Hardware-software co-design of 3D data compression for real-time ultrasonic imaging applications

Abstract

Many of the ultrasonic NDE and imaging applications require processing of huge amount of data in real-time. Compression of acquired data helps to reduce the storage and to rapidly transmit information to remote locations for further analysis. Signal fidelity, computational speed and resource utilization are the major parameters to be considered while designing the architecture for the compression algorithm. The objective of this study is to implement discrete wavelet transform (DWT) based ultrasonic 3D data compression algorithm on a reconfigurable ultrasonic system-on-chip hardware platform targeted for real-time ultrasonic imaging applications. The reconfigurable platform allows analysis of multiple architectures to suit various applications. The algorithm is implemented as a hardware-only design and hardware-software co-design. Both implementations provide a high signal compression ratio of about 98% with good quality signal reconstruction. This study demonstrates that, compressing 33 MBytes of experimental ultrasonic 3D data into 0.4 MBytes requires only one-fourth of a second for hardware-only design, and one minute for hardware-software co-design, making both designs highly suitable for real-time ultrasonic imaging applications.