摘要
针对3 mm波段毫米波探测器信号处理系统的实时性要求,鉴于传统小波算法在硬件平台上的运算速度慢不能满足要求,对几种简化系数提升格式9-7小波的部分系数作了近似处理,使所有系数的分母都为2的整数次幂,算法只涉及整数的加法、乘法和移位运算。利用Matlab仿真,分析比较了它们对探测器回波信号的去噪效果及算法复杂度,并应用到基于DSP构建的探测器信号处理平台上。实验表明,能够满足信号去噪的实时性要求。
According to the requirement of 3 mm band MMW detector signal processing to real-time performance and considering that traditional wavelet algorithm is slow when operating on handware platform and can't meet the requirement, we made approximate treatment to some simplified lifting coefficients of 9- 7 wavelet. All the lifting coefficient denominators were made strictly to be power of two. Then the algorithm only involved integer addition, multiplication and shift. The detector echo signal denoising effect and the algorithm complexity were analyzed and compared based on simulation using Matlab. The algorithm was implemented on signal processing platform constructed by DSP. The experiments show that it can satisfy the real-time requirement to signal denoising.
出处
《电光与控制》
北大核心
2010年第2期74-77,共4页
Electronics Optics & Control
基金
武器装备预研基金项目资助(9140A05070107BQ0204)