Embedded Systems @ De Nayer Instituut

Multifunctional Embedded Digital Camera System

Publications

Here we'll present all publications that we made in relation to the research project.

Motion Detection

This paper describes an algorithm to detect moving objects within a frame. It is designed to be robust and efficient at the same time.

You can download it here.

Auto Exposure

This algorithm describes a way to automatically modify the exposure time of a CMOS sensor in order to obtain more optimal images.

You can read all about it here.

JPEG Compression

When working with images, one of the more important modules of an embedded system is the image compression. The project team decided to implement a standard JPEG compression.

You can read all about it here.

Sub-pixel edge detection

We have developped an algorithm for edge detection, that can provide the exact location of an edge even on a sub-pixel level. This can be very usefull in applications like calibration, modeling, ...

You can download it here.

Linux on the MPC5200

We've implemented Linux using the help of the DULG at DENX and documented it completely. It runs a 2.4 kernel, tinylogin, busybox etc. Wolfgang Denk has read it and made some corrections to the article.

You can download it here.

Inventory of embedded image processing systems

As a part of the research project, an inventory of existing image processing systems had to be composed. The project engineers made a list of existing systems, they found on the World Wide Web. This list, however, is far from complete. When the reader notices an image processing system that is not included in our inventory, please notify the project team and they will update the list.

The latest revision of the inventory can be downloaded here.

Autofocus

The first embedded design, made by the project team is now fully documented. The document describes a system that can determine the focal sharpness of a picture. The algorithm was evaluated in plain C, before it was transferred to the embedded platforms. Tests were done on the ML300 board from Xilinx and on the MPC5200 platform from Motorola.

Download the documentation here.

CMOS versus CCD Sensors

The heart of every image processing system is the image sensor. When we look at digital sensors our choice is narrowed to CMOS or CCD technology. The presentation here gives an overview on these technologies. The heart of every image processing system is the image sensor. When we look at digital sensors our choice is narrowed to CMOS or CCD technology.

The presentation here gives an overview on these technologies. You can download the presentation here.