Optimal speckle noise reduction filter for range gated laser illuminated imaging

ABSTRACT Laser illuminated imaging has a number of applications in the areas of night time air-to-ground target surveillance, ID, and pointing and tracking. Using a laser illuminator, the illumination intensity and thus the signal to noise ratio can be controlled. With the advent of high performance range gated cameras in the short-wave infra-red band, higher spatial resolution can be achieved over passive thermal night imaging cameras in the mid-wave in fra-red due to the shorter wave-length. If a coherent illuminator is used the resulting imagery often suffers from speck le noise due to the scattering o ff of a rough target surface, which gives it a grainy “salt and pepper” appearance. The pr obability density function for the intensity of focal plane speckle is well understood to follow a negative exponential distribution. This can be exploited to develop a Bayesian speckle noise filter. The filte r has the advantage over simple frame averaging approaches in that it preserves target features and motion while reducing speckle noise without smearing or blurring the images. The resulting filtered images have the appearance of passive imagery and so are more amenable to sensor fusion with simultaneous mid-wave infra-red thermal images for enhanced target ID. The noise filter improvement is demonstrated usin g examples from real world laser imaging tests on tactical targets.