Chuting Yu
ABSTRACT
This research has developed a neural network model based on deep learning, which consists of a recognition part and a counting part. The recognition part is realized by the Fully Convolutional Neural Network (FCN), and thecounting part is realized by the NAS-CNN model. Labelme (online labeling tool) was used to manually mark the two types of targets, the remnant bait and the white prawn, and a special data set for the identification of the white prawn was created. Experiments show that the FCN model has achieved excellent recognition results, with a training recognition accuracy rate of 99%, and a verification recognition accuracy rate of 96.13%. When counting the residual bait, the connected-component labeling (CCL) isused for the image with only sporadic residual bait. The accuracy rate reached 97.5%. The pictures with severe adhesion residual bait are counted using the NAS-CNN model training, and the verification accuracy rate can reach 88.52%.Experiments have proved that combined with the deep learning method, the white shrimp and the residual bait can be quickly and accurately identified and theresidual bait can be counted at the same time. This method provides a reference for exploring the scientific breeding methods of Penaeus vannamei, and can be extended to use in other aquaculture environments.
Chuting Yu, Yu Li and Xinjie Yu.2021. A Method for Recognizing and Counting Residual Bait of Penaeus Vannamei Based on Deep Learning. In Proceedings of The 10th international workshop on intelligent data processing (IDP2021). ACM,Melbourne, Vic,Australia, 4 pages.
- FAO, 2018. Fishery and Aquaculture Statistics 2016. Food and Agriculture Organization of the United Nations, Rome, Italy.Google Scholar
- Gatlin, D.M., Barrows, F.T., Brown, P., Dabrowski, K., Gaylord, T.G., Hardy, R.W., Herman, E., Hu, G., Krogdahl, Å., Nelson, R., Overturf, K., Rust, M., Sealey, W., Skonberg, D., J Souza, E., Stone, D., Wilson, R., Wurtele, E., 2007. Expanding the utilization of sustainable plant products in aquafeeds: a review. Aquac. Res. 38, 551–579.Google ScholarCross Ref
- Md.moazzem Hossain.,Md.LokmanAlf.,Saleha Khan.,Md.Mahfuzul Haque.,Md.Shahjahan.,2020.Use of Asian watergrass as feed of grass carp .Aquaculture Reports.Volume 18,100434Google Scholar
- Zhou, C., Zhang, B., Lin, K., Xu, D., Chen, C., Yang, X., Sun, C., 2017. Near-infrared imaging to quantify the feeding behavior of fish in aquaculture. Comput. Electron. Agric. 135, 233–241. https://doi.org/10.1016/j.compag.2017.02.013.Google ScholarDigital Library
- François Chollet.,Deep Learning with Python.,2018,pp110-111Google Scholar
- Krizhevsky A, Sutskever I, Hinton GE. ImageNet classifification with deep convolutional neural networks. Adv Neural Inf Process Syst 2012;25(2):11, 6–14.Google Scholar
Recommendations
Face Recognition Based on Deep Learning
Human Centered ComputingAbstractAs one of the non-contact biometrics, face representation had been widely used in many circumstances. However conventional methods could no longer satisfy the demand at present, due to its low recognition accuracy and restrictions of many ...
Research Progress of Facial Expression Recognition Based on Deep Learning
CAIH2020: Proceedings of the 2020 Conference on Artificial Intelligence and HealthcareFacial expression recognition is a practical application in the field of artificial intelligence. This article focuses on the introduction of the principles of each facial expression recognition method based on deep learning in chronological order, as ...
A novel trilinear deep residual network with self-adaptive Dropout method for short-term load forecasting
Highlights- A novel deep residual network with self-adaptive Dropout method is proposed.
- A ...
AbstractShort-term load forecasting underlies the effective energy management. However, accurate short-term load forecasting is a big challenge owing to the significant uncertainty and volatility of load demand. Deep learning has successfully ...
Comments