Date of Award


Document Type

Master Thesis

Degree Name

Master of Engineering (Research)


Electronic Engineering

First Advisor

Dr. Joseph Connell


This work provides extensions of a standard D-Cam/Firewire digital camera platform for use within a multispectral imaging setting and includes real-time lossless image compression targeted at the re-configurable environment of an FPGA. The control and image acquisition structures specific to multispectral systems are detailed with respect to an embedded firmware solution, complementary host PC software driver routines, and camera hardware control infrastructures. The augmentation of the standards based implementation is discussed without compromising compliance with underlying specifications. The mandate for image data acquisition includes glue logic to efficiently transmit data via a high-speed communications pipe, IEEE - 1394 in this case, providing compact pixel representation via a packing process.

This packing procedure is evolved to include image compression using a similar approach taken by the lossless variant of the JPEG standard. A pipelined approach is taken to ensure fulfilment of the real-time constraint for consumption of image data from the sensor electronics. The design is configurable to the requirements of the system, either at compile time or in system. RGB is a subset of the multispectral image type, and the significant correlation present between spatially localised pixels across spectral offer additional compression ratio gains. An algorithm to exploit the inter-colour correlations is described and evaluated using a common set of image benchmarks.

Access Level