China
Africa
Explore chapters and articles related to this topic
Underwater Target Tracking Control of an Untethered Robotic Fish with a Camera Stabilizer
Published in Junzhi Yu, Xingyu Chen, Shihan Kong, Visual Perception and Control of Underwater Robots, 2021
Junzhi Yu, Xingyu Chen, Shihan Kong
Despite successfully implementing vision-based target tracking with self-propelled robotic fish, the proposed control methods have some limitations. First, the image stabilization scheme based on one-DoF camera stabilizer is compact, yet unable to suppress slight roll jitter. As a prerequisite of a miniaturized integrated propulsion platform with a fishlike shape, designing two-DoF or even three-DoF camera stabilizers as much as possible will make the image stabilization effect better. Second, dramatic changes in target trajectory go along with a large hysteresis of the tracking system. The primary cause is that image acquisition and processing bring a pure hysteresis element with a large time constant for the overall active vision tracking system. In practice, this pure hysteresis element is likely to cause the active vision tracking system to be extremely unstable. Thus, it needs to determine the time constant of this pure hysteresis and further to develop corresponding compensation algorithms for better tracking performance. At last, the convergence time of the employed DDPG-based learning system is a bit long, possibly leading to the instability of the overall control system during the transition process. To avoid this dilemma, one possible alternative is to explore enhanced learning control algorithms that are well suited to bioinspired swimming and real-time implementation.
Path Planning of Unmanned Aerial Systems for Visual Inspection of Power Transmission Lines and Towers
Published in IETE Journal of Research, 2023
M. D. Faiyaz Ahmed, J. C. Mohanta, Alok Sanyal, Pankaj Singh Yadav
To calculate the effectiveness of quality in images, a video camera is mounted on the quadcopter that remains stable. Using a camera stabilizer, (Tarot 3D gimbal) holds the camera stable regardless of quadcopter movement. In addition, using a camera lens to capture high-definition images with consistent pixel density is proposed. As a result, the distance between the image sensor and the lens is FL millimeters. The size of the image is set to be 4:3 ratios in our case. The number of pixels with respect to images is defined by VNP. To match the rules of Department General of Civil Aviation (DGCA), the quadcopter must be less than 25 kg in weight by considering all the onboard components [19]. The surface of the tower is used to represent the coverage ratio of tower by a set of rectangular and triangular elements as Φs. As a result, the total area of tower is computed by adding the numerous surface features observed in each of the UAS images acquired while hovering.