OpenCV-AI-Competition-UTAR4Vision

We are team UTAR4Vision!
This is the github repository for our team project - Visually Impaired Assistance in COVID-19 Pandemic Outbreak
Please refer the links before for the demonstrations of the project.
Demonstration Sample Images
Demonstration Video

Terminology

Some important terminologies for a better understanding on the readme

1. OAK-D

an OpenCV AI Kit (with depth) device
has a stereo camera setup, and have Intel VPU (Vision Processing Unit) chip in it
OAK-D specs

2. Pipeline

a pipeline is the workflow on the OAK-D side,
it consists of nodes (such as monocam node, neural network node, and etc.), together with the connections between them

Utilites files

The utilities files are Python Modules consisting useful functions or classes that will be used in the main code and demo codes.

1. depthai_utils

Contains the code for each of the Pipeline required in this project

The Pipelines involved are:

Navigation

obstacle detection
obstacle avoidance
social distancing

Walkable road path segmentation

road path segmentation for pedestrian's perspective

Pedestrian traffic light detection

to detect and classify the traffic light for pedestrian

Protective Equipment (PPE) detection

to detect hand sanitizer/face mask/thermometer

2. task_utils

Speech Recognition

an OOP object for speech recognition and voice command
designed for multithreading

Navigator

contains the functions for obstacle avoidance, social distancing

Sample Codes

Python file starting with "oakd_" prefixes are demonstration code for each of the tasks we addressed

1. main.py

the consolidated code

2. oakd_navigation.py

demo code for navigation

3. oakd_segmentation.py

demo code for road segmentation

4. oakd_traffic_light.py

demo code for pedestrian traffic light detection

5. oakd_ppes.py

demo code for ppe detection
use voice command to search for ppe items

6. oakd_ppes_no_audio.py

demo code for ppe detection (without voice command)
user has to type the item that he/she wants to search for
this is demonstration code for anyone not using the Google Speech to Text API

7. gps.py

the demo code for our gps system
it is seperated out from the main.py because it is hardware-dependant code

yjwong1999/OpenCV-AI-Competition-UTAR4Vision

About

Resources

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages