Crown-of-Thorn Starfish (COTS) outbreaks are a major cause of coral loss on the Great Barrier Reef (GBR) and substantial surveillance and control programs are ongoing to manage COTS populations to ecologically sustainable levels. In this paper, we present a comprehensive real-time machine learning-based underwater data collection and curation system on edge devices for COTS monitoring. In particular, we leverage the power of deep learning-based object detection techniques, and propose a resource-efficient COTS detector that performs detection inferences on the edge device to assist marine experts with COTS identification during the data collection phase. The preliminary results show that several strategies for improving computational efficiency (e.g., batch-wise processing, frame skipping, model input size) can be combined to run the proposed detection model on edge hardware with low resource consumption and low information loss.

We describe an experimental platform that uses an evolutionary algorithm to automatically tune the gains of a cascaded PID quadcopter controller. All parameters are tuned simultaneously, few platform assumptions are necessary, and no modeling is required. The platform is able to run back-to-back experiments for over 24 hours without human intervention. In a sample experiment, we apply the system to solve a hovering task - the behaviors generated by an initially-random population of gain vectors are evaluated and gradually improved, with the attainment of high fitness hover controllers reported within 12 hours.

This paper demonstrates the feasibility of accomplishing real-world inspection tasks beyond visual range with an autonomous helicopter using simple but effective methods. We propose a LIDAR-based perception and guidance system that enables a helicopter to perform obstacle detection and avoidance, terrain following, and close-range inspection. The system has been implemented on board the CSIRO unmanned helicopter and flight tested in a number of different mission scenarios in unknown environments. We have successfully completed 37 missions and recorded more than 14 hours of autonomous flight time. Missions beyond visual range were executed without a backup pilot. The system has achieved a high success rate and has proven to be dependable.

This paper presents Multi-Step A∗ (MSA∗), a search algorithm based on A∗ for multi-objective 4-D vehicle motion planning (three spatial and one time dimensions). The research is principally motivated by the need for offline and online motion planning for autonomous unmanned aerial vehicles (UAVs). For UAVs operating in large dynamic uncertain 4-D environments, the motion plan consists of a sequence of connected linear tracks (or trajectory segments). The track angle and velocity are important parameters that are often restricted by assumptions and a grid geometry in conventional motion planners. Many existing planners also fail to incorporate multiple decision criteria and constraints such as wind, fuel, dynamic obstacles, and the rules of the air. It is shown that MSA∗ finds a cost optimal solution using variable length, angle, and velocity trajectory segments. These segments are approximated with a grid-based cell sequence that provides an inherent tolerance to uncertainty. The computational efficiency is achieved by using variable successor operators to create a multiresolution memory-efficient lattice sampling structure. The simulation studies on the UAV flight planning problem show that MSA∗ meets the time constraints of online replanning and finds paths of equivalent cost but in a quarter of the time (on average) of a vector neighborhood-based A∗.

We present a method for camera calibration and metric reconstruction of the three-dimensional structure of scenes with several, possibly small and nearly planar objects from one or more images. We formulate the projection of object models explicitly according to the pin-hole camera model in order to be able to estimate the pose parameters for all objects as well as relative poses and the focal lengths of the cameras. This is accomplished by minimising a multivariate non-linear cost function. The only information needed is simple geometric object models, the correspondence between model and image features, and the correspondence of objects in the images if more than one view of the scene is used. Additionally, we present a new method for the projection of circles using projective invariants. Results using both simulated and real images are presented. keywords: Least-squares model fitting, model-based vision, 3-D reconstruction, camera calibration, projective invariants. 1 Introduction ...