Obstacle detection and quantification for vehicle mobility in winter conditions

Abstract

Recently, the focus for military operations has shifted from the desert to cold climates, causing a corresponding shift in the military's need to better understand the mobility of our current vehicle fleet in these areas. Therefore, this work investigated the effects of winter conditions on military vehicle mobility. The main objective was to detect obstacles on the scene. This study developed and tested a method for automatic obstacle detection in the digital elevation model of a scene. The method detects statistical anomalies relative to an estimated background image that contains no obstacles. The sensitivity of the detection can be adjusted by a specified probability of false alarms, and the obstacle detection confidence is characterized by a probability of detection. The visible height of obstacles above the snow is related to the actual height of the obstacles above the ground. Compared to other detection techniques, the developed method is fast, calibrates itself to the cluttered images, operates with a single given image, and aligns with a detection quantification adopted in the receiver operating characteristic framework. The examples considered in this paper demonstrate high efficiency and applicability of the developed approach to the military vehicle mobility missions.

Snow-covered obstacles’ effect on vehicle mobility

“…A number of methods were investigated, including edge detection and texture detection. As part of this project, Vecherin et al (2020) found that the texture method generated better results than the edge-detection method, allowing the detection of snow-covered obstacles with fewer false positives. It developed an additional algorithm to determine the probability of false positives.…”

“…It developed an additional algorithm to determine the probability of false positives. Vecherin et al (2020) fully describes the obstacle detection algorithms.…”