Starting from the history of driverless cars and analyzing the current scenario of car manufacturers, in terms of self-driving cars, it was decided to study this phenomenon by building a prototype that could replicate its behavior. A model was created that was able to recognize lane lines and move autonomously within them. From the hardware point of view, the Raspberry Pi model B + was used on which a 5MP camera module was mounted. Everything was connected to two motor inductors connected to two rubber wheels and to the dual h bridge Motor driver. On the software side, a Phyton script was produced, using the OpenCV library for color detection, so that the prototype can stop in front of the detection of the red color, and of the edges of the carriageway lines in each frame thanks to the algorithm implemented (Canny Edge, Region of Interest, Hough Transform and Mean Slope Intercept). A web app was also built in Flask, whose goal was to stream the video stream retrieved from the camera.
Partendo dalla storia delle self-driverless car e analizzando lo scenario attuale delle case automobilistiche, in fatto di macchine a guida autonoma, si è scelto di studiare questo fenomeno costruendo un prototipo che sapesse replicarne il comportamento. Si è creato un modellino che fosse in grado di riconoscere delle linee di corsia e muoversi autonomamente all’interno delle stesse. Dal punto di vista hardware è stato utilizzato il Raspberry Pi modello B+ su cui è stato montato un modulo camera da 5MP. Il tutto è stato collegato a due motoinduttori connessi a due ruote gommate e al dual h bridge Motor driver. Lato software è stato prodotto uno script Phyton, utilizzando la libreria OpenCV per il rilevamento del colore, affinché il prototipo potesse fermarsi dinnanzi al rilevamento del colore rosso, e dei bordi delle linee di carreggiata in ogni frame grazie all’algoritmo implementato (Canny Edge, Region of Interest, Trasformata di Hough e Average slope intercept). È stata anche costruita una web app in Flask, il cui obiettivo è stato quello di trasmettere in streaming il flusso video recuperato dalla camera.
PROGETTAZIONE E SVILUPPO DI UN SISTEMA AUTONOMO DI GUIDA
CIFARATTI, VALERIA
2021/2022
Abstract
Starting from the history of driverless cars and analyzing the current scenario of car manufacturers, in terms of self-driving cars, it was decided to study this phenomenon by building a prototype that could replicate its behavior. A model was created that was able to recognize lane lines and move autonomously within them. From the hardware point of view, the Raspberry Pi model B + was used on which a 5MP camera module was mounted. Everything was connected to two motor inductors connected to two rubber wheels and to the dual h bridge Motor driver. On the software side, a Phyton script was produced, using the OpenCV library for color detection, so that the prototype can stop in front of the detection of the red color, and of the edges of the carriageway lines in each frame thanks to the algorithm implemented (Canny Edge, Region of Interest, Hough Transform and Mean Slope Intercept). A web app was also built in Flask, whose goal was to stream the video stream retrieved from the camera.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi di Laurea Cifaratti per Segreteria.pdf
File consultabile da postazioni biblioteca
Descrizione: Tesi di Laurea Cifaratti Valeria
Dimensione
32.61 MB
Formato
Adobe PDF
|
32.61 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14253/65345