Target-aided fixed-quality-of-service compression of SAR imagery for transmission over noisy wireless channels

We present a target-aided, fixed-quality-of-service (FQoS) system for coding synthetic aperture radar (SAR) imagery, whereby regions of interest and background information are coded independently of each other. A multiresolution, constant-false-alarm-rate (CFAR) detection scheme is used to discriminate between target regions and natural clutter. Based on the detected target regions, we apply less compression to targets, and more compression to background data. This methodology preserves the relevant features of targets for further analysis, and preserves the background only to the extent of providing contextual information. The background data is coded using channel-optimized trellis-coded quantization (COTCQ), while the target regions are coded using a combination of COTCQ and channel coding, depending upon the channel state. The combination of COTCQ and channel coding provides fixed target detection performance at the decoder, independent of the channel bit error probability. Additionally, all-pass filtering is employed, which significantly increases the detection performance at low encoding bit rates, while increasing the security level of the encoded bit stream. The proposed coder is evaluated in terms of target detection performance on the decoded SAR image for a wide variety of channel conditions and encoding bit rates. The resulting system dramatically reduces the bandwidth requirements of digital SAR imagery, and is shown to provide outstanding performance in adverse channel conditions.