Applying convolutional networks to underwater tracking without training

Huang, Renjie; Lai, Yi-Chung; Tsao, Chun-Yu; Kuo, Yen-Chun; Wang, Jung-Hua; Chang, Chung‐Cheng

doi:10.1109/icasi.2018.8394604

Cited by 10 publications

(4 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In individual fish behavior analysis, numerous studies employ deep learning method to track the individual fish in controllable laboratory conditions (Fig. 1 (a-f)), and then behavior characteristic parameters including swimming speed and direction are extracted from tracking trajectories (Chuang et al, 2017;Huang et al, 2018;Jonas et al, 2017;G. Wang et al, 2017).…”

Section: Introductionmentioning

confidence: 99%

A dual attention network based on efficientNet-B2 for short-term fish school feeding behavior analysis in aquaculture

Yang

Cheng

et al. 2021

Computers and Electronics in Agriculture

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

A dual attention network based on efficientNet-B2 for short-term fish school feeding behavior analysis in aquaculture

Yang

Cheng

et al. 2021

Computers and Electronics in Agriculture

View full text Add to dashboard Cite

“…Shkurti et al (2017) developed several lightweight neural network architectures for visual convoying, and devised a ReducedYOLO model for object tracking. Huang et al (2018) proposed an underwater surveillance system where the GMM-based background subtraction scheme is used for object detection. Further, multi-target tracking is implemented with Fast-CNT2.…”

Section: Deep Learning Based Object Trackingmentioning

confidence: 99%

Passive Visual Underwater Surveillance: A Survey

Subudhi

Kapoor

Rout

et al. 2023

Preprint

View full text Add to dashboard Cite

The objective of this article is to provide a detailed review of the state-of-the-art underwater surveillance process and the new trends of the same. Underwater surveillance has recently getting lots of attention because of its potential applications including the security of the coastal border, effective fish farming, deep-sea exploration, preservation of rare aquatic animals, etc. In the case of underwater, the light that is sensed by the camera got degraded due to many underwater phenomena such as haze, low illumination, scattering, absorption, diffraction, and refraction. The actual color of the object, as well as the scene, gets degraded as the light can not travel deep underwater. In a situation, where the scene of view is not clear, it is very difficult to detect the moving objects present in the scene. Further, the identification of the object of concern and tracking of that becomes more challenging. In this survey, we categorize underwater surveillance as a combination of the three blocks: enhancement, object detection, and object tracking. We categorize all these three blocks based on the underwater complexity and motion models considered. In this article, we tried to enumerate the most detailed descriptions of each category. We also discuss the future directions of research in the area of underwater surveillance.

show abstract

“…Using image technologies to analyze fish behavior enables us to provide important information to farmers for their use in making feeding decisions. In general, image techniques based on tracking are used to extract fish trajectories and then utilize extracted characteristic from trajectories to quantify individual fish feeding behavior [4][5][6][7]. However, these methods are not suitable for intensive aquaculture since fishes are generally managed at the group level, and motion of fish school is complex [8].…”

Section: Introductionmentioning

confidence: 99%

An FSFS-Net Method for Occluded and Aggregated Fish Segmentation from Fish School Feeding Images

et al. 2023

View full text Add to dashboard Cite

Smart feeding is essential for maximizing resource utilization, enhancing fish growth and welfare, and reducing environmental impact in intensive aquaculture. The image segmentation technique facilitates fish feeding behavior analysis to achieve quantitative decision making in smart feeding. Existing studies have largely focused on single-category object segmentation, ignoring issues like occlusion, overlap, and aggregation amongst individual fish in the fish feeding process. To address the above challenges, this paper presents research on fish school feeding behavior quantification and analysis using a semantic segmentation algorithm. We propose the use of the fish school feeding segmentation method (FSFS-Net), together with the shuffle polarized self-attention (SPSA) and lightweight multi-scale module (LMSM), to achieve two-class pixel-wise classification in fish feeding images. Specifically, the SPSA method proposed is designed to extract long-range dependencies between features in an image. Moreover, the use of LMSM techniques is proposed in order to learn contextual semantic information by expanding the receptive field to extract multi-scale features. The extensive experimental results demonstrate that the proposed method outperforms several state-of-the-art semantic segmentation methods such as U-Net, SegNet, FCN, DeepLab v3 plus, GCN, HRNet-w48, DDRNet, LinkNet, BiSeNet v2, DANet, and CCNet, achieving competitive performance and computational efficiency without data augmentation. It has a 79.62% mIoU score on annotated fish feeding datasets. Finally, a feeding video with 3 min clip is tested, and two index parameters are extracted to analyze the feeding intensity of the fish. Therefore, our proposed method and dataset provide promising opportunities for the urther analysis of fish school feeding behavior.

show abstract

Applying convolutional networks to underwater tracking without training

Cited by 10 publications

References 6 publications

A dual attention network based on efficientNet-B2 for short-term fish school feeding behavior analysis in aquaculture

A dual attention network based on efficientNet-B2 for short-term fish school feeding behavior analysis in aquaculture

Passive Visual Underwater Surveillance: A Survey

An FSFS-Net Method for Occluded and Aggregated Fish Segmentation from Fish School Feeding Images

Contact Info

Product

Resources

About