An Integrated Visual System for Unmanned Aerial Vehicles Following Ground Vehicles: Simulations and Experiments

Abstract

The vision of unmanned aerial vehicles is very significant for UAV-related applications such as search and rescue, landing on a moving platform, etc. In this work, we have developed an integrated system for the UAV landing on the moving platform, and the UAV object detection with tracking in the complicated environment. Firstly, we have proposed a robust LoG-based deep neural network for object detection and tracking, which has great advantages in robustness to object scale and illuminations compared with typical deep networkbased approaches. Then, we have also improved based on the original Kalman filter and designed an iterative multi-modelbased filter to tackle the problem of unknown dynamics in real circumstances of motion estimations. Next, we implemented the whole system and do ROS Gazebo-based testing in two complicated circumstances to verify the effectiveness of our design. Finally, we have deployed the proposed detection, tracking, and motion estimation strategies into real applications to do UAV tracking of a pillar and obstacle avoidance. It is demonstrated that our system shows great accuracy and robustness in real applications.

Industrial UAV-Based Unsupervised Domain Adaptive Crack Recognitions: From Database Towards Real-Site Infrastructural Inspections

“…Most importantly, a significant module for Unmanned Aerial Vehicle (UAV) intelligent inspection systems is to develop computer vision algorithms for processing images captured and detecting cracks and structural damages. The visual sensors have been emerging as novel techniques for modeling the environment [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], and has been widely adopted for robotic applications. Various visual sensors and Light Detection And Ranging Sensors (LiDAR) sensors have been widely deployed in various robotics platforms such as automatically navigated aerial robots and ground robots [16], [17], [18], [19], [20], [21].…”

“…Various visual sensors and Light Detection And Ranging Sensors (LiDAR) sensors have been widely deployed in various robotics platforms such as automatically navigated aerial robots and ground robots [16], [17], [18], [19], [20], [21]. For example, the SLAM is widely adopted for robot localization, mapping, and navigation applications [11], [22], [23], [12], [24], [25], [26], [27], [28], [29], [30], [31], [22].…”

A LiDAR-Inertial-Visual SLAM System with Loop Detection

“…The poses are also refined within the sliding window. The LVI-SAM [12], [14]- [18] framework is based on the traditional LOAM pipeline, which supports both the joint optimizations of the LiDAR odometry and the visual odometry and the individual optimizations of the VIO and the LIO system. The individual working mechanisms of LIO and VIO are activated when a failure occurs in one of the systems.…”

Semi-Supervised Confidence-Level-based Contrastive Discrimination for Class-Imbalanced Semantic Segmentation

“…And the heavy reliance of current deep learning-based semantic segmentation approaches on abundant richly labeled training data has prevented the application of them into real industrial applications, such as the infrastructural crack segmentation, and the industrial road segmentation in the transportation networks. Typically, the LiDAR is still the most popular and prominent sensor because it is the one of the most fast and accurate robotic sensors [1]- [3], [8]- [15]. However, the visual systems such as the cameras are also important for perceiving and interacting with the real-world.…”