTURE
-Step motor controlling a conveyor belt with raspberry pi
-Raspberry Pi controlled by PC via UDP on same network.
-PC doing detection and send cameras frames to GUI
-GUI showing these frames + controlling conveyor belt for starting and stoping.
FlowChart
-Capture Frame
-Do Detection
-Send frames to GUI
-Count Objects -------->1 Thread to push these information to cloud. (Power BI)