Self-Driving Car Prototype using OpenCV and Neural Networks

碩士 === 元智大學 === 通訊工程學系 === 105 ===   In recent years, the importance of computer vision is increasing, and it brings great convenience and market impact for human beings. Moreover, the embedded system has the advantages of portability, small size, high extensibility, low cost and easy operation to d...

Full description

Bibliographic Details
Main Authors: Shao-Lei Hsing, 邢紹磊
Other Authors: Po-Chiang Lin
Format: Others
Language:zh-TW
Published: 2017
Online Access:http://ndltd.ncl.edu.tw/handle/gyp93u
Description
Summary:碩士 === 元智大學 === 通訊工程學系 === 105 ===   In recent years, the importance of computer vision is increasing, and it brings great convenience and market impact for human beings. Moreover, the embedded system has the advantages of portability, small size, high extensibility, low cost and easy operation to design and multi sensor, flexible integration and other advantages, so let life has more applications. In this paper, real-time image recognition system is proposed, which combines the image processing and computer vision of Self-Driving Car. It purpose at the application of unmanned vehicles in the future. Expecting that Self-Driving car can drive smoothly in an unknown structural space without any collisions. This paper focuses on Self-Driving Car use ultrasonic sensors and image recognition as aids, and use Arduino to control Raspberry Pi send images to computer finally make self-adjustment in the environmental changes to achieve the final purpose of Self-Driving.