Robust mud detection is a critical perception requirement for Un manned Ground Vehicle (UGV) auton omous offroad navigation. A military UGV stuck in a mud body during a mission may have to be sacrificed or rescued, both of which are unattractive options. There are several characteristics of mud that may be detectable with appropriate UGV-mounted sensors. For example, mud only occurs on the ground surface, is cooler than surrounding dry soil during the daytime under nominal weather conditions, is generally darker than surrounding dry soil in visible imagery, and is highly polarized. However, none of these cues are definitive on their own. Dry soil also occurs on the ground surface, shadows, snow, ice, and water can also be cooler than surrounding dry soil, shadows are also darker than surrounding dry soil in visible imagery, and cars, water, and some vegetation are also highly polarized. Shadows, snow, ice, water, cars, and vegetation can all be disambiguated from mud by using a suite of sensors that span multiple bands in the electromagnetic spectrum. Because there are military operations when it is imperative for UGV’s to operate without emitting strong, detectable electromagnetic signals, passive sensors are desirable.

A General Dynamics Robotic Systems (GDRS)experimental unmanned vehicle (XUV) navigatesthrough a muddy grass field during a data collectionfor the Daytime Mud Detection System.
JPL has developed a daytime mud detection capability using multiple passive imaging sensors. Cues for mud from multiple passive imaging sensors are fused into a single mud detection image using a rule base, and the resultant mud detection is localized in a terrain map using range data generated from a stereo pair of color cameras. Thus far at the time of this reporting, JPL has:

  1. Performed daytime data collections, on wet and dry soil, with several candidate passive imaging sensors, including multi-spectral (blue, green, red, and near-infrared bands), short-wave infrared, mid-wave infrared, long-wave infrared, polarization, and a stereo pair of color cameras.
  2. Characterized the advantages and disadvantages of each passive imaging sensor to provide cues for mud.
  3. Implemented a first-generation mud detector that uses a stereo pair of color cameras and a polarization camera. For each set of input images, the innovators calculate degree of linear polarization (DOLP), back-project polarization pixels that have high DOLP into the left color image, generate a stereo range image (which is registered with the left color image), and insert detected mud into a world map using the stereo range data. As it is only expected for mud to occur on the ground surface, stereo range data are used to isolate ground surface pixels from the other pixels corresponding to ground clutter. Ground clutter pixels with high DOLP (such as vegetation) are ignored.

Techniques to estimate soil moisture content have been studied for decades for agricultural applications; however, mud detection for UGV autonomous navigation is a relatively new research area. Ground vehicle methods of soil moisture estimation have used passive microwave sensors, but the antennas tend to be bulky and have been mounted directly downwards. This requires a UGV to drive on potentially hazardous terrain in order to characterize it. This work involves detecting mud hazards from a UGV without having to drive on the hazard first.

Mud detection is a terrestrial application; however, the intermediate image processing steps and world modeling techniques performed for this task are valuable to terrain hazard assessment in general, terrestrial, or planetary situations.

This work was done by Arturo L. Rankin and Larry H. Matthies of Caltech for NASA’s Jet Propulsion Laboratory. For more information, contact This email address is being protected from spambots. You need JavaScript enabled to view it.. NPO-46624