Review of Building Extraction Methods Based on High-Resolution Remote Sensing Images

Lecture notes in electrical engineering, Journal Year: 2024, Volume and Issue: unknown, P. 499 - 506

Published: Jan. 1, 2024

Language: Английский

---

Regularizing building outlines extracted from remote sensing images by integrating multiple algorithms

Geocarto International, Journal Year: 2024, Volume and Issue: 39(1)

Published: Jan. 1, 2024

Extracting building from remote sensing images is crucial, but the extracted outlines still face issues such as point redundancy and lack of right-angle features, relying on further regularization. This study proposed an integrated regularization method to combine strengths multiple algorithms. 22 metrics were calculated describe geometric characteristic outlines, followed by principal component analysis for selecting key metrics. Further, a supervised learning model was constructed analyze these determine most suitable algorithm among three candidates–rectangle transformation, recursive regression, feature edge reconstruction–for regularizing each building. Experimental results demonstrated that our can adaptively select appropriate based metrics, achieving superior those obtained using algorithms independently. Compared with two existing correction-based methods, excels in preserving orientation, area, shape. Our also has advantages over learning-based methods maintaining orthogonality.

Language: Английский

---

Impact of Call Drop Ratio Over 5G Network

Advances in wireless technologies and telecommunication book series, Journal Year: 2023, Volume and Issue: unknown, P. 201 - 224

Published: March 3, 2023

The 5G network is the main topic of this investigation. expected to be far more advanced than 4G since it makes use three distinct bands spectrum. acronym “5G” refers latest generation wireless communications. With 5G, communications will improve all across world. purpose research examine consequences increasing call drop ratio in networks. In other words, if a user's current session interrupted, they need make new connection continue using service. One area where excels over its predecessors, (and LTE), reducing latency. Also, there would fewer lost calls for individuals utilizing VoIP because uptime have risen significantly. Reduced failure rates lead happier customers.

Language: Английский

---

GAN-Based Sub-Instance Augmentation for Open-Pit Mine Change Detection in Remote Sensing Images

IEEE Transactions on Geoscience and Remote Sensing, Journal Year: 2023, Volume and Issue: 62, P. 1 - 19

Published: Nov. 28, 2023

Remote sensing change detection (CD) for open-pit mines plays a critical role in both mineral development and environmental conservation. The performance of supervised deep-learning-based CD is often limited by the amount available data, necessitating data augmentation. Current augmentation methods mainly focus on image-level, region-level instance-level transformations. However, mining area changes occur at finer sub-instance level, leading to suboptimal existing methods. Therefore, this study proposes novel method named generative adversarial network-based (GSIA). This enables generation realistic diverse samples using unchanged from areas address issue scarcity mine CD. GSIA comprises three steps. In first two steps, achieves sub-instance-level transformations sequentially applying GAN-based local editing labels images areas. third step, constructs synthetic dataset randomly combining bitemporal data. effectiveness was evaluated comparing it with fourteen other five models, outperformed all them. addition, training solely generated achieved an overall accuracy 97.64% F1-score 78.65%, which were comparable real Furthermore, can insufficient correlation between test sets domain adaptation. significant guiding

Language: Английский

---

Python Based Smart Trial Room

Published: Dec. 21, 2022

Other material endeavors hit during these challenging intervals, as the overwhelming majority of things moved to virtual mode and challengers attempted advance beyond one. Every small shop can assist their customers with an internet-based trial room a without danger contracting illness. The advantage using this method is that it saves time effort evaluate clothing accurately. Everything, from food apparel, now available online. Instead travelling shopping malls textile stores, where they have try on items in see if fit correctly how appear themselves, go internet. Furthermore, most limited number rooms are overcrowded. As result, people wait long time, which inconvenient. Trial not so secure no one be certain hidden cameras present. online buying websites extremely safe. People simply choose variety outfits. On other hand, do know piece will them until brought them. by just selecting goods seeing appears article uses python create "Smart Room" allows users garments virtually. This project comprises homepage several clothing-related products comparable e-commerce website. HTML CSS were used Users product like, launch webcam show thing looks Multiple also at same time. major goal experiment greater interaction features clothes increase sales.

Language: Английский

---