Webcam-Object-Detector-Realtime

A light weight webapp for Object detection using Webcam

This WebApp has been tranined to detect ["background", "aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"]

[Notice -- This WebApp is now up and running on https://share.streamlit.io/siddharthksah/webcam-object-detector-realtime/main/main.py, It is still in the beta phase, please report bugs.]

Have fun! Happy browsing!

Below are instructions to implement in in your local system using a separate development environment using the Conda package management system which comes bundled with the Anaconda Python distribution provided by Continuum Analytics.

Step 1:

Fork and clone a copy of this repository on to your local machine.

Step 2:

Create a conda environment called web-object-detector and install all the necessary dependencies, the environment.yml file is uploaded in the repo for ease:

$ conda env create -n web-object-detector python=3.8 -y

Step 3:

Install the extra dependencies, it is not required but helps in making sure the jupyter notebook is running in the right conda env:

$ pip install -r requirement.txt
$ conda install nb_conda

Step 4:

Activate the `web-object-detector' environment:

$ source activate web-object-detector

To confirm that everything has installed correctly, type

$ which pip

at the terminal prompt. You should see something like the following:

$ ~/anaconda/envs/web-object-detector/bin/pip

which indicates that you are using the version of pip that is installed inside the web-object-detector Conda environment and not the system-wide version of pip that you would normally use to install Python packages.

Step 5:

Change into your local copy of the this repo:

$ cd Webcam-Object-Detector-Realtime

Step 6:

Run the code:

$ streamlit run main.py