juzi-z / lane_change_prediction

Lane_Change_Prediction

Lane Departure Prediction prototype system implementation

copyright: University of Michigan Dearborn - Intelligent System Laboratory editor: Yuan Ma email: [email protected]

---

Table of contents

[TOC]

---

Project description

This project wants to predict driver's lane departure event in advanced using drivers' physiological signal feature.

Dataset

Note:

• Dataset is forbidden to share on Github.
• For more details about data description, please read dataDescription.md file.

---

Program Architecture

Please run this project program in the following sequence:

Step index	Objective	Code path
1	Environmental Variable Configuration	./preamble
2	Data Preparation	./dataPreparation
3	Data Preprocessing	./dataPreprocessing
4	Data Clean	./dataClean
5	Data Synchronization	./dataSynchronization
6	Signal Selection	./signalSelection
7	Feature Generation	./featureGeneration
8	Trainingset Generation	./trainingsetGeneration
9	Model Selection	./modelSelection
10	Event-based Training and Testing	./eventTrainTest
11	Real-time Training and Testing	./realtimeTrainTest

Environmental Variable Configuration

Note: For more details, please go to folder preamble and read the file preamble.md.

Data Preparation

Note: For more details, please go to folder dataPreparation and read the file dataPreparation.md.

Data Preprocessing

Note: For more details, please go to folder dataPreprocessing and read the file dataPreprocessing.md.

Data Clean

Note: For more details, please go to folder dataClean and read the file dataClean.md.

Data Synchronization

Note: For more details, please go to folder dataSynchronization and read the file dataSynchronization.md.

Signal Selection

Note: For more details, please go to folder signalSelection and read the file signalSelection.md.

Feature Generation

Note: For more details, please go to folder featureGeneration and read the file featureGeneration.md.

Trainingset Generation

Note: For more details, please go to folder trainingsetGeneration and read the file trainingsetGeneration.md.

Model Selection

Note: For more details, please go to folder modelSelection and read the file modelSelection.md.

Event-based Training and Testing

Note: For more details, please go to folder eventTrainTest and read the file eventTrainTest.md.

Real-time Training and Testing

Note: For more details, please go to folder realtimeTrainTest and read the file realtimeTrainTest.md.

---

Support StackEdit