Floating Raft Aquaculture Area Automatic Extraction Based on Fully Convolutional Network

Cui, Binge; Zhong, Yong; Dong, Fang; Zhang, Yonghui; Liu, Rongjie; Chu, Jialan; Zhao, Jianhua

doi:10.2112/si90-011.1

Cited by 21 publications

(13 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The learning rate of the model is set as 0.0001 and the batch-size is 8. Adam optimizer and the cross-entropy loss function are adopted in the model, and many scholars use them to optimize the network to identify aquaculture zones and achieve remarkable results [34,35]. Adam is an algorithm for first-order gradient-based optimization of stochastic objective functions.…”

Section: Extracting Aquaculture Areas With Fcnmentioning

confidence: 99%

A New Method for Extracting Laver Culture Carriers Based on Inaccurate Supervised Classification with FCN-CRF

Pan

Jiang

Zhang

et al. 2020

JMSE

View full text Add to dashboard Cite

Timely monitoring of marine aquaculture has considerable significance for marine ecological protection and maritime safety and security. Considering that supervised learning needs to rely on a large number of training samples and the characteristics of intensive and regular distribution of the laver aquaculture zone, in this paper, an inaccurate supervised classification model based on fully convolutional neural network and conditional random filed (FCN-CRF) is designed for the study of a laver aquaculture zone in Lianyungang, Jiangsu Province. The proposed model can extract the aquaculture zone and calculate the area and quantity of laver aquaculture net simultaneously. The FCN is used to extract the laver aquaculture zone by roughly making the training label. Then, the CRF is used to extract the isolated laver aquaculture net with high precision. The results show that the k a p p a coefficient of the proposed model is 0.984, the F 1 is 0.99, and the recognition effect is outstanding. For label production, the fault tolerance rate is high and does not affect the final classification accuracy, thereby saving more label production time. The findings provide a data basis for future aquaculture yield estimation and offshore resource planning as well as technical support for marine ecological supervision and marine traffic management.

show abstract

Section: Extracting Aquaculture Areas With Fcnmentioning

confidence: 99%

A New Method for Extracting Laver Culture Carriers Based on Inaccurate Supervised Classification with FCN-CRF

Pan

Jiang

Zhang

et al. 2020

JMSE

View full text Add to dashboard Cite

show abstract

“…The Food and Agriculture Organization of the United Nations (FAO) reports that the global marine aquaculture industry has been developing rapidly year by year [1]. As a critical component of mariculture, the rapid growth of marine floating raft aquaculture (FRA) has huge economic benefits but may harm the marine ecological environment, maritime traffic safety, and mariculture sustainable development [2][3][4][5]. Therefore, monitoring the distribution and quantity of FRA areas is of great significance for marine aquaculture planning and food security assessment.…”

Section: Introductionmentioning

confidence: 99%

Extraction of Floating Raft Aquaculture Areas from Sentinel-1 SAR Images by a Dense Residual U-Net Model with Pre-Trained Resnet34 as the Encoder

et al. 2022

View full text Add to dashboard Cite

Marine floating raft aquaculture (FRA) monitoring is significant for marine ecological environment and food security assessment. Synthetic aperture radar-based monitoring is considered to be an effective means of FRA identification because of its capability for all-weather applications. Considering the poor generalization and extraction accuracy of traditional monitoring methods, a semantic segmentation model called D-ResUnet is proposed to extract FRA areas from Sentinel-1 images. The proposed model has a U-Net-like structure but combines the pre-trained ResNet34 as the encoder and adds dense residual units into the decoder. For this model, the final layer and cropping operation of the original U-Net model are removed to eliminate the model parameters. The mean and standard deviation of Precision, Recall, Intersection over Union (IoU), and F1 score are calculated under a five-fold training strategy to evaluate the model accuracy. The test experiments indicated that the proposed model performs well with the F1 of 92.6% and IoU of 86.24% in FRA extraction tasks. In particular, the ablation experiments and application experiments proved the effectiveness of the improvement strategy and the portability of the proposed D-ResUnet model, respectively. Compared with the other three state-of-the-art semantic segmentation models, the experiments demonstrate a clear accuracy advantage of the D-ResUnet model. For the FRA extraction task, this paper presents a promising approach that has refined extraction capability, high accuracy, and acceptable model complexity.

show abstract

“…Visual interpretation is a classical method used to extract aquaculture areas from remote sensing images [7][8][9][10], but its operation is subjective, and the interpreter's prior knowledge has a significant influence on extraction accuracy, resulting in low efficiency. Using spectral features, some scholars have tried to automatically extract aquaculture areas according to spectral classification [11][12][13][14][15][16][17][18], or by constructing a spectral feature index [19][20][21][22][23]. Spectral classification, however, is susceptible to the phenomenon of "same object different spectrum" and "foreign object in the same spectrum", and the spectral characteristic index is constructed primarily for a certain region or a certain sensor, which has low portability and robustness.…”

Section: Introductionmentioning

confidence: 99%

Extraction of Aquaculture Pond Region in Coastal Waters of Southeast China Based on Spectral Features and Spatial Convolution

Wang

Zhang

et al. 2022

Water

View full text Add to dashboard Cite

To control the negative effects resulting from the disorderly development of aquaculture ponds and promote the development of the aquaculture industry, rapid and accurate identification and extraction techniques are essential. An aquaculture pond is a special net-like water body divided by complex roads and dikes. Simple spectral features or spatial texture features are not sufficient to accurately extract it, and the mixed feature rule set is more demanding on computer performance. Supported by the GEE platform, and using the Landsat satellite data set and corresponding DEM combined with field survey data, we constructed a decision-making model for the extraction of aquaculture ponds in the coastal waters, and applied this method to the coastal waters of Southeast China. This method combined the image spectral information, spatial features, and morphological operations. The results showed that the total accuracy of this method was 93%, and the Kappa coefficient was 0.86. The overlapping proportions of results between the automated extraction and visual interpretation for test areas were all more than 90%, and the average was 92.5%, which reflected the high precision and reliability of this extraction method. Furthermore, in 2020, the total area of coastal aquaculture ponds in the study area was 6348.51 km2, which was distributed primarily in the cities of Guangdong and Jiangsu. Kernel density analysis suggested that aquaculture ponds in Guangdong and Jiangsu had the highest degree of concentration, which means that they face higher regulatory pressure in the management of aquaculture ponds than other provinces. Therefore, this method can be used to extract aquaculture ponds in coastal waters of the world, and holds great significance to promote the orderly management and scientific development of fishery aquaculture.

show abstract

Floating Raft Aquaculture Area Automatic Extraction Based on Fully Convolutional Network

Cited by 21 publications

References 0 publications

A New Method for Extracting Laver Culture Carriers Based on Inaccurate Supervised Classification with FCN-CRF

A New Method for Extracting Laver Culture Carriers Based on Inaccurate Supervised Classification with FCN-CRF

Extraction of Floating Raft Aquaculture Areas from Sentinel-1 SAR Images by a Dense Residual U-Net Model with Pre-Trained Resnet34 as the Encoder

Extraction of Aquaculture Pond Region in Coastal Waters of Southeast China Based on Spectral Features and Spatial Convolution

Contact Info

Product

Resources

About