基于自适应差分量化的红外背景抑制新方法  

A Novel Infrared Background Suppression Method Based on Adaptive Differencing Quantization

作　　者：汪大宝[1] 刘上乾[1] 张峰[1] 

机构地区：[1]西安电子科技大学技术物理学院,陕西西安710071

出　　处：《红外技术》2009年第3期163-166,共4页Infrared Technology

基　　金：国家自然科学基金资助项目(60377034)

摘　　要：针对复杂背景下红外弱小目标图像的背景抑制难题,从图像采集环节考虑,根据自适应差分量化理论,提出了一种基于自适应背景抑制的红外弱小目标图像采集方法。其基本思想是:根据红外弱小目标图像背景杂波的相关性,利用自适应预测器,由已采集的像素信息实时估计出下一时刻背景杂波的最佳估计,并将其反馈至原始图像信号输出端,与实际采集图像信号相比较。通过量化残差图像信号来获得预测增益,从而提高采集图像的信噪比。理论分析与仿真实验表明,与传统的直接图像采集方法相比,此方法不仅等效地提高模数转化器的性能,而且能够很大程度地提高采集图像的信噪比(SNR),SNR可由1.43提高到4.57。Aiming at the difficultie of infrared DIM-small targets image background suppression, a novel infrared DIM-small targets image acquisition method based on background suppression is proposed according to the adaptive differencing quantization algorithm. The basic concept is to estimate the best estimation of the background according to the correlation of the background clutter with adaptive prediction filter, and the residual image signal is acquisitied by subtract this best estimation from original image signal, and then, the residual image signal was quantized by the acquisition system. Theoretical analysis and simulation results show that, comparing with the traditional background suppression method, this method equivalent improve the ADC＇s performance and can effectively increase the SNR of the image, the SNR increase to 4.57 from 1.43.

关 键 词：红外图像采集 自适应差分量化 背景预测 模数转换 

分 类 号：TN957.52[电子电信—信号与信息处理]