FPGA Based Smart Camera for Real-time 2D Bar Code Recognition

Type: Master's Assignment
Contacts: Sabih Gerez, Bert Molenkamp, Berend Dekens
Student: Zhongli Liu
Project:  
Location: CAES, University of Twente

Description

A smart camera is able to generate specific information from the images that the sensor has captured. The application of this project is a real-time 2D bar code recognition which primarily requires high data rate processing and based on low-cost devices.

 

The intended bar code is called TRIP code which was proposed as a visual tag and the corresponding recognition algorithm needs both hardware and software implementations. The hardware part which requires massive repetitive computation and data parallelism will be implemented in the FPGA core on an Altera DE1 board. The software part which needs floating-point support will be implemented on the Nios II processor which can be customized in the same FPGA. The target system is expected first to capture the frames via a camera module (THDB-D5M) and then do hardware processing on the frames to generate binary edge images. Eventually the code will be recognized in the processings on the Nios II host, and the result will be transferred and displayed on the PC.

The system will be built with decisions of practical trade-offs during speed, accuracy, robustness, and cost. Different design methods as well as modifications of the algorithm are being considered in the first stage of the project. And some basic image factors such as resolution and brightness will also be taken into account. The performance of all aspects will be analyzed based on different conditions for the further research and potential applications.


Free Joomla Templates designed by Web Hosting Top