Eye-Tracking Virtual Keyborad
=============================
Was created with the help of
[Abd-El-Rahman Elhawary](https://github.com/AbdelrahmanElhawary)
[Omar Ayman](https://github.com/OmarAymanMahfouz)
[Moamon Ahmed](https://github.com/MoamenAhmedMostafa)
Software description
--------------------
The main goal of our project is to help paralyzed people, who can’t make any gesture or express their feelings,
deal with other people and communicate with them using a virtual keyboard that is displayed on the monitor and
use their eyes as a mouse to choose letters from that keyboard in order to display the chosen letters as a meaningful word that express the patient needs.
Helps limit the spread of the Coronavirus by using the apps without having to touch the mouse and keyboard.
Main Functions
-
* User will be able to write on the virtual keyboard using his eyes
* User will be able to control the mouse using his eyes.
* User will be able to convert text to speech.
* System provide word suggestion feature
How it works
-
In [EyeTrackerV2.py](https://github.com/AhmedHafez98/EyeTracker/blob/master/EyeTrackerV2.py)
* Using open CV, dlib and numpy module to detect the image.
* Eye blanking detection.
* Gaz Detection.
* Every 10 frames we locate the landmark points of the right and left eye.
* Send the eyes landmarks to Eye blanking detection and gaz detection to get eye state for each frame.
* Find most frequent state for each 10 frames as Eye State.
* Send this state to controller
In [WordPrediction.py](https://github.com/AhmedHafez98/EyeTracker/blob/master/WordPrediction.py)
* We use machine learning algorithm here.
* We Import big.txt (huge chunk of text).
* Clean the text using regular expression.
* Divide the text into inputs and outputs.
* We use one hot encoding to make the training easier.
* Now we move to building the model using embedding layer, LSTM layer and dense layers and train it for 500 epoch.
* take text input from the controller and we return the top 5 prediction from our model.
In [Threads.py](https://github.com/AhmedHafez98/EyeTracker/blob/master/Threads.py)
we distributed every main function into different threads so that they can work in parallel
In [Controller.py](https://github.com/AhmedHafez98/EyeTracker/blob/master/Controller.py)
* here we Design VK
* connect VK buttons to keyboard keys
* connect curser tracker
* connect with eye tracker
* control VK
* connect to text to speech
* connect to word prediction
* connect to mouse
* control Mouse Tracker
© 2020 GitHub, Inc.
Terms
Privacy
Security
Status
Help
Contact GitHub
Pricing
API
Training
Blog
About
没有合适的资源?快使用搜索试试~ 我知道了~
资源推荐
资源详情
资源评论
收起资源包目录
基于图像处理的眼动追踪虚拟键盘并输入内容.zip (34个子文件)
基于图像处理的眼动追踪虚拟键盘并输入内容
presentaion.pptx 457KB
Controller.py 16KB
WordPrediction.py 4KB
Main.py 263B
Tester.py 2KB
EyeTrackerV2.py 6KB
Threads.py 3KB
GUI
Keyboardv2.ui 119KB
Icons
up-arrow.png 638B
down-arrow.png 629B
backspace.png 367B
computer.png 462B
next.png 591B
audio.png 4KB
windows.png 310B
return.png 629B
test.ui 2KB
testk.py 2KB
VKDesign.py 120KB
__pycache__
VKDesign.cpython-38.pyc 72KB
VKDesign.cpython-37.pyc 71KB
testk.cpython-38.pyc 2KB
Resources
CSV
MapKeys.csv 881B
TwoDButtons.csv 609B
landmarks_points.png 14KB
shape_predictor_68_face_landmarks.dat 95.08MB
Prediction
word_dict 164KB
tokenizer.pkl 403KB
keras_next_word_model.h5 5.97MB
seq_len.pkl 5B
__pycache__
Threads.cpython-38.pyc 2KB
Controller.cpython-38.pyc 9KB
README.md 3KB
Documentation.doc 1.53MB
共 34 条
- 1
资源评论
小码蚁.
- 粉丝: 2534
- 资源: 4146
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 基于Javascript的影视动画设计源码 - cad
- 基于Java和深度学习的瓦斯浓度预测系统后端设计源码 - 瓦斯浓度预测后端
- Screenshot_20240528_103010.jpg
- 基于Python的新能源承载力计算及界面设计源码 - HAINING-DG
- 基于Java的本科探索学习项目设计源码 - 本科探索
- 基于Javascript和Python的微商城项目设计源码 - MicroMall
- 基于Java的网上订餐系统设计源码 - online ordering system
- 基于Javascript的超级美眉网络资源管理应用模块设计源码
- 基于Typescript和PHP的编程知识储备库设计源码 - study-php
- Screenshot_2024-05-28-11-40-58-177_com.tencent.mm.jpg
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功