Multiple Object Tracking: Case of Aircraft Detection and Tracking

Ellouze, Ameni; Ksantini, Mohamed; Delmotte, François; Karray, Mohamed

doi:10.1109/ssd.2019.8893202

Cited by 8 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Numerous traditional machine learning methods such as SVM [31,32], RF [10,33], and boosted DTs [34,35] have been widely used for feature extraction from remote sensing data. In recent years, deep learning has shown increasing success in a variety of computational vision tasks and is increasingly being used in remote sensing applications also [15,36,37]. The authors of [14] used deep learning methods in both pixel-based and object-based remote sensing applications and showed that they demonstrated superior performance over traditional machine learning methods.…”

Section: Remote Sensing Analysis Methodsmentioning

confidence: 99%

A Generic Self-Supervised Learning (SSL) Framework for Representation Learning from Spectral–Spatial Features of Unlabeled Remote Sensing Imagery

Zhang,

Han

2023

Remote Sensing

View full text Add to dashboard Cite

Remote sensing data has been widely used for various Earth Observation (EO) missions such as land use and cover classification, weather forecasting, agricultural management, and environmental monitoring. Most existing remote-sensing-data-based models are based on supervised learning that requires large and representative human-labeled data for model training, which is costly and time-consuming. The recent introduction of self-supervised learning (SSL) enables models to learn a representation from orders of magnitude more unlabeled data. The success of SSL is heavily dependent on a pre-designed pretext task, which introduces an inductive bias into the model from a large amount of unlabeled data. Since remote sensing imagery has rich spectral information beyond the standard RGB color space, it may not be straightforward to extend to the multi/hyperspectral domain the pretext tasks established in computer vision based on RGB images. To address this challenge, this work proposed a generic self-supervised learning framework based on remote sensing data at both the object and pixel levels. The method contains two novel pretext tasks, one for object-based and one for pixel-based remote sensing data analysis methods. One pretext task is used to reconstruct the spectral profile from the masked data, which can be used to extract a representation of pixel information and improve the performance of downstream tasks associated with pixel-based analysis. The second pretext task is used to identify objects from multiple views of the same object in multispectral data, which can be used to extract a representation and improve the performance of downstream tasks associated with object-based analysis. The results of two typical downstream task evaluation exercises (a multilabel land cover classification task on Sentinel-2 multispectral datasets and a ground soil parameter retrieval task on hyperspectral datasets) demonstrate that the proposed SSL method learns a target representation that covers both spatial and spectral information from massive unlabeled data. A comparison with currently available SSL methods shows that the proposed method, which emphasizes both spectral and spatial features, outperforms existing SSL methods on multi- and hyperspectral remote sensing datasets. We believe that this approach has the potential to be effective in a wider range of remote sensing applications and we will explore its utility in more remote sensing applications in the future.

show abstract

Section: Remote Sensing Analysis Methodsmentioning

confidence: 99%

A Generic Self-Supervised Learning (SSL) Framework for Representation Learning from Spectral–Spatial Features of Unlabeled Remote Sensing Imagery

Zhang,

Han

2023

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Due to the existence of several objects in our lives, for instance, humans being (pedestrians [28], sport players [6], shoppers), vehicles (cars [29], motorcycles, buses), animals (fishes [30], birds [31], cats), other (cells [32], insects), and so on; many scientific studies have explored implementing multiple object tracking using different algorithms. These initiatives are divided into research areas or domains; see Fig.…”

Section: Research Areasmentioning

confidence: 99%

A WSM-based Comparative Study of Vision Tracking Methodologies

Bouraya¹,

Belangour²

2021

IJACSA

View full text Add to dashboard Cite

Vision tracking is a key component of a video sequence. It is the process of locating single or multiple moving objects over time using one or many cameras. The latter's function consists of detecting, categorizing, and tracking. The development of the trustworthy solution for video sequence analysis opens up new horizons for a variety of applications, including intelligent transportation systems, biomedical, agriculture, human-machine interaction, augmented reality, video surveillance, robots, and many crucial research areas. To make efficient models, there are challenges in video observation to deal with, such as problems with the environment, light variation, pose variation, motion blur, clutter, occlusion, and so on. In this paper, we present several techniques that addressed the issues of detecting and tracking multiple targets on video sequences. The proposed comparative study relied on different methodologies. This paper's purpose is to list various approaches, classify them, and compare them, using the Weighted Scoring Model (WSM) comparison method. This includes studying these algorithms, selecting relevant comparison criteria, assigning weights for each criterion, and lastly computing scores. The obtained results of this study will reveal the strong and weak points of each algorithm mentioned and discussed.

show abstract

“…Extracting features related to black smoke emissions from vehicles and combining them with classifiers can enable automatic detection of black smoke. Among the various methods, deep neural networks have been used to build object detection models that are categorized into two-stage and single-stage models [1][2][3][4]. Cao et al [5] utilized the Incep-tionv3 convolutional neural network to capture spatial information in surveillance videos with suspected black smoke frames, while a long short-term memory network learned the temporal dependencies between frames.…”

Section: Introductionmentioning

confidence: 99%

A Lightweight Model for Real-Time Detection of Vehicle Black Smoke

Chen,

Wang,

Zhai

2023

Sensors

View full text Add to dashboard Cite

This paper discusses the application of deep learning technology in recognizing vehicle black smoke in road traffic monitoring videos. The use of massive surveillance video data imposes higher demands on the real-time performance of vehicle black smoke detection models. The YOLOv5s model, known for its excellent single-stage object detection performance, has a complex network structure. Therefore, this study proposes a lightweight real-time detection model for vehicle black smoke, named MGSNet, based on the YOLOv5s framework. The research involved collecting road traffic monitoring video data and creating a custom dataset for vehicle black smoke detection by applying data augmentation techniques such as changing image brightness and contrast. The experiment explored three different lightweight networks, namely ShuffleNetv2, MobileNetv3 and GhostNetv1, to reconstruct the CSPDarknet53 backbone feature extraction network of YOLOv5s. Comparative experimental results indicate that reconstructing the backbone network with MobileNetv3 achieved a better balance between detection accuracy and speed. The introduction of the squeeze excitation attention mechanism and inverted residual structure from MobileNetv3 effectively reduced the complexity of black smoke feature fusion. Simultaneously, a novel convolution module, GSConv, was introduced to enhance the expression capability of black smoke features in the neck network. The combination of depthwise separable convolution and standard convolution in the module further reduced the model’s parameter count. After the improvement, the parameter count of the model is compressed to 1/6 of the YOLOv5s model. The lightweight vehicle black smoke real-time detection network, MGSNet, achieved a detection speed of 44.6 frames per second on the test set, an increase of 18.9 frames per second compared with the YOLOv5s model. The mAP@0.5 still exceeded 95%, meeting the application requirements for real-time and accurate detection of vehicle black smoke.

show abstract

Multiple Object Tracking: Case of Aircraft Detection and Tracking

Cited by 8 publications

References 4 publications

A Generic Self-Supervised Learning (SSL) Framework for Representation Learning from Spectral–Spatial Features of Unlabeled Remote Sensing Imagery

A Generic Self-Supervised Learning (SSL) Framework for Representation Learning from Spectral–Spatial Features of Unlabeled Remote Sensing Imagery

A WSM-based Comparative Study of Vision Tracking Methodologies

A Lightweight Model for Real-Time Detection of Vehicle Black Smoke

Contact Info

Product

Resources

About