| 摘要: |
| 针对深空次表层探测雷达相邻帧道数据相似的特性，提出了一种对相邻帧道差值数据进行分块自适应量化的算法——帧间差分分块自适应量化（Frame Difference Block Adaptive Quantization，FD-BAQ）。该算法首先对数据进行分块，然后进行帧间差分，并对差值数据或原始数据进行Lloyd-Max量化，最后用量化后的数据进行重构。在选择对差值数据或原始数据进行量化时，提出用数据方差作为量化误差的衡量指标，当子块差值数据方差小于原始数据方差时，对差值数据进行量化来替代对原始数据量化，否则直接量化原始数据，从而大幅减小量化误差。将该方法与已有改进型BAQ算法比较，实验结果表明，在相同压缩比条件下，FD-BAQ算法在数据域和图像域均能取得更好的压缩效果。 |
| 关键词: 次表层探测雷达 帧间差分 分块自适应量化 量化误差 |
| DOI：10.3969/j.issn.1672-2337.2018.05.002 |
| 分类号:TN957.52 |
| 基金项目:国家自然科学基金（No.U1531141） |
|
| Subsurface Sounding Radar Data Compression Based on Frame Difference |
|
LI Peng，LIU Xiaojun，FANG Guangyou
|
|
1.Key Laboratory of Electromagnetic Radiation and Sensing Technology，Chinese Academy of Sciences，Beijing 100190，China;2.Institute of Electronics，Chinese Academy of Sciences，Beijing 100190，China;3.University of Chinese Academy of Sciences，Beijing 100049，China
|
| Abstract: |
| In this study，an algorithm (frame difference block adaptive quantization，FD-BAQ) based on the similarity of deep space subsurface sounding radar adjacent frame data is proposed to quantify the difference data adaptively.The algorithm firstly blocks the raw data，secondly makes a frame difference for the adjacent frame data，then uses the Lloyd-Max quantizer to quantify the difference data，and reconstructs the raw data utilizing the quantified data at last.Before determining whether to quantify the difference data or the raw data，the algorithm utilizes variance as an index measuring the quantization error.When the variance of the difference data is less than that of the raw data，the algorithm quantifies the difference data instead of the raw data，otherwise，the raw data can be quantized directly，which greatly reduces the quantization error.The experimental results indicate that the proposed FD-BAQ algorithm can get better performance in both the data domain and image domain compared with the existing improved BAQ algorithms. |
| Key words: subsurface sounding radar frame difference block adaptive quantization quantization error |
