A object detection and tracking method for security in intelligence of unmanned surface vehicles

Zhang, Wei; Gao, Xian-Zhong; Yang, Chengxiang; Jiang, Feng; Chen, Zhiyuan

doi:10.1007/s12652-020-02573-z

Cited by 19 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, the method proposed in [11] is based on MobileNet for feature extraction and SSD for fast multi-scale detection to achieve realtime marine object detection of high-speed USVs. Zhang et al [12] proposed a method for marine object detection and tracking based on improved YOLOv3 and used their method on a real USV experiment platform. The authors of [13] fused DenseNet in YOLOv3 for robust detection of marine objects under various weather conditions.…”

Section: Object Detection On Water Surfacesmentioning

confidence: 99%

A Millimeter-Wave Radar-Aided Vision Detection Method for Water Surface Small Object Detection

Zhu,

Yang,

Cheng

2023

JMSE

View full text Add to dashboard Cite

Unmanned surface vehicles (USVs) have wide applications in marine inspection and monitoring, terrain mapping, and water surface cleaning. Accurate and robust environment perception ability is essential for achieving autonomy in USVs. Small object detection on water surfaces is an important environment perception task, typically achieved by visual detection using cameras. However, existing vision-based small object detection methods suffer from performance degradation in complex water surface environments. Therefore, in this paper, we propose a millimeter-wave (mmWave) radar-aided vision detection method that enables automatic data association and fusion between mmWave radar point clouds and images. Through testing on real-world data, the proposed method demonstrates significant performance improvement over vision-based object detection methods without introducing more computational costs, making it suitable for real-time application on USVs. Furthermore, the image–radar data association model in the proposed method can serve as a plug-and-play module for other object detection methods.

show abstract

Section: Object Detection On Water Surfacesmentioning

confidence: 99%

A Millimeter-Wave Radar-Aided Vision Detection Method for Water Surface Small Object Detection

Zhu,

Yang,

Cheng

2023

JMSE

View full text Add to dashboard Cite

show abstract

“…The integration and refinement of these algorithms are of significant importance to the promotion of sustainable development within the maritime economy [6]. While infrared thermal imagers may lack detailed imagery, and laser radars are prone to sea wave reflections, visible light cameras offer a cost-effective and robust solution for long-range object detection and have become the most common choice for USVs [7,8].…”

Section: Introductionmentioning

confidence: 99%

“…Current water surface object detection models predominantly focus on lightweight models and the detection of small targets [7,[12][13][14], often neglecting the unique characteristics of water surface objects, such as the submerged nature of their bottom boundaries, as well as the distinctive attributes of water surface imagery, which typically features pronounced separation between the foreground and background. This oversight may limit the effectiveness of these models in accurately detecting and classifying objects within the aquatic environment.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Water Surface Object Detection with Dynamic Task-Aligned Sample Assignment and Attention Mechanisms

Zhao,

Qiu,

Chen

2024

Sensors

View full text Add to dashboard Cite

The detection of objects on water surfaces is a pivotal technology for the perceptual systems of unmanned surface vehicles (USVs). This paper proposes a novel real-time target detection system designed to address the challenges posed by indistinct bottom boundaries and foggy imagery. Our method enhances the YOLOv8s model by incorporating the convolutional block attention module (CBAM) and a self-attention mechanism, examining their impact at various integration points. A dynamic sample assignment strategy was introduced to enhance the precision of our model and accelerate its convergence. To address the challenge of delineating bottom boundaries with clarity, our model employs a two-strategy approach: a threshold filter and a feedforward neural network (FFN) that provides targeted guidance for refining these boundaries. Our model demonstrated exceptional performance, achieving a mean average precision (mAP) of 47.1% on the water surface object dataset, which represents a 1.7% increase over the baseline YOLOv8 model. The dynamic sample assignment strategy contributes a 1.0% improvement on average precision at the intersection over union (IoU) threshold of 0.5 (AP0.5), while the FFN strategy fine-tunes the bottom boundaries and achieves an additional 0.8% improvement in average precision at IoU threshold of 0.75 (AP0.75). Furthermore, ablation studies have validated the versatility of our approach, confirming its potential for integration into various detection frameworks.

show abstract

“…The rapid growth and development of digital technology and the availability of video capture-based devices such as digital cameras and mobile phones with cameras are driving the explosive rise of network storage devices [1][2][3]. This is consistent with the nearly daily increase in the number of cars, but it is not accompanied by major volume changes [4]. Consequently, with a substantial increase in the number of vehicles and a constant volume of roadways, there will be an accumulation of vehicles, which ultimately leads to congestion [5,6].…”

Section: Introductionmentioning

confidence: 99%

Vehicle Detection Using Machine Learning Model with the Gaussian Mixture Model (GMM)

Rosnelly¹,

Riza²,

Wahyuni³

et al. 2022

JOWUA

View full text Add to dashboard Cite

Motion tracking apps are used for a lot of different things, like finding traffic jams and counting the number of cars going through a traffic light. The datasets come from many places on the internet, like YouTube and public dataset archives. There are about 20 videos that are tagged with the words "traffic" and "traffic camera video" and run for 10 to 30 seconds. The Gaussian Mixture Models (GMM) method is the proposed model. It separates the background from the tracked object, which is needed to do motion tracking. Then, the GMM method groups pixel data based on the background color of each pixel. After the cluster is made, the input is matched as a distribution, with the most common distribution used as the background. The analysis was done using MATLAB2019B. The results of this study show that the GMM method can adapt to the background. This is shown by the fact that testing of some of the given conditions went well.

show abstract

A object detection and tracking method for security in intelligence of unmanned surface vehicles

Cited by 19 publications

References 20 publications

A Millimeter-Wave Radar-Aided Vision Detection Method for Water Surface Small Object Detection

A Millimeter-Wave Radar-Aided Vision Detection Method for Water Surface Small Object Detection

Enhanced Water Surface Object Detection with Dynamic Task-Aligned Sample Assignment and Attention Mechanisms

Vehicle Detection Using Machine Learning Model with the Gaussian Mixture Model (GMM)

Contact Info

Product

Resources

About