Perception-Aware Receding Horizon Trajectory Planning for Multicopters With Visual-Inertial Odometry

Abstract

Visual inertial odometry (VIO) is widely used for the state estimation of multicopters, but it may function poorly in environments with few visual features or in overly aggressive flights. In this work, we propose a perception-aware collision avoidance trajectory planner for multicopters, that may be used with any feature-based VIO algorithm. Our approach is able to fly the vehicle to a goal position at high speed, avoiding obstacles in an unknown stationary environment while achieving good VIO state estimation accuracy. The proposed planner samples a group of minimum jerk trajectories and finds collisionfree trajectories among them, which are then evaluated based on their speed to the goal and perception quality. Both the motion blur of features and their locations are considered for the perception quality. Our novel consideration of the motion blur of features enables automatic adaptation of the trajectory's aggressiveness under environments with different light levels. The best trajectory from the evaluation is tracked by the vehicle and is updated in a receding horizon manner when new images are received from the camera. Only generic assumptions about the VIO are made, so that the planner may be used with various existing systems. The proposed method can run in real time on a small embedded computer on board. We validated the effectiveness of our proposed approach through experiments in both indoor and outdoor environments. Compared to a perception-agnostic planner, the proposed planner kept more features in the camera's view and made the flight less aggressive, making the VIO more accurate. It also reduced VIO failures, which occurred for the perception-agnostic planner but not for the proposed planner. The ability of the proposed planner to fly through dense obstacles was also validated. The experiment video can be found at https://youtu.be/qO3LZIrpwtQ.

An autonomous Unmanned Aerial Vehicle exploration platform with a hierarchical control method for post‐disaster infrastructures

“…For instance, perception‐aware path planners aim to maximise the visibility of features within the camera's Field of View (FOV) to assist visual SLAM processes. As a result, their cost functions consider the amount and resolution of visual features [16]. In scenarios where the objective is to explore an unknown environment, Ref.…”

A Model Predictive-based Motion Planning Method for Safe and Agile Traversal of Unknown and Occluding Environments

“…For example, despite the existence of sophisticated nonlinear controller techniques [7], linear PID low level controllers offer sub-optimal performance but remain ubiquitous in many real-world applications for low level tracking due to speed and ease of implementation [8]. Aerodynamic effects [9], [10] and errors within the visual odometry pipeline [11], [12] are additional factors that can degrade trajectory tracking, which in turn introduces a risk of collision.…”

“…For example, ρ(•) may be a function that approximates some notion of risk that may be hard to write by hand, e.g., a learned policy trained on simulated or experimental data. Inclusion of the risk measure defined by (12) in the cost function has two different effects. First, trajectories are generally planned to be spatially away from obstacles in order to increase d obs .…”

“…A regression line dH (v max ) was fit to the data to capture the 95 th percentile, giving a conservative estimate of the actual set difference, dH (v max ) ≈ d H (•). This estimator of the set difference dH was used in(12) to find an estimate of the risk ρ(•).…”

Robust MADER: Decentralized and Asynchronous Multiagent Trajectory Planner Robust to Communication Delay

“…Equipping agents with sensors, often cameras, is a common method to detect and avoid unfamiliar obstacles [17]- [24]. This provides agents with real-time situational awareness, facilitating informed decisions for collision avoidance in dynamic settings.…”