Final Project - Advanced Embedded Linux Development

This repository contains the individual portion of the final project submission.

Project Overview

For the complete project overview, architecture, and team information, please see the Project Overview Wiki Page.

Project board

All the issues and sprint planning can be found in the link below. The project consists of three sprints and it will approximately take 3 weeks. Tracking board: Github projects

Project Outline

A video of this project can be found here: Video

The video demonstrates:

This video shows my setup including a Raspberry pi 4B (64bit), an Intel RealSense D415 and connection to a display. The camera outputs RGB image, Depth information based on laser emitters, and IR camera output. The system is headless (without Desktop) so the out put directly will be shown on the display.

Challenges:

• Displaying video on headless system.
• Compatibility of diffrernt library version and linux kernel version.
• Providing enough power to the board.
• Finding the right serial port to show information on serial inerface.
• Time consuming build process for Yocto on old machines.

Lessons Learned:

• Always use the latest Yocto (in this case I am using Scarthgap)
• To debug the system having a system with desktop envirnoment can be helpfull (like using core-image-sato)

Repository Structure

This repository contains:

• default.xml - Repo tool manifest for managing Yocto layers
• meta-raspsense/ - Custom Yocto layer with core-image-raspsense recipe
• Yocto build configuration and custom layers
• Individual project components and scripts
• Documentation and build instructions

Project Description

This project implements a headless embedded Linux system on Raspberry Pi 4B (64-bit) with Intel RealSense D415 depth camera. The system streams video from the RealSense camera and displays it on an HDMI monitor without requiring a desktop environment.

Key Technologies:

• Yocto Project for custom Linux image
• Intel RealSense SDK (librealsense2)
• GStreamer for video streaming
• DRM/KMS for headless HDMI output

Links

• Project Overview Wiki
• Schedule Wiki
• Sprint 1 Issues Template

Quick Start

For Proposal Submission:

1. See SETUP_GUIDE.md for step-by-step instructions
2. Copy content from PROJECT_OVERVIEW.md to GitHub Wiki
3. Copy content from SCHEDULE.md to GitHub Wiki
4. Create GitHub Issues using templates in SPRINT1_ISSUES.md

For Yocto Build Setup:

1. See REPO_SETUP.md for instructions on using default.xml with repo tool
2. Initialize repo: repo init -u <this-repo-url> -m default.xml
3. Sync layers: repo sync
4. See BUILD_INSTRUCTIONS.md for building core-image-raspsense and generating .wic files

Building the Custom Image:

# After repo sync, set up build environment
source poky/oe-init-build-env build

# Add the custom layer
bitbake-layers add-layer ../meta-raspsense

# Configure for .wic files (in build/conf/local.conf)
# IMAGE_FSTYPES = "wic wic.bmap"
# MACHINE ??= "raspberrypi4-64"

# Build the image
bitbake core-image-raspsense

The .wic file will be generated at: tmp/deploy/images/raspberrypi4-64/core-image-raspsense-raspberrypi4-64.wic