Region of interest identification in unmanned aerial vehicle imagery

Author(s):  
Jeffrey L. Solka ◽  
David J. Marchette ◽  
George W. Rogers ◽  
Evelyn C. Durling ◽  
John E. Green ◽  
...  
2018 ◽  
Vol 14 (11) ◽  
pp. 160
Author(s):  
Yao Yao ◽  
Qing-le Quan ◽  
Hong-hui Zhang ◽  
Qiong Li

<p class="0abstract"><span lang="EN-US">In order to study the power patrol technology of unmanned aerial vehicle, the tracking algorithm was applied. The automatic patrolling of power lines was discussed in terms of algorithms. An unmanned aerial vehicle transmission line inspection method based on machine vision was proposed. The image and video of the unmanned aerial vehicle inspection of the power line had a complex background. By Wiener filtering de-noising and probability density functions, the image clarity was improved. According to the existing tracking techniques and algorithms, a Camshaft target tracking algorithm based on lossless Kalman filter was proposed. The method of non-destructive Kalman filter was adopted to predict the region of interest of power line identification. Using the Camshaft algorithm, the prediction of the window was searched and the size of the window was adjusted. Transmission lines were tracked in real time. The results showed that the restoration effect of the algorithm was obvious. The clarity of the image was improved. It prepared for the extraction and tracking of the future transmission lines. Therefore, the proposed method provides a feasible way for the UAV power line inspection technology based on machine vision.</span></p>


2021 ◽  
Vol 33 ◽  
pp. 221-236
Author(s):  
Zoya Hubenova ◽  
Konstantin Metodiev ◽  
Svetla Dimitrova ◽  
Liubomir Alexiev

This article proposes yet another approach towards looking into causes for attention distribution of an operator of unmanned aerial vehicle. During examination, the operator is being tested at dedicated flight simulator while data are gathered and visualized through a mobile eye tracker. Two work stages are considered sequentially, i.e. building a geometric 2D transformation of region of interest (homography) within an image, and overlaying a dynamic heatmap as well. In the former stage, spontaneous movements of the operator’s head, recorded by the video, are eliminated thus enabling the operator to use the mobile eye tracker instead of a desktop-based one. During the latter stage, the distribution of operator’s attention over time is displayed. In order to implement the current research, a source code has been developed in C++ for some features readily available in OpenCV library to be used. In addition, data gathered after carrying out flight session are processed and discussed thoroughly.


Author(s):  
Yifeng Cai ◽  
◽  
Kosuke Sekiyama

Cognitive sharing of objects is fundamental in a heterogeneous robot system composed of a Unmanned Aerial Vehicle and a ground robot. Since the viewpoint of a UAV is greatly different from a ground robot, they may have different perceptions about the same objects. That makes it difficult to realize cognitive sharing. In this paper, we proposed a cognitive sharing method which is based on Geometric Relation-based Triangle Representations. The method is able to make a UAV and a ground robot identify the same object from similar objects without sharing appearance information in unstructured environment. To copy with the problem of increasing computational cost for the recognition of objects in the Region of Interest, entropy evaluation is employed to evaluate and select unique representations. We illustrated the proposed method with robots in real world.


Author(s):  
Zhouyu Zhang ◽  
Yunfeng Cao ◽  
Meng Ding ◽  
Likui Zhuang ◽  
Jiang Tao

Autonomous carrier landing is regarded as a crucial problem among the flight stages of carrier-based unmanned aerial vehicle. In recent years, vision-based guidance has become a promising solution for unmanned aerial vehicle autonomous carrier landing. In this paper a new vision-based navigation scheme is proposed for unmanned aerial vehicle autonomous carrier landing. The scheme aims at dealing with two core problems: searching the carrier by using the images obtained from the airborne forward-looking camera and estimating the relative position and attitude between the unmanned aerial vehicle and the carrier. In order to solve the first problem, the spectral residual-based saliency analysis method is firstly adopted to obtain the Region of Interest. Then the locality-constraint linear coding-based feature learning method is proposed for feature extraction, and the region of interest containing the carrier is finally recognized by the linear support vector machine. In order to solve the second problem, five feature points are firstly selected on the surface of the carrier. Then, a new carrier-fixed moving reference coordinate system is set up. The six landing parameters including three attitude parameters and three position parameters are finally obtained by using orthogonal iteration. The experiment results verify the superiority and effectiveness of the algorithms proposed in this paper.


2020 ◽  
Vol 20 (4) ◽  
pp. 332-342
Author(s):  
Hyung Jun Park ◽  
Seong Hee Cho ◽  
Kyung-Hwan Jang ◽  
Jin-Woon Seol ◽  
Byung-Gi Kwon ◽  
...  

2018 ◽  
pp. 7-13
Author(s):  
Anton M. Mishchenko ◽  
Sergei S. Rachkovsky ◽  
Vladimir A. Smolin ◽  
Igor V . Yakimenko

Results of experimental studying radiation spatial structure of atmosphere background nonuniformities and of an unmanned aerial vehicle being the detection object are presented. The question on a possibility of its detection using optoelectronic systems against the background of a cloudy field in the near IR wavelength range is also considered.


Author(s):  
Amir Birjandi ◽  
◽  
Valentin Guerry ◽  
Eric Bibeau ◽  
Hamidreza Bolandhemmat ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document