# OpenCV-AI-Competition-UTAR4Vision

**Repository Path**: Cpdc1/OpenCV-AI-Competition-UTAR4Vision

## Basic Information

- **Project Name**: OpenCV-AI-Competition-UTAR4Vision
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-03-19
- **Last Updated**: 2025-03-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# OpenCV-AI-Competition-UTAR4Vision
We are team **UTAR4Vision**!<br>
This is the github repository for our team project - **Visually Impaired Assistance in COVID-19 Pandemic Outbreak**<br>
Please refer the links before for the demonstrations of the project.<br>
[Demonstration Sample Images](https://github.com/yjwong1999/OpenCV-AI-Competition-UTAR4Vision/blob/main/demo/Demo.md)<br>
[Demonstration Video](https://youtu.be/kVfaFBObXq0)

# 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](https://docs.luxonis.com/projects/hardware/en/latest/pages/BW1098OAK.html#bw1098oak)
### 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<br>
<img src="https://user-images.githubusercontent.com/55955482/126036064-95e69b4f-7579-44f1-bb06-b003ed24fb72.png" 
     alt="depthai" width=600><br>

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

## 1. depthai_utils
Contains the code for each of the Pipeline required in this project<br><br>
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<br>

### 1. main.py
- the consolidated code<br>
<img src="https://raw.githubusercontent.com/yjwong1999/OpenCV-AI-Competition-UTAR4Vision/main/others/program%20flowchart.png" 
     alt="flowchart" width=600><br>
### 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