Image processing techniques for determining dust optical density in Apollo videos have been developed. The software generates histograms, and calculates the mean and standard deviation, which are then used to match dusty and clear images for the purpose of estimating an effective optical density. A dust thickness model, based on the tilt of the camera and increasing thickness of the dust layer towards the top of the image, is used to account for the distance light travels through dust.

Previous methods relied on comparing specific features in clear versus dusty images, which severely limited the ability to analyze video frames. This method compares the statistical nature of a clear image to the statistical nature of a dusty image, assuming that the average scene’s description (as characterized by an image histogram) is invariant throughout the frame sequence. This assumption fails when shadows show up on the scene, which is evident in the last 20 seconds of the landing descent. For the last 20 seconds, histogram matching is not used, and is replaced by a manual choice of extinction model parameters based on the trending before shadows appear.

Using this image processing technique, mass erosion rate and total mass ejected from a surface due to rocket plume interaction can be estimated by analysis of the Apollo cockpit video sequence. Mass erosion rate of soil is measured by optical density (the second moment of particle size distribution), and is analogous to rainfall mass rate as measured by radar reflectivity (sixth moment of drop size distribution). The breakthrough in this approach is the realization that the National Weather Service (NWS) radar/rainfall correlation methodology can be used where various soil size distribution models can be substituted for different lunar locations.

This work was done by Philip Metzger of Kennedy Space Center and John Lane of EASI. KSC-13831/2