Embedded Image Processing Application on a Multi-core Platform with FPGA

This project is a comprehensive project with FPGA. It is a course project of IL2212 - Embedded Software Lab course held at KTH in VT19. We designed and implemented the FPGA hardware architecture using Qsys and Verilog.Also, we developed three image detection and processing applications that can run on different hardware resources:

1. on a single core without operating system i.e. Bare-metal;
2. on a single core using the real-time operating system MicroC/OS-II;
3. on the multiprocessor without operating system while satisfying design constraints on worst case throughput and memory footprint.

In addition to these, we also developed a debug mode for software testing. For more details of what we have done, please refer to the report under the folder "report".

Contributors

• Jun Zhang
• Jaakko Laiho

Contents

This repository is organized as follows:

• app contains source files for the applications, grouped in subfolders by project. It also contains build automation scripts for helping with the lab tasks.
• bsp is an empty folder which will contain the generated Board Support Packages for different configurations.
• hardware contains files describing the custom hardware platform.
• model contains the (ready-to-build) ForSyDe-Haskell project which represents the executable functional specification of the application that needs to be implemented on the given hardware platform.
• test contains input images and scripts that help in generating input data, output plots and testing the implementations against the golden model.

For more detailed usage instructions, check the README files found in each folder.