ycbq999 / OpenCV_Streamlit_APP

Try to make an online application with Streamlit App combined with modern vision ML technology, YOLO, OpenCV. Deploy it on Azure with CI/CD.

Deployment TBD

OpenCV_Streamlit_APP

Try to make an online application with Streamlit App combined with modern vision ML technology, YOLO, OpenCV. Deploy it on Azure with CI/CD.

Running Locally

• create virtual environment in current folder

  virtualenv .venv

• start virtual environment

  source .venv/bin/activate

• deactivate virtual environment

  deactivate

Automatically go to virtual vironment in VSCode:

• first setting all up and activate your virtual environment

• then choose file --> Preference -->settings select workspace tab then click open setting json icon from top right

• put this code in

  { "python.terminal.activateEnvInCurrentTerminal": true, "python.defaultInterpreterPath": "/home/ycbq999/face_recognition_venv/bin/python" }

• see reference [https://www.youtube.com/watch?v=g7bg2ADfx4c&ab_channel=JohnSolly]

Streamlit

• I include stream tutorial folder called 3_Streamlit_crash_course

• Streamlit treats multiple apps as pages, you need to make a fold named "pages" exactly to activate this feature. also name 1_,2_,--- infront of each page name so it will detect sequence. Remember that

see reference in udemy [https://www.udemy.com/course/yolo-custom-object-detection/learn/lecture/33903494#notes]

Web Video Streaming

• you need to install webrtc library to work on video stream project

• streamlit has streamlit-webrtc so you just need to install it from requirement txt.

• Reference [https://github.com/whitphx/streamlit-webrtc]

We will first use google colab to train the model

• Google Drive URL: [https://drive.google.com/drive/folders/16qo-WE2glfSrz5gEDGDMzCi7wLKpUeAu?usp=sharing]

• YOLO GitHub Repository: [https://github.com/ultralytics/yolov5]

• URL to Clone Repository: [https://github.com/ultralytics/yolov5.git]

Later we will use azure to train and deploy the model.

-reference [https://learn.microsoft.com/en-us/azure/machine-learning/tutorial-train-model?view=azureml-api-2]

-webrtc may not work in this environment due the security on the network. but still can deploy for image object detection.

HTTPS
streamlit-webrtc uses getUserMedia() API to access local webcams, and this method does not work in an insecure context.

This document says

A secure context is, in short, a page loaded using HTTPS or the file:/// URL scheme, or a page loaded from localhost.

So, when hosting your app on a remote server, it must be served via HTTPS.