Methods and datasets on semantic segmentation for Unmanned Aerial Vehicle remote sensing images: A review

ISPRS Journal of Photogrammetry and Remote Sensing, Journal Year: 2024, Volume and Issue: 211, P. 1 - 34

Published: April 2, 2024

Language: Английский

---

Semantic Segmentation of Aerial Imagery: A Novel Approach Leveraging Hierarchical Multi-scale Features and Channel-based Attention for Drone Applications

International journal of engineering. Transactions B: Applications, Journal Year: 2024, Volume and Issue: 37(5), P. 1022 - 1035

Published: Jan. 1, 2024

Drone semantic segmentation is a challenging task in computer vision, mainly due to inherent complexities associated with aerial imagery. This paper presents comprehensive methodology for drone and evaluates its performance using the ICG dataset. The proposed method leverages hierarchical multi-scale feature extraction efficient channel-based attention Atrous Spatial Pyramid Pooling (ASPP) address unique challenges encountered this domain. In study, of compared several state-of-the-art models. findings research highlight effectiveness tackling segmentation. outcomes demonstrate superiority over models, showcasing potential accurate results contribute advancement drone-based applications, such as surveillance, object tracking, environmental monitoring, where precise crucial. obtained experimental that outperforms these existing approaches regarding Dice, mIOU, accuracy metrics. Specifically, achieves an impressive scores 86.51%, 76.23%, 91.74%, respectively.

Language: Английский

---

Road Extraction in Diverse Urban Environments Using UAV Data and nDSM Perturbations: A Case of Bhopal, India

Remote Sensing Applications Society and Environment, Journal Year: 2025, Volume and Issue: unknown, P. 101465 - 101465

Published: Jan. 1, 2025

Language: Английский

---

Semantic Image Segmentation Employing U-Net-Based Ensemble Model

Advances in computational intelligence and robotics book series, Journal Year: 2025, Volume and Issue: unknown, P. 305 - 328

Published: March 7, 2025

Image segmentation is an important topic in computer vision, playing role wide range of applications such as medical image analysis, scene understanding, tumour boundary extraction among many others. aims to identify groups pixels and parts images that are similar belong together Semantic a classification with labels partitioning the objects. By applying segmentation, we can all objects image. The brain dataset utilizing for BRATS'20, which contains 317 images. proposed ensemble approach combining U-Net variants Mask RCNN models outperforms individual models. While method yielded improved Dice scores, using union six other methods achieved highest accuracy, indicated by superior scores. Specifically, model score 71.10 IoU 81.98. Additionally, demonstrated strong performance terms precision, reaching 84.96, recall value 81.90.

Language: Английский

---

Robust unsupervised domain adaptation by retaining confident entropy via edge concatenation

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 238, P. 122120 - 122120

Published: Oct. 17, 2023

Language: Английский

---

Packaging style design based on visual semantic segmentation technology and intelligent cyber physical system

PeerJ Computer Science, Journal Year: 2023, Volume and Issue: 9, P. e1451 - e1451

Published: July 10, 2023

The integration of image segmentation technology into packaging style design significantly amplifies both the aesthetic allure and practical utility product design. However, conventional algorithm necessitates a substantial amount time for analysis, rendering it susceptible to loss vital features yielding unsatisfactory results. Therefore, this study introduces novel network, G-Lite-DeepLabV3+, which is seamlessly incorporated cyber-physical systems (CPS) enhance accuracy efficiency segmentation. In research, feature extraction network DeepLabV3 replaced with Mobilenetv2, integrating group convolution attention mechanisms proficiently process intricate semantic improve network’s responsiveness valuable characteristics. These adaptations are then deployed within CPS, allowing G-Lite-DeepLabV3+ be integrated processing module CPS. This facilitates remote real-time images in virtual environment.Experimental findings demonstrate that excels at segmenting diverse graphical elements images. Compared original DeepLabV3+ intersection over union (IoU) metric shows remarkable increase 3.1%, while mean pixel (mPA) exhibits an impressive improvement 6.2%. Additionally, frames per second (FPS) experiences significant boost 22.1%. When successfully accomplishes tasks enhanced efficiency, maintaining high levels accuracy.

Language: Английский

---

EvolveNet: Evolving Networks by Learning Scale of Depth and Width

Published: July 26, 2023

Convolutional Neural Networks (CNNs) are largely hand-crafted, which leads to inefficiency in the constructed network. Various other algorithms have been proposed address this issue, but inefficiencies resulting from human intervention not addressed. Our EvolveNet algorithm is a task-agnostic evolutionary search that can find optimal depth and width scales automatically an efficient way. The configurations found using grid search, instead evolved existing This eliminates emanate hand-crafting, thus reducing drop accuracy. framework through large space of subnetworks until suitable configuration found. Extensive experiments on ImageNet dataset demonstrate superiority method by outperforming state-of-the-art methods.

Language: Английский

---

МЕТОД ПОШУКУ ОБ’ЄКТІВ ІНТЕРЕСУ ЗА СПЕКТРАЛЬНИМИ ОЗНАКАМИ НА ЗОБРАЖЕННЯХ З АКТИВНОЇ ОПТИКО-ЕЛЕКТРОННОЇ СИСТЕМИ СПОСТЕРЕЖЕННЯ

Системи управління навігації та зв’язку Збірник наукових праць, Journal Year: 2024, Volume and Issue: 2(76), P. 5 - 10

Published: April 30, 2024

Предметом вивчення в статті є метод пошуку об’єктів інтересу за спектральними ознаками на основі активної оптико-електронної системи, у якій якості джерела випромінювання застосовується набір багатоспектральних лазерних випромінювачів, з подальшою комп’ютерною обробкою отриманого зображення. Метою розробка методу із застосуванням системи Завдання: аналіз особливостей побудови і функціонування активних оптико-електронних систем динамічною спектральною обробкою; активною оптико-електронною системою зображень; математичної моделі, яка дозволяє прийняти рішення про наявність об’єкта шляхом визначення ділянок зображення елементами, що мають найбільше значення яскравості. Використовуваними методами є: методи цифрової обробки зображень, математичного моделювання, теорії оптимізації, аналітичні та емпіричні аналізу зображень після їх обробки. Отримані такі результати. Проаналізовано особливості обробкою. Розроблено зображень. Розроблена математична модель, маючими Висновки. Особливістю розробленого застосування початковому етапі для вимірювання спектрального коефіцієнту відбиття досліджуваної ділянки, який подальшому разом зі характеристиками інтересу, використовується обчисленні апаратної функції (вектора фільтру) забезпечення підвищення контрасту інтересу. На заключному проводиться комп’ютерна обробка метою розміщено об’єкт елементами найбільшим значенням Проведено математичне моделювання зоні спостереження. За результатами було прийнято ділянок, складові елементи яких найвищі показники

---

COMPARATIVE ANALYSIS OF SPECTRAL ANOMALIES DETECTION METHODS ON IMAGES FROM ON-BOARD REMOTE SENSING SYSTEMS

Advanced Information Systems, Journal Year: 2024, Volume and Issue: 8(2), P. 48 - 57

Published: June 4, 2024

The subject matter of the article is methods detecting spectral anomalies on images from remote sensing systems. goal to conduct a comparative analysis for tasks are: main systems; processing systems using basic anomalies; assessment quality monitoring used digital image processing, mathematical apparatus matrix theory, modeling, optimization analytical and empirical comparison. following results are obtained. were analyzed. Processing was carried out. A Conclusions. difference considered revealed by value information indicators - Euclidean distance, Mahalanobis brightness contrast, Kullback-Leibler divergence. Mathematical modeling with relatively “simple” complicated background It established that when searching anomaly an background, method based divergence can be more effective than other methods, but not optimal. When determining several areas high indicators, they should additionally investigated specified in order accurately determine position anomaly.

Language: Английский

---

Semantic segmentation of oblique UAV video based on ConvLSTM in complex urban area

Earth Science Informatics, Journal Year: 2024, Volume and Issue: unknown

Published: June 8, 2024

Language: Английский

---