Adaptive Semantic Segmentation for Unmanned Surface Vehicle Navigation

Zhan, Wenqiang; Xiao, Changshi; Wen, Yuanqiao; Zhou, Chunhui; Yuan, Haiwen; Xiu, Supu; Zou, Xiong; Xie, Cheng; Li, Qiliang

doi:10.3390/electronics9020213

Cited by 16 publications

(10 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Later, some researchers [30,31] extended the urban landscape segmentation by designing a network that classifies pixels into water, land, or sky using a CRF method. This is particularly relevant for urban areas with diverse landscapes, from water bodies to green spaces [32].…”

Section: Literature Reviewmentioning

confidence: 99%

Urban Aquatic Scene Expansion for Semantic Segmentation in Cityscapes

Yue,

Lo,

et al. 2024

Urban Science

View full text Add to dashboard Cite

In urban environments, semantic segmentation using computer vision plays a pivotal role in understanding and interpreting the diverse elements within urban imagery. The Cityscapes dataset, widely used for semantic segmentation in urban scenes, predominantly features urban elements like buildings and vehicles but lacks aquatic elements. Recognizing this limitation, our study introduces a method to enhance the Cityscapes dataset by incorporating aquatic classes, crucial for a comprehensive understanding of coastal urban environments. To achieve this, we employ a dual-model approach using two advanced neural networks. The first network is trained on the standard Cityscapes dataset, while the second focuses on aquatic scenes. We adeptly integrate aquatic features from the marine-focused model into the Cityscapes imagery. This integration is carefully executed to ensure a seamless blend of urban and aquatic elements, thereby creating an enriched dataset that reflects the realities of coastal cities more accurately. Our method is evaluated by comparing the enhanced Cityscapes model with the original on a set of diverse urban images, including aquatic views. The results demonstrate that our approach effectively maintains the high segmentation accuracy of the original Cityscapes dataset for urban elements while successfully integrating marine features. Importantly, this is achieved without necessitating additional training, which is a significant advantage in terms of resource efficiency.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Urban Aquatic Scene Expansion for Semantic Segmentation in Cityscapes

Yue,

Lo,

et al. 2024

Urban Science

View full text Add to dashboard Cite

show abstract

“…This is a computer vision technique that divides an image into segments and that has found application in many aspects of image processing. In [47], the authors proposed adaptive semantic segmentation for visual perception of water scenes. Adaptive filtering and progressive segmentation for shoreline detection are shown in [48].…”

Section: Related Workmentioning

confidence: 99%

Computer Vision-Based Position Estimation for an Autonomous Underwater Vehicle

Zalewski,

Hożyń

2024

Remote Sensing

View full text Add to dashboard Cite

Autonomous Underwater Vehicles (AUVs) are currently one of the most intensively developing branches of marine technology. Their widespread use and versatility allow them to perform tasks that, until recently, required human resources. One problem in AUVs is inadequate navigation, which results in inaccurate positioning. Weaknesses in electronic equipment lead to errors in determining a vehicle’s position during underwater missions, requiring periodic reduction of accumulated errors through the use of radio navigation systems (e.g., GNSS). However, these signals may be unavailable or deliberately distorted. Therefore, in this paper, we propose a new computer vision-based method for estimating the position of an AUV. Our method uses computer vision and deep learning techniques to generate the surroundings of the vehicle during temporary surfacing at the point where it is currently located. The next step is to compare this with the shoreline representation on the map, which is generated for a set of points that are in a specific vicinity of a point determined by dead reckoning. This method is primarily intended for low-cost vehicles without advanced navigation systems. Our results suggest that the proposed solution reduces the error in vehicle positioning to 30–60 m and can be used in incomplete shoreline representations. Further research will focus on the use of the proposed method in fully autonomous navigation systems.

show abstract

“…As for shoreline detection, some researchers have attempted to introduce semantic segmentation techniques into this field [15] by first extracting the water surface area with a semantic segmentation network and using an edge detection algorithm on the obtained result to detect the shoreline. Based on this, some works have improved existing semantic segmentation models, making them more suitable for water surface area segmentation [16][17][18], and other studies have introduced pretrained methods [19] or the use of transfer learning [20] to solve the problem of an insufficient amount of data for training, as well as use self-supervised training approaches [21] to address the problem of insufficient labeled data. The trained models have strong robustness, which is good for solving the interference of environmental factors present in traditional image-based approaches but still cannot address the processing speed problem due to the use of a large network architecture, the large scale of the feature maps, etc.…”

Section: Introductionmentioning

confidence: 99%

Improved UNet-Based Shoreline Detection Method in Real Time for Unmanned Surface Vehicle

Zhao

Song

Gong

et al. 2023

JMSE

View full text Add to dashboard Cite

Accurate and real-time monitoring of the shoreline through cameras is an invaluable guarantee for the safety of near-shore navigation and berthing of unmanned surface vehicles; existing shoreline detection methods cannot meet both these requirements. Therefore, we propose an improved shoreline detection method to detect shorelines accurately and in real time. We define shoreline detection as the combination of water surface area segmentation and edge detection, the key to which is segmentation. To detect shorelines accurately and in real time, we propose an improved U-Net for water segmentation. This network is based on U-Net, using ResNet-34 as the backbone to enhance the feature extraction capability, with a concise decoder integrated attention mechanism to improve the processing speed while ensuring the accuracy of water surface segmentation. We also introduce transfer learning to improve training efficiency and solve the problem of insufficient data. When obtaining the segmentation result, the Laplace edge detection algorithm is applied to detect the shoreline. Experiments show that our network achieves 97.05% MIoU and 40 FPS with the fewest parameters, which is better than mainstream segmentation networks, and also demonstrate that our shoreline detection method can effectively detect shorelines in real time in various environments.

show abstract

Adaptive Semantic Segmentation for Unmanned Surface Vehicle Navigation

Cited by 16 publications

References 35 publications

Urban Aquatic Scene Expansion for Semantic Segmentation in Cityscapes

Urban Aquatic Scene Expansion for Semantic Segmentation in Cityscapes

Computer Vision-Based Position Estimation for an Autonomous Underwater Vehicle

Improved UNet-Based Shoreline Detection Method in Real Time for Unmanned Surface Vehicle

Contact Info

Product

Resources

About