Absolute scale in structure from motion from a single vehicle mounted camera by exploiting nonholonomic constraints

Abstract

In structure-from-motion with a single camera it is well known that the scene can be only recovered up to a scale. In order to compute the absolute scale, one needs to know the baseline of the camera motion or the dimension of at least one element in the scene. In this paper, we show that there exists a class of structure-from-motion problems where it is possible to compute the absolute scale completely automatically without using this knowledge, that is, when the camera is mounted on wheeled vehicles (e.g. cars, bikes, or mobile robots). The construction of these vehicles puts interesting constraints on the camera motion, which are known as "nonholonomic constraints" The interesting case is when the camera has an offset to the vehicle's center of motion. We show that by just knowing this offset, the absolute scale can be computed with a good accuracy when the vehicle moves. We give a mathematical derivation and provide experimental results on both simulated and real data. To our knowledge this is the first time nonholonomic constraints of wheeled vehicles are used to estimate the absolute scale. We believe that the proposed method can be useful in those research areas involving visual odometry and mapping with vehicle mounted cameras.

Robust Scale Estimation in Real-Time Monocular SFM for Autonomous Driving

“…Prior knowledge of the environment is used to counter scale drift in several monocular SFM systems, such as nonholonomic constraints for wheeled robots [19] or geometry of circular pipes [10]. Like ours, other systems also handle scale drift by estimating camera height above the ground plane [9,20,21].…”

Robust Monocular Visual Odometry for a Ground Vehicle in Undulating Terrain

“…This approach can recover motion at a low computational cost with a single feature point, and shows significantly improved accuracy compared to unconstrained cases. Scaramuzza also shows that a monocular camera placed with an offset to the vehicle rotation center can recover scale when the vehicle is turning [17]. In straight driving, however, the formulation degenerates and the scale is no longer recoverable.…”

“…In [4][5][6]17], the methods all assume a planar ground model. However, violation of the assumption can make motion estimation fail.…”

Full scaled 3D visual odometry from a single wearable omnidirectional camera

“…Loethe et al [12] use the prior knowledge of the distance of the camera to the ground plane to compute the scale factor of the scene, which is well suited for camera mounted on vehicles. Scaramuzza et al [21] exploit non-holonomic motion constraints of wheeled vehicles to resolve the absolute scale, although this is only posible when the vehicle turns.…”