论文标题
逆机光学截面
Inverse Airborne Optical Sectioning
论文作者
论文摘要
我们提出了逆空气中光学截面（IAOS）与逆合成孔径雷达（ISAR）的光学类比。可以通过固定的光学传感器（例如，在森林上方的悬停相机无人驾驶飞机的无人驾驶飞行器）对移动的目标，例如步行者，被植被大量阻塞的人。我们介绍了IAOS的原理（即，逆合孔径成像），解释如何通过过滤图像积分的ra radon变换来进一步抑制封闭器的信号，并介绍如何手动和自动估算目标运动参数。最后，我们表明，尽管在传统的航空图像中跟踪遮挡目标是不可行的，但在IAOS产生的积分图像中，它变得有效。
We present Inverse Airborne Optical Sectioning (IAOS) an optical analogy to Inverse Synthetic Aperture Radar (ISAR). Moving targets, such as walking people, that are heavily occluded by vegetation can be made visible and tracked with a stationary optical sensor (e.g., a hovering camera drone above forest). We introduce the principles of IAOS (i.e., inverse synthetic aperture imaging), explain how the signal of occluders can be further suppressed by filtering the Radon transform of the image integral, and present how targets motion parameters can be estimated manually and automatically. Finally, we show that while tracking occluded targets in conventional aerial images is infeasible, it becomes efficiently possible in integral images that result from IAOS.
