1 UA V Placement for Real-time Video Acquisition A Tradeoff between Resolution and Delay

2025-04-28 1 0 576.16KB 8 页 10玖币

侵权投诉

UAV Placement for Real-time Video Acquisition: A

Tradeoff between Resolution and Delay

Xiao-Wei Tang, Member, IEEE, and Xin-Lin Huang, Senior Member, IEEE

Abstract—Recently, UAVs endowed with high mobility, low

cost, and remote control have promoted the development of UAV-

assisted real-time video/image acquisition applications, which

have a high demand for both transmission rate and image

resolution. However, in conventional vertical photography model,

the UAV should ﬂy to the top of ground targets (GTs) to capture

images, thus enlarge the transmission delay. In this paper, we

propose an oblique photography model, which allows the UAV

to capture images of GTs from a far distance while still satisfying

the predetermined resolution requirement. Based on the proposed

oblique photography model, we further study the UAV placement

problem in the cellular-connected UAV-assisted image acquisition

system, which aims at minimizing the data transmission delay

under the condition of satisfying the predetermined image

resolution requirement. Firstly, the proposed scheme is ﬁrst

formulated as an intractable non-convex optimization problem.

Then, the original problem is simpliﬁed to obtain a tractable

suboptimal solution with the help of the block coordinate descent

and the successive convex approximation techniques. Finally,

the numerical results are presented to show the effectiveness

of the proposed scheme. The numerical results have shown that

the proposed scheme can largely save the transmission time as

compared to the conventional vertical photography model.

Index Terms—Unmanned aerial vehicles, UAV placement, im-

age acquisition, and oblique photography model.

I. INTRODUCTION

The great progress of aviation, energy and artiﬁcial intel-

ligence (AI) technology promotes the rapid development of

unmanned aerial vehicles (UAVs), empowering them many

advantages including low cost, controllable mobility, and line-

of-sight (LoS) link with ground users [1]. At the same time,

UAVs are endowed with the ability of real-time ultra high-

deﬁnition image transmission in the ﬁfth generation (5G)

of mobile network, thus giving birth to many UAV-assisted

image/video acquisition applications such as live broadcast,

disaster monitoring, agriculture precision, and virtual real-

ity/actual reality (VR/AR) [2].

Different from conventional UAV-assisted applications, e.g.,

remote sensing, the emerging UAV-assisted image/video acqui-

sition applications need to transmit the captured image/video

data back to the base station (BS) in real time, which are thus

facing severe challenges. Firstly, these applications have an

extremely large demand on bandwidth due to the huge amount

of video/image amount [3]. Then, in general, these applications

usually bear rigorous quality of experience (QoE) requirements

since users expect to receive videos with low frame loss rate,

tolerable end-to-end delay and little jitter [4]. Last but not

least, the performance of these applications is still restricted

by UAV’s communication range and ﬂight endurance owing

to the limited power supply [5].

The challenges mentioned above are becoming irreconcil-

able when adopting conventional vertical photography tech-

nique to capture images/videos. To be speciﬁc, in conventional

vertical photography model, the UAV can only capture images

at the top of the ground target to make sure it is located in

the center of the image/video so that users can easily focus

on it [6]. Although this can provide high-deﬁnition images

for users, the transmission rate will be very low when the

GT is far away from the BS. Even worse, once the distance

between the GT and the BS is greater than a certain threshold,

the communication between the UAV and the BS will be

interrupted due to the limited transmit power of the UAV,

which seriously degrading user’s QoE. Fortunately, oblique

photography makes up for the limitation that images can

only be photographed from vertical angles in the past [7].

Speciﬁcally, the UAV doesn’t need to ﬂy to the point above

the GT, but can choose a location between the BS and the

GT when capturing images/videos, thus providing a ﬂexible

tradeoff between image resolution and communication quality.

A handful of research work on the oblique photography has

been done in recent years [8]–[14]. Speciﬁcally, the concept

of the oblique photography ﬁrst appeared in aerial survey for

visualization purposes. By carrying multiple sensors on the

UAV and collecting images from 5 different angles including

1 vertical and 4 oblique angles at the same time, the real

and intuitive image effect that conforms to human vision can

be generated via a series of data processing methods such

as multi-vision image joint adjustment and dense matching

of multi-vision images [8], [9]. In addition to aerial survey,

object measurement is also one of the main uses of obilque

photography. H¨

ohle et al. proposed to measuring the distances,

coordinates, elevations or areas of objects via oblique images

[10]. Zhou et al. acquired images of plantations with different

ages via UAV oblique photography and then extracted tree

heights according to reconstructed three dimension (3D) point

clouds [11]. Aghaei et al. employed the UAV to ﬂy over

a test laboratory in order to capture images at different

altitudes, in order to investigate the correlation between aerial

image capture altitude and potential defect identiﬁcation on

photovoltaic modules [12]. Lin et al. adopted an electric

ﬁxed-wing UAV loaded with a digital camera to take oblique

photographs of a sparse subalpine coniferous forest in the

source region, aiming at extracting individual tree heights

with the help of generated point cloud data obtained from the

overlapping photographs [13]. Zhang et al. proposed a UAV-

based panoramic oblique photogrammetry (POP) approach to

arXiv:2210.04677v1 [cs.MM] 10 Oct 2022

achieve georeferenced panoramic images and real 3D models

together by using panorama image projection algorithms [14].

Fig. 1 The tradeoff between the image resolution and channel quality

in the UAV-assisted image acquisition system.

Although the above-mentioned literatures have made a

good progress in aerial survey and object measurement, the

oblique photography model they adopted is not suitable for

UAV-assisted image acquisition. On one hand, the oblique

photography model they adopted still can’t make sure that

the GT is imaged at the center of the photograph all the time

which is not convenient for users to watch. On the other hand,

the resolution metric is only slightly modiﬁed according to the

traditional vertical photography model (i.e., adding the effect

of the oblique angle) [15], which lacks a more comprehensive

formula related to the UAV’s 3D coordinates. To address these

issues, we propose a novel oblique photography model in this

paper which quantitatively analyzes the inﬂuence of UAV’s

space position on the image quality. Based on the proposed

oblique photography model, we further consider a UAV-

assisted image acquisition and transmission system where the

UAV is adopted to capture images of a GT via the carried

camera and then transmit the captured image to the BS through

the wireless backhaul. Fig. 1 shows a common scenario which

is often encountered in practice: 1) at point PA, the captured

image has a low resolution, but the channel quality is good,

and 2) at point PB, the captured image has a high resolution,

but the channel quality is poor. How to choose between these

two points may make people confused. Therefore, the goal

of this paper is to study the UAV deployment problem where

the transmission delay can be minimized while satisfying the

predetermined resolution requirement.

Contributions: The main contributions of this paper are three-

fold:

1) A UAV-assisted oblique photography model is created

where the resolution of the captured image is determined

by the 3D coordinate of the UAV as well as the scale

of the GT, which thus provides several new potential

research directions.

2) The UAV deployment problem is modeled as a non-

convex optimization problem, aiming at minimizing the

data transmission delay while ensuring that the predeter-

mined resolution requirement can be satisﬁed. We ﬁrstly

simplify the three-variable original problem into a two-

variable one and then propose a suboptimal solution to

the simpliﬁed problem by using the block coordinate de-

scent (BCD) and successive convex approximation (SCA)

techniques.

3) Detailed numerical results are provided to verify the

effectiveness of the proposed system. Firstly, the effects

of horizontal and vertical coordinates on the resolution

of UAV are analyzed. In addition, the performance of

the UAV deployment problem are compared under three

different solutions including 1) conventional scheme, 2)

the proposed scheme solved by ES, and 3) the proposed

scheme solved by BCD and SCA.

The reminder of the paper is organized as follows. Section

II describes models of UAV-BS channel, oblique photography,

and image transmission. In Section III, the original opti-

mization problem is stated and simpliﬁed. In Section IV, an

effective BCD and SCA-based algorithm is proposed to solve

the simpliﬁed non-convex problem. In Section V, the numerical

results are provided to show the effectiveness of the proposed

system. In Section VI, we conclude this paper. In Section VII,

we provide some potential research directions.

II. SYSTEM MODEL

Consider an UAV-assisted image acquisition system where a

rotary-wing UAV is deployed to capture an image for a ground

target (GT) and transfer the captured image data back to the

base station (BS) immediately. In the following subsections,

models of UAV-BS channel, oblique photography, and image

transmission are described, respectively.

A. UAV-BS Channel Model

In this paper, we consider a real-time image acquisition

system with a UAV, a BS, and a GT, where a three-dimensional

(3D) Cartesian coordinate system is adopted. We assume that

the coordinates of the BS and the GT are known a priori to the

UAV. Let wT

b, zbdenote the 3D coordinate of the BS, where

wb=[xb, yb]T∈R2×1represents the horizontal coordinate and

zbrepresents the vertical coordinate, respectively. Similarly,

let qT, zdenote the 3D coordinate of the UAV, where

q= [x, y]T∈R2×1represents the horizontal coordinate and

zrepresents the vertical coordinate, respectively. As such, the

distance between the UAV and the BS, denoted by du,b, is

given by

du,b =qkq−wbk2+(z−zb)2.(1)

For ease of analysis, we assume that the wireless channel

between the UAV and BS is dominated by the LoS link.1

1Note that this work can be extended to more complex channel models such

as probabilistic LoS model or Rician fading model [16]. However, we may

start with the simplest case under the LoS channel, where a trade-off between

the image resolution and transmission rate still exists due to the possibly large

horizontal distances between the GTs and the BS.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

1UAVPlacementforReal-timeVideoAcquisition:ATradeoffbetweenResolutionandDelayXiao-WeiTang,Member,IEEE,andXin-LinHuang,SeniorMember,IEEEAbstractRecently,UAVsendowedwithhighmobility,lowcost,andremotecontrolhavepromotedthedevelopmentofUAV-assistedreal-timevideo/imageacquisitionapplications,whichhaveahi...

展开>> 收起<<

1 UA V Placement for Real-time Video Acquisition A Tradeoff between Resolution and Delay.pdf

共8页,预览2页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

1 UA V Placement for Real-time Video Acquisition A Tradeoff between Resolution and Delay

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: