论文部分内容阅读
对应用双波段红外数据探测地面目标的差分加权信号处理算法作了研究,评估了算法的三种变式:(1)单差值,(2)最小噪声和(3)最大信噪比。将热红外多光谱扫描传感器在不同地域收集到的两种场景的测试结果与理论分析结果作了比较。相关系数和彩色比一定时,可以预示出信噪比增益的总量。然而,目标输出信噪比与彩色比例可能有较大的变化。就本文估算的目标和场景而言,所得的典型增益可从几个dB损失(对无色目标)衰减到最大值约20dB。
We study the algorithm of differential weighted signal processing based on dual-band infrared data to detect ground targets, and evaluate three variants of the algorithm: (1) single difference, (2) minimum noise and (3) maximum signal-to-noise ratio. The results of the two scenarios collected in different regions of the thermal infrared multi-spectral scanning sensor are compared with the theoretical analysis results. When the correlation coefficient and color ratio are constant, the total amount of signal-to-noise gain can be predicted. However, the target output signal to noise ratio and color ratio may have a greater change. For the purposes and scenarios evaluated in this article, the resulting typical gain can be attenuated from a few dB losses (to a colorless target) to a maximum of about 20 dB.