Sinkage definition and visual detection for planetary rovers wheels on rough terrain based on wheel-soil interaction boundary

Wheel sinkage detection is of tremendous significance for planetary rover mobility optimization control and prevention of serious wheel sinking. Wheel sinkage is regarded as the distance between the lowest point of the wheel in the soil and the horizontal flat terrain on relative research. For rovers moving on rough terrains, the above definition is not reasonable because the horizontal flat terrain cannot be found. New wheel sinkage definition and detection method are proposed, based on vision. The sinkage definition rationality is analyzed for various terrain conditions. The discrete mathematical wheel sinkage calculation model, for which the input is acquired through a visual method, is built. Saturation of wheel-soil interaction image is adjusted by a dynamic piecewise nonlinear adjustment method The image is processed into a binary image based on HSI (Hue, Saturation, Intensity) color space. The wheel-soil boundary is extracted from the corrected wheel region outline according to its morphological features. The wheel sinkage and equivalent terrain interface angles are calculated through the discrete mathematical wheel sinkage calculation model. Sinkage definition rationality and detection method applicability are experimentally validated in the four typical terrain conditions (flat, bulgy, sunken, and uneven terrains). Particularly, the experimental results prove the sinkage detection method has high accuracy for flat terrain, for which, the deviation of the sinkage modulus is less than 2% of the wheel radius. The sinkage detection method is proved to have fairly reasonable adaptability to complex illumination conditions, based on the experimental results for different illumination conditions. (C) 2017 Published by Elsevier B.V.