Detection of Cities Vehicle Fleet Using Yolo V2 and Aerial Images

Lechgar, Hatim; Bekkar, H.; Rhinane, Hassan

doi:10.5194/isprs-archives-xlii-4-w12-121-2019

Cited by 9 publications

(9 citation statements)

References 4 publications

(5 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since all sensor data processing takes place on embedded computer boards and yet a high frame rate is required, the focus here is on the use of one-stage methods. The YOLO detector is based on this concept and has been successfully used several times in the literature for the application considered here [ 9 , 37 , 39 , 40 ]. Therefore, and because it represents the properties of common algorithms very well, the YOLOv3 [ 17 ] detector was chosen as the test algorithm for the investigations carried out.…”

Section: Methodsmentioning

confidence: 99%

“…To investigate these issues, vehicle detection on aerial imagery based on two-dimensional bounding boxes has been selected as an exemplary application in this paper. This is a current research area especially in UAV missions [ 9 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 ], furthermore, it forms the basis for more advanced applications such as object counting or tracking, and enables a direct performance comparison through appropriate metrics.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Based Vehicle Detection on Real and Synthetic Aerial Images: Training Data Composition and Statistical Influence Analysis

Krump

Stütz

2023

Sensors

View full text Add to dashboard Cite

The performance of deep learning based algorithms is significantly influenced by the quantity and quality of the available training and test datasets. Since data acquisition is complex and expensive, especially in the field of airborne sensor data evaluation, the use of virtual simulation environments for generating synthetic data are increasingly sought. In this article, the complete process chain is evaluated regarding the use of synthetic data based on vehicle detection. Among other things, content-equivalent real and synthetic aerial images are used in the process. This includes, in the first step, the learning of models with different training data configurations and the evaluation of the resulting detection performance. Subsequently, a statistical evaluation procedure based on a classification chain with image descriptors as features is used to identify important influencing factors in this respect. The resulting findings are finally incorporated into the synthetic training data generation and in the last step, it is investigated to what extent an increase of the detection performance is possible. The overall objective of the experiments is to derive design guidelines for the generation and use of synthetic data.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep Learning Based Vehicle Detection on Real and Synthetic Aerial Images: Training Data Composition and Statistical Influence Analysis

Krump

Stütz

2023

Sensors

View full text Add to dashboard Cite

show abstract

“…While large companies use reports of the number of vehicles in their parking lots as a crucial variable in developing models to predict store earnings over a given period (Cisek et al, 2017), academia has not paid much attention to the subject. Most papers focus solely on developing models of parking lot classification without emphasis on practical applications (Lechgar et al, 2019;Minetto et al, 2021).…”

Section: Satellite Imagesmentioning

confidence: 99%

Satellite imagery and machine learning for channel member selection

Brei,

Rech,

Bozkaya

et al. 2023

IJRDM

View full text Add to dashboard Cite

PurposeThis study aims to propose a new method to predict retail store performance using publicly available satellite imagery data and machine learning (ML) algorithms. The goal is to provide manufacturers and other practitioners with a more accurate and objective way to assess potential channel members and mitigate information asymmetry in channel selection and negotiation.Design/methodology/approachThe authors developed an open-source approach using publicly available Google satellite imagery and ML algorithms. A computer vision algorithm was used to count cars in store parking lots, and the data were processed with a CNN. Linear regression and various ML algorithms were used to estimate the relationship between parked cars and sales.FindingsThe relationship between parked cars and sales was nonlinear and dependent on the type of channel member. The best model, a Stacked Ensemble, showed that parking lot occupancy could accurately predict channel member performance.Research limitations/implicationsThe proposed approach offers manufacturers a low-cost and scalable solution to improve their channel member selection and performance assessment process. Using satellite imagery data can help balance the marketing channel planning process by reducing information asymmetry and providing a more objective way to assess potential partners.Originality/valueThis research is unique in proposing a method based on publicly available satellite imagery data to assess and predict channel member performance instead of forward-looking sales at the firm and industry levels like previous studies.

show abstract

“…In deep learning, a computer model learns to perform classification tasks using different types of information such as texts, sounds or images. The models implemented in deep learning are formed using a large number of tagged data and neural network architectures that contain various layers [8]. The word "deep" refers to the number of hidden layers in the neural network.…”

Section: General Considerationsmentioning

confidence: 99%

Deep Learning and YOLOv3 Systems for Automatic Traffic Data Measurement by Moving Car Observer Technique

Guerrieri

Parla

2021

Infrastructures

View full text Add to dashboard Cite

Macroscopic traffic flow variables estimation is of fundamental interest in the planning, designing and controlling of highway facilities. This article presents a novel automatic traffic data acquirement method, called MOM-DL, based on the moving observer method (MOM), deep learning and YOLOv3 algorithm. The proposed method is able to automatically detect vehicles in a traffic stream and estimate the traffic variables flow q, space mean speed vs. and vehicle density k for highways in stationary and homogeneous traffic conditions. The first application of the MOM-DL technique concerns a segment of an Italian highway. In the experiments, a survey vehicle equipped with a camera has been used. Using deep learning and YOLOv3 the vehicles detection and the counting processes have been carried out for the analyzed highway segment. The traffic flow variables have been calculated by the Wardrop relationships. The first results demonstrate that the MOM and MOM-DL methods are in good agreement with each other despite some errors arising with MOM-DL during the vehicle detection step due to a variety of reasons. However, the values of macroscopic traffic variables estimated by means of the Drakes’ traffic flow model together with the proposed method (MOM-DL) are very close to those obtained by the traditional one (MOM), being the maximum percentage variation less than 3%.

show abstract

Detection of Cities Vehicle Fleet Using Yolo V2 and Aerial Images

Cited by 9 publications

References 4 publications

Deep Learning Based Vehicle Detection on Real and Synthetic Aerial Images: Training Data Composition and Statistical Influence Analysis

Deep Learning Based Vehicle Detection on Real and Synthetic Aerial Images: Training Data Composition and Statistical Influence Analysis

Satellite imagery and machine learning for channel member selection

Deep Learning and YOLOv3 Systems for Automatic Traffic Data Measurement by Moving Car Observer Technique

Contact Info

Product

Resources

About