Yingcai Bi

Design and implementation of an unmanned aerial vehicle for autonomous firefighting missions

This paper presents a design and implementation of an unmanned aerial vehicle (UAV) for outdoor firefighting application. The proposed UAV firefighting system consists of a self-designed quadcopter as platform, a transmission system to collect and release water, a real time kinematic (RTK) based navigation system and a mission control system to monitor and coordinate the UAV. In our proposed autonomous cooperative framework, the UAV finds an optimal path (with respect to distance and power consumption) to the fire spot at first. After arriving at the target on fire, the mission control system will guide the UAV to suppress the fire. The proposed framework has been demonstrated in the 2015 International Micro Aerial Vehicle Outdoor Competition, held in Aachen, Germany and help our team V-Lion obtain the second place. Our designed framework has successfully achieved all the requirements in firefighting mission even in a challenging wind gusts environment. The video of the firefighting mission can be found at: https://youtu.be/YREngbun7zQ.

An autonomous quadrotor for indoor exploration with laser scanner and depth camera

In this paper, we present a fully autonomous quadrotor for indoor exploration. The quadrotor is fully customized and capable of localization, mapping, planning and flying in unknown indoor environment with all real-time computations performed onboard. Two laser scanners are equipped to determine the 3D position of the quadrotor. The position measurements are further fused with Inertial Measurement Unit (IMU) to get a robust 6DOF state estimation. A depth camera is deployed to build 3D maps for the environment. Also, vision-based algorithms are designed to detect visual targets and perform precision landing. The whole system is verified in Singapore Amazing Flight Machine Competition (SAFMC). As Unmanned Systems Research Group from the National University of Singapore, we rank top in the fully autonomous category.

A stereo and rotating laser framework for UAV navigation in GPS denied environment

Recent developments in robotics sensing using either the active sensor-laser range finder (LRF) or passive sensors-camera systems have shown that the existing approaches are able to estimate the motion and reconstruct the environment in a typical GPS-denied environment such as indoor environments. However, for a 2D LRF, it can only provide a planar measurement due to the hardware limitations and the cost for 3D LRF is still too high for robotic systems. For camera systems, the 3D perception capability and lightweight are promising while it is not effective in long range, low illumination and low textured environment compared to the LRF. In the proposed framework, our first contribution is a senor integration system that combines a stereo camera with a rotating LRF. The stereo camera can achieve a fast and smooth motion estimation and the rotating LRF could construct a dense 3D environment. Our second contribution is an integration of a fast feature-based motion estimation with an accurate shape matching refinement. The proposed approaches have been built in our customized unmanned aerial vehicle (UAV) platform. We have verified our proposed approach through a series of extensive evaluations in clustered indoor environments and open outdoor environments.

A 3D Rotating Laser-Based Navigation Solution for Micro Aerial Vehicles in Dynamic Environments

In this paper, we present a 3D rotating laser-based navigation framework for micro aerial vehicles (MAVs) to fly autonomously in dynamic environments. It consists of a 6-degree of freedom (DoF) loc...

Semi-dense motion segmentation for moving cameras by discrete energy minimization

We present an approach for two-view motion segmentation for freely moving cameras, by formulating the epipolar constraint and spacial consistency into a discrete energy minimization problem, which can be efficiently solved using graph cut algorithms. With dense optical flow and proper sampling, a set of matched points is acquired for computing the fundamental matrix and the corresponding epipolar lines. The points distance to the epipolar lines, and their position on the image plane are used to construct a Markov Random Field (MRF) with discrete label-space. The 2-dimension label-space, i.e. motion area or static area, is computed using graph cut algorithms. We demonstrate the effectiveness of our method with the Johns Hopkins 155 motion dataset.

BIT*-based path planning for micro aerial vehicles

This paper presents a 3D on-line path planning algorithm for micro sized aerial vehicles (MAVs). The proposed approach adopts a two-layered planning framework. The first layer of the algorithm utilizes a sampling-based planner named Batch Informed Trees (BIT*) to quickly find an geometric obstacle-free passage. The second layer takes into account the dynamic constrains of the vehicle. By adopting a two-point boundary value problems (TPBVPs) approach, dynamically feasible trajectories can be generated efficiently within the previously found passage for lower-level controller. The main contribution of this work is proposing a complete on-line 3D path planning algorithm which can be implemented on the MAV with limited computational power.

A brief survey of visual odometry for micro aerial vehicles

Recently, visual odometry (VO) has experienced a rapid growth, which makes it viable for a range of applications. This survey paper attempts to provide a timely and comprehensive review of this field, focusing specifically on micro aerial vehicles (MAVs), with monocular, stereo or RGB-D cameras onboard. In this survey, the milestones in the development of VO will be reviewed, followed by an illustration of its general workflow, the commonly used datasets. The survey is concluded by an overall discussion.