A Literature Review on Vehicle Detection and Tracking in Aerial Image Sequence using Deep Learning Technique

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 06 | June 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1982
A Literature Review on
Vehicle Detection and Tracking in Aerial Image Sequence using Deep
Learning Technique
Niyazahamad S K1, Sanjay Subramanya2, Rishab Chandrashekar Shirur3, Ratan Prakash
Shikhari4
1,2,3,4 Department Of Computer Science And Engineering, Dayananda Sagar College of Engineering, Bengaluru
----------------------------------------------------------------------***-----------------------------------------------------------------------
Abstract: Visual multi-object tracking that is
both robust and high-performing is indeed a key
difficulty in computer vision, particularly with in context
of drones. Small target recognition and tracking in UAV
situations is problematic for standard Multi-Object
Tracking (MOT) techniques based on the tracking-by-
detection paradigm. We will be performing real time
vehicle detection and tracking on Aerial Image
Sequences using different AIML approach which
comprehensively includes techniques such as Image
Processing, Pattern Recognition and Computer Vision. It
has a wide variety of applications encompassing Visual
Surveillance, Traffic Control, Digital Forensics and
Human-Computer Interaction.
I. INTRODUCTION
Visually multi-object monitoring that is both resilient &
high-performing is indeed a major difficulty in computer
vision(CV), particularly in the context of UAV. Also with
huge popularisation for commercialized unmanned aerial
vehicles (UAVs) as well as the advancement of
OpenCV & AIML technologies, drone detection methods are
becoming a hot topic for researchers. Auto-navigation,
campus security surveillance, & catastrophe assistance
have all become easier thank to effective video image
computational techniques and powerful deep neural
networks.
II. SURVEY
[1] Multiple Object Tracking Performance: The
CLEAR MOT Metrics
Metrics that describe, he quality and key characteristics
in numerous object tracking systems must be studied and
compared in accordance to carefully analyse and evaluate
their performance. Regrettably, there has yet to be
agreement on such a range of generally valid measures.
They present two new measures for evaluating MOT
systems in this paper. Multiple object tracking precision
(MOTP) as well as multiple object tracking accuracy
(MOTA) are suggested benchmarks that can be used for a
variety of monitoring activities & permit for objective
contrast of tracking systems' primary features, like
accuracy at locating targets, precision at recognising
target configurations, but also way to detect targets on
consistent bases. They put the proposed metrics to the test
in a series of global evaluation workshops to see how useful
and expressive they were. The CLEAR workshops in 2006
and 2007 featured a wide range of monitoring activities
whereby a big number of models were tested & evaluated.
Their studies findings reveal that its suggested measures
accurately reflect the numerous methods' qualities and
shortcomings in a simple and direct manner, helps in easy
evaluation in performance, thus relevant toward a wide
range of circumstances.
[2] Fully-Convolutional Siamese Networks for Object
Tracking
Traditionally, the problem of arbitrary target tracking
was tackled by developing a system of the targets arrival
entirely online, with only the video as training data. Despite
their effectiveness, these approaches' online-only
methodology limits using depth information which can be
studied. Many efforts have actually been developed
towards harnessing deep convolutional networks'
descriptive ability. Once the target to monitor isn't
determined ahead of time, Stochastic Gradient Descent
online is required in adjusting the network's parameters,
risking overall system's speed. For object detection in
video, a basic tracking method is combined with a novel
fully-convolutional Siamese network that has been trained
end-to-end on the ILSVRC15 dataset. The tracker reaches
state of art success in various tests with the minimal
brevity. It works at fps that are faster than actual.
[3] Simple Online And Realtime Tracking
This research looks at a realistic approach for
monitoring many items, with the primary objective of
associating objects successfully for online and real-time
operations. The study claims that recognition ability is a
critical component in determining detection accuracy, with
modifying its detector boosting tracking efficiency by up as
18.9 percentage. In contrast to many batch-based tracking
systems, this research focuses on online tracking, where

the tracker is only shown detections from the previous and
current frames. Despite just employing a simple mix of
existing techniques such as the Kalman Filter and the
Hungarian algorithm for the tracking components, this
approach achieves tracking accuracy similar to state-of-
the-art online trackers. This research looks at a realistic
approach for monitoring many items, with the primary
objective of associating objects successfully for online and
real-time applications. The study claims that detection
quality is a critical factor in determining tracking
performance, with changing the detector boosting tracking
performance by up to 18.9%. The tracker also updates at a
rate of 260 Hz, which is nearly 20 times faster than other
state-of-the-art trackers due to the simplicity of the
tracking method.
[4] High-Speed Tracking-by-Detection without Using
Image Information
As the performance in object detectors increases, the
foundation for a tracker becomes significantly more
trustworthy. The problems for a successful tracker have
changed as a result of this, as well as the increased use of
higher frame rates. As a result of this shift, considerably
simpler tracking algorithms may now compete with more
complex systems for a fraction of the processing cost. This
paper outlines and illustrates such a method by conducting
extensive tests with a range of object detectors. The
proposed technique can easily operate at 100K fps on the
DETRAC vehicle tracking dataset, beating the state-of-the-
art. The notion of a passive detection filter is used to
analyse a very simple tracking technique in this research.
Due to its modest computing footprint, the suggested
approach can serve as a basic predictive model for other
trackers and provide an appraisal of the necessity of
additional efforts in the tracking algorithm. It also permits
reviewing tracking benchmarks to evaluate if the specific
concerns they indicate (for instance, missed detections,
frame rate, etc.) are within the capabilities of existing
algorithms.
[5] Cascade R-CNN: Delving into High Quality Object
Detection
The this study, they presented the Cascade R-CNN, a
multi-stage object recognition framework for developing
high-quality object detectors. Overfitting during learning
and quality disparity during inference have both been
demonstrated to be avoided using this design. On the hard
COCO and famous PASCAL VOC datasets, the Cascade R-
substantial CNN's and consistent recognition
improvements show that effective object detection
necessitates modelling and knowledge of several
corroborating aspects. The Cascade RCNN has been shown
to work with a wide range of object detection architectures.
They hope it will be useful in a variety of future object
detection research projects.
[6] Real-Time ‘Actor-Critic’ Tracking
Object tracking is one of the most important step in
object detection and tracking. In this paper Actor-Critic
framework been used, where the 'Actor' model seeks to
infer the best option in a continuous action space, causing
the tracker to move the bounding box to the object's
current location. The 'Critic' model is used for offline
training to create 'Actor-Critic' framework along with
reinforcement learning as well as a Q-value for directing
both the 'Actor' and 'Critic' deep network learning
processes. Visual tracking is viewed as a dynamic search
process in which the 'Actor' model outputs only one action
to locate the tracked object in each frame. Offline training
of better policy for finding the best result is done using
reinforcement learning. Furthermore, the 'Critic' network
serves as a verification system for both offline and online
instruction. Using popular benchmarks, The suggested
tracker gets contrasted to certain state of art trackers, as
well as the stimulation finding reveal that it performs well
in actual.
[7] Hybrid Task Cascade for Instance Segmentation
At instance level, segmentation process is basic
computer vision job which identifies objects per-pixel. In
real-world settings like automated driving and video
surveillance, precise and reliable feature extraction is
challenging to achieve. Cascade would be a basic but
efficient design which has increased results on a wide
range of workloads. A basic Cascade R-CNN and Mask R-
CNN combination produces just a little boost. The secret to
good instance segmentation cascade would be to
completely use inverse interaction across detection as well
as segmentation so that to discover a more effective
technique.
They propose Hybrid Task Cascade (HTC) in this paper,
that is different in two key ways: (i) it intertwines these
two tasks for simultaneous multi-stage computation, rather
than conducting cascaded refinement on them individually;
and (ii) employs a convolutional section in order to give
spatial features, that help distinguish difficult frames in
cluttered background. Bounding box analysis plus masked
predictions is coupled inside a multi-tasking way at each
step of HTC. At certain stages, easily applicable within its
masked sections are also given - the masked characteristics
in every step are combined and supplied to the next. The
whole design increases data flow inside the activities as
well as stages, resulting in good refined predictions at all
levels and more reliable forecasts overall. HTC is simple
in setting up & may be programmed from beginning to end.
It gained 2.6 % & 1.4 % greater masked AP than that of the
Masked R-CNN as well as Cascade Masked R-CNN
benchmarks, respectively, on difficult COCO benchmark.

[8] MMDetection: Open MMLab Detection Toolbox and
Benchmark
CV tasks like target recognition & instance segmentation
are both fundamental. Also the detection framework's
pipeline is typically more complicated than that of
classification jobs, and various implementation parameters
might produce drastically different results. In this paper
they are having the goal of providing a high-quality
codebase and unified benchmark, for this they have built
MMDetection model. Major features of MMDetection (1)
Modular design - They split the detection framework into
separate components, allowing users to quickly build a
customised object detection framework by combining
different modules. (2) Out-of-the-box support for multiple
frameworks. Popular and current detection frameworks
are supported by the toolbox. (3) High efficiency - GPUs
handle all fundamental bbox and mask operations. Other
codebases, such as Detectron, maskrcnn-benchmark, and
SimpleDet, have training speeds that are quicker or
equivalent. (4) State-of-the-art - The toolbox is based on
the software created by the MMDet team, who won the
2018 COCO Detection Challenge.
[9] A Cost Effective Tracking System for Small
Unmanned Aerial Systems
The paper describes a object tracking system for
UAV systems is described. It is built on object recognition
with a sum of absolute differences similarity measure.
Every algorithm loop contains a template change that
incorporates new data so that the former stays valid,
allowing the track to proceed even if the object's
appearance changes. It also enables the human operator to
enter targets manually and receive data from other picture
processing units..
The paper supports the importance of uav vision model
suggesting a small, imprecise, non-gyro-stabilized
architectural method for detecting & monitoring many
targets in real-time by not making any rigid assertions on
trajectories, depending onto limited input of their
surroundings and also with avoiding necessity motion
compensation, which usually necessitates the need for a
inertial measurement unit (IMU). This approach eliminates
the additional expense & cost of a binocular model and
those of gyro-stabilized turret instead opting for small,
unstable system. N notion that camera doesn't always need
to be corrected for such model to work reduces the
building & configuration procedure even further. Even
more simplifies the assembling and setup procedure. The
idea can run in real time on low-cost, low-power computer
systems thanks to its ability to track several things at a
cheap computational cost. It's highly crucial because
automated driving vehicles with restricted payload
capacity, whom it is impractical to have a substantial
energy backup for sustaining the computer. Despite the fact
that current advances in recent computer architecture &
powerful batteries assurances that meet the issue by
supplying more productive, advanced machineries and
greater power intensity battery, a urge of a simpler style to
vision systems integration will almost definitely stay
important. This is particularly true for smaller unmanned
aerial systems, such as the Maxxi Joker 2 system employed
in this research.
[10] The research on visual object tracking algorithm
based on an adaptive combination kernel
The research suggested a visual object tracking
approach based on the Adaptive Combination Kernel to
increase the resilience to intricate transitions of multiple
objects and a complex backdrop picture. The object
tracking approach has indeed been divided down into
separate subtasks to approximate the object's details:
Translation Filter as well as Scale Filter. At commence, the
Translation Kernel Tracker utilizes a new Linear Kernel
Filter as well as Gaussian Kernel Filter(GKF) pair. The goal
function, which incorporates not just overfitting problem
but also the highest amount of response output for every
kernel, was used to determine the weight coefficients in
both the Linear Kernel filter as well as GKF. The advantages
of both the local plus global kernels are combined in the
Adaptive Combination Kernel. Next, the tracker position
was identified by using response result of the dynamic
combination kernel correlation filter. Furthermore, the
interpretation filter has been constructed with a scene-
adaptive training data depending on the highest response
score. The effective learning speed can be used to update
the translating filters. Lastly, the item scale was estimated
using a 1D magnitude filters. Compared to previous
techniques, the proposed algorithm is more robust to
deformation and occlusion.
[11] Dual-Channel Convolutional Neural Networks-
Based Image Super-Resolution Algorithm
By using picture super-resolution technique on single
network, it's really challenging in accomplishing
combined high-quality pattern reconstruction and quick
converging. To overcome the drawbacks of earlier
approaches, the current study proposes a picture high-
resolution strategy built upon dual-channel CNN (DCCNN).
A deep tunnel as well as a shallow tunnel were created in
the system model's novel architecture. The shallow
tunnel was largely employed to maintain the original
picture's general shape, whereas the deep tunnel has been
used to retrieve specific feature data. To begin, during
feature extraction phase, the leftover frame was altered, as
well as the channel's nonlinear mapping capability was
increased. After the characteristic mapping scale was
reduced, the picture's effective features were recovered.

During up sampling step, deconvolutional kernel's
general variables are tweaked, and high
bandwidth network degradation were minimised. High-
resolution texture area may get recreated iteratively
employing long as well as the small data chunks in
recreation steps, increasing texture information recovery
even further. Second, the convolutional kernel of narrow
network was adjusted for decreasing the variables, Suring
that the whole contour of picture is recovered & the
channel narrowed faster. Lastly, double channel error
rate is simultaneously changed to increase capability to fit
features in terms of achieving a most recent high-resolution
picture result.
[12] Enhanced Hierarchical Principal Component
Analysis Method for Detecting Saliency
A saliency identification technique formed upon
Hierarchical Principal Component Analysis (HPCA) was
created with in study to solve the challenges of existing
important item identification approaches, such as severe
environmental noise, less accuracy, and high processing
performance. After transforming the RGB picture to
monochrome, the monochrome image was split in eight
layers and used the digital surface stratification approach.
Important subject data correlates to the layered image
characteristics in each picture layer. Second, the
monochrome picture is reinitialised using the
monochrome colour conversion technique, which uses the
initial image's colour layout like a source images, resulting
in a tiered picture that not only represents the initial
structural characteristics and also effeciently conserves the
initial picture's color information. Moreover, Principle
Component Analysis (PCA) was used on layered picture to
identify the structural and colour differential features of
every tier with in principle component orientation.
To get a saliency layout having great resilience as well
as to improve our findings further, two characteristics have
been combined: recognized prviously were added to
picture organisation, which may pinpoint its photograph's
topic around the image's centre. Finally, the entropy
computation was used to produce the idealized picture as
from multilayered saliency layout; the best layout contains
the lowest background as well as most clearly saliency
entities compared to the others.
[13] Online Multi-Object Tracking Using CNN-based
Single Object Tracker with Spatial-Temporal Attention
Mechanism
A CNN-based architecture for online MOT is proposed in
this research. This framework makes use of the advantages
of single object trackers When it comes to altering the
shape of models & searching for the objective within next
frame, There are concerns with processing efficiency and
occlusion-induced drifting outcomes when employing the
single object tracker to MOT. The given approach achieves
computational efficiency besides wanting to share
characteristics and using ROI-Pooling
for obtaining individual characteristic to every target. The
appearance model is adjusted in every target using any
target-specific CNN layers learned online. Introduction
of the spatial-temporal attention mechanism (STAM) in the
framework was carried out to deal with drift caused by
occlusion and target engagement. A target's visibility plan
was learnt and used to infer its spatial attention plan.
Characteristics are then weighted using the spatial
attention map. Furthermore, the occlusion state can be
evaluated using the visibility mapping, that regulates a
continuous updating mechanism using weighted loss upon
training samples having varying occlusion states over
multiple frames. It's possible to think of it as a temporal
attention process. On the rigorous MOT15 and MOT16
benchmark datasets, the suggested approach obtains 34.3
percent and 46.0 percent in MOTA, respectively.
[14] Deep Learning in Video Multi-Object Tracking
Although the target's presence is understood in
advance in SOT, MOT needs the detection stage for
detecting the objects which might leave but rather reenter
the frame. One of most difficult aspect of monitoring many
objects at once is the numerous occlusions as well as
interactions amongst target, which might sometimes seem
toward being identical. Tracking-by-detection is the
standard strategy used in MOT algorithms: The tracking
procedure is guided by a collection of detections (i.e.
enclosing box denoting all objects inside picture) extracted
out from image sequences.
The following steps are found in the great majority of
MOT algorithms:
• Object detection stage: The target identification
method analyses every input video frame using bounding
boxes to discover target belonging here to target class,
often called by 'detections' as in domain of MOT..
• Extraction of features prediction stage: One or
many feature retrieving techniques investigate these
identification plus tracklets in order to obtain information
about look, movement, and interactivity. Each monitored
target's upcoming location may be predicted using a
motion prediction.
• Affinity stage: By giving the same ID for
detections which indicate the same target,
similarity metrics are utilised to correlate detections with
tracklets pertaining to target.
• Association stage: By giving the same ID for
identification which indicate the same object,

similarity metrics have been used to correlate
identification with tracklets pertaining to the target.
[15] Joint Multi-view Face Alignment in the Wild
In this paper, Using face detection model followed with
a deformable model fitting on the bounding box is the de
facto approach to estimate facial landmarks. It entails two
major issues:
 A detection & deformable fitting processes were
done separately, as well as the detector may or
may not offer the optimum initialization to fitting
step,
 A look of a face varies greatly with postures,
making changeable facial fitting extremely difficult,
necessitating the usage of several images.
They demonstrate the first joint. To the best of their
knowledge, multiview convolutional networks can handle
significant postural variances across facial with in field and
neatly integrate facial identification and facial landmark
localisation tasks. Currently present combined
facial detection & landmark positioning methods mainly
evaluate limited number of landmarks. Their technique can
recognise as well as match a huge proportion of markers in
mid-frontal (68 landmarks) as well as profiles (39
landmarks) faces. They put their system through its paces
on range of samples, namely COFW, 300W, IBUG, as well as
the current Menpo Standard both for mid-frontal as well
as profile faces. There is considerable enhancement over
state of art methods in deformable face recognition on
300VW benchmark. Also on FDDB and MALF benchmark,
they also gives state of the art face recognition results.
[16] Object Detection from Video Taken by Drone
through CNN
The study used CNN to recognise objects in footage
taken by the drone. They suggest employing CNNs in this
work to allow drones can discriminate between different
entity types like buildings, automobiles, trees, and humans.
Despite the fact that convolutional neural networks are
computationally expensive, they are trained with smaller
picture datasets using the approach transfer learning. They
used TensorFlow's sophisticated object detection API in
their project, which allowed them to quickly build a new
model and deploy it for detection. According to the data,
the two models' detection rate for houses, trees, cars, &
pedestrians is fairly good, with just an mean of more than
85 % and a peak of 99 percent. They compared the memory
consumption, speed, and accuracy of two current object
detectors in an experiment. In comparison to Faster RCNN
models, SSD concepts give more importance to size, ratio,
as well as predicted sample location, with an average frame
time of 115 ms and a low target identification rate. R-CNN
that is faster and precise but also discovers many items in
the image; about 95% among all targets in frame can be
recognised, yet every frame takes on average 140
milliseconds. In general, their experiments show that
SSD outperforms the HDD model in the long term.
[17] Rich feature hierarchies for accurate object
detection and semantic segmentation
According to this paper, Object detection performance
has been stagnant in previous years. Complex ensembles
combine several low-level picture features with high-level
context from object detectors and scene classifiers to
produce the best results. This research proposes a simple
and scalable object detection technique that improves on
the best recent findings on PASCAL VOC 2012 by 30
percent. They were able to attain this level of success
because to two key observations. This one was to localise
and segment objects using bottom-up region
recommendations utilising high-capacity CNN. The 2nd is a
strategy of teaching large CNNs when labelled training data
is limited. They demonstrate that pre-training the network
for an auxiliary task with a large
amount of information (object classification) and
afterwards tuning the system for primary goal with
minimal input is very effective (detection). They believe
that the "guided pre-training/domain-specific finetuning"
approach will work well for a range of vision difficulties
with little data. They end by emphasising how remarkable
it is that they were able to obtain these results utilising a
combination of traditional computer vision(CV) as well as
deep learning(DL) tools. Instead being antagonistic paths of
rational research, they are closely intertwined.
[18] Deep Residual Learning for Image Recognition
According to paper, hierarchical neural networks
are difficult to be trained. They offer an residual learning
strategy for significantly deeper training models than
earlier used models. Researchers specifically reformulate
the levels as training residual methods with relation to the
levels inputs, rather than learning unreferenced functions.
Researchers presented significant experimental proof
suggesting that residual networks is simpler to implement
and that increasing complexity can boost performance.
Authors used the ImageNet data to evaluate residual nets
having put to 152 levels of complexity, that are 8 layers
deeper to VGG networks whereas have less intricacy. The
aggregation of such residue nets scores 3.57 % error just
on ImageNet test set. The work won 1st spot with in
ILSVRC 2015 classification problem. CIFAR-10 study using
100 - 1000 levels is shown. The complexity in depictions is
crucial for many image identification tasks. Only because of
their exceptionally deep depiction can they accomplish an
28 % notable improvement on COCO object recognition
samples. Their contributions towards ILSVRC & COCO 2015

contests used deep residual networks as the basis, and thus
won 1st spot in the ImageNet identification, ImageNet
localisation, COCO recognition, as well as in COCO
segmentation tasks.
REFERENCES
1. Bernardin K, Stiefelhagen R (2008) Evaluating multiple
object tracking performance: the CLEAR MOT metrics.
EURASIP Journal on Image and Video Processing, 2008,
1–10
2. Bertinetto L, Valmadre J, Henriques JF, Vedaldi A, Torr
PH (2016) Fully-convolutional siamese networks for
object tracking. In European conference on computer
vision (pp. 850–865). Springer, Cham
3. Bewley A, Ge Z, Ott L, Ramos F, Upcroft B (2016) Simple
online and realtime tracking. In 2016 IEEE International
Conference on Image Processing (ICIP) (pp. 3464–
3468). IEEE
4. Bochinski E, Eiselein V, Sikora T (2017) High-speed
tracking-by-detection without using image information.
In 2017 14th IEEE International Conference on
Advanced Video and Signal Based Surveillance (AVSS)
(pp. 1–6). IEEE
5. Cai Z, Vasconcelos N (2018) Cascade R-CNN: Delving into
high quality object detection. In Proceedings of the IEEE
conference on computer vision and pattern recognition
(pp. 6154–6162)
6. Chen B, Wang D, Li P, Wang S, Lu H (2018) Real-
time'Actor-Critic'Tracking. In Proceedings of the
European Conference on Computer Vision (ECCV) (pp.
318–334)
7. Chen K, Pang J, Wang J, Xiong Y, Li X, Sun S ... Loy CC
(2019) Hybrid task cascade for instance segmentation.
In Proceedings of the IEEE conference on computer
vision and pattern recognition (pp.4974–4983)
8. Chen K, Wang J, Pang J, Cao Y, Xiong Y, Li X, ... Zhang Z
(2019) MMDetection: Open mmlab detection toolbox
and benchmark. arXiv preprint arXiv:1906.07155
9. Chen Y, Wang J, Liu S, Chen X, Xiong J, Xie J, Yang K
(2019)Multiscale fast correlation filtering tracking
algorithm based on a feature fusion model. Concurrency
and Computation: Practice and Experience, e5533
10. Chen Y, Wang J, Xia R, Zhang Q, Cao Z, Yang K (2019)
The visual object tracking algorithm research based on
adaptive combination kernel. J Ambient Intell
Humanized Comput 10(12):4855–4867
11. Chen Y, Wang J, Chen X, Sangaiah AK, Yang K, Cao Z
(2019) Image super-resolution algorithm based on
dual-channel convolutional neural networks. Appl Sci
9(11):2316
12. Chen Y, Tao J, Zhang Q, Yang K, Chen X, Xiong J, ... Xie J
(2020) Saliency Detection via the Improved Hierarchical
Principal Component Analysis Method. Wireless
Communications and Mobile Computing, 2020
13. Chu Q, Ouyang W, Li H, Wang X, Liu B, Yu N (2017)
Online multi-object tracking using CNN-based single
object tracker with spatial-temporal attention
mechanism. In Proceedings of the IEEE International
Conference on Computer Vision (pp. 4836–4845)
14. Ciaparrone G, Sánchez FL, Tabik S, Troiano L, Tagliaferri
R, Herrera F (2020) Deep learning in video multi-object
tracking: A survey. Neurocomputing 381:61–88
15. Deng J, Trigeorgis G, Zhou Y, Zafeiriou S (2019) Joint
multi-view face alignment in the wild. IEEE
Transactions on Image Processing 28(7):3636–3648
16. Fan, D. P., Wang, W., Cheng, M. M., & Shen, J. (2019).
Shifting more attention to video salient object detection.
In Proceedings of the IEEE conference on computer
vision and pattern recognition (pp. 8554–8564).
17. Girshick R, Donahue J, Darrell T, Malik J (2014) Rich
feature hierarchies for accurate object detection and
semantic segmentation. In Proceedings of the IEEE
conference on computer vision and pattern recognition
(pp. 580–587)
18. He K, Zhang X, Ren S, Sun J (2016) Deep residual
learning for image recognition. In Proceedings of the
IEEE conference on computer vision and pattern
recognition (pp. 770–778)

A Literature Review on Vehicle Detection and Tracking in Aerial Image Sequence using Deep Learning Technique

More Related Content

Similar to A Literature Review on Vehicle Detection and Tracking in Aerial Image Sequence using Deep Learning Technique (20)

More from IRJET Journal (20)

Recently uploaded (20)

A Literature Review on Vehicle Detection and Tracking in Aerial Image Sequence using Deep Learning Technique