Shuangxing Feng's research while affiliated with Beijing Academy of Agriculture and Forestry Sciences and other places

Quantifying feeding intensity of fish is important in developing intelligent feeding control system, thus improving feed utilization rate and reducing water pollution. The current study explored a real-time, high-precision and lightweight 3D ResNet-Glore fish feeding intensity quantification network, which can accurately locate the four levels of fish feeding intensities in video stream. In this network, the lightweight GloRe module is expanded in 3D space, and the Residual block in the 3D ResNet network is modified to form the 3D GloRe module. The relational reasoning is achieved through graph convolution in the interactive space to improve accuracy of discrimination. In addition, the sliding window and the frame extraction processing of the video data significantly reduces the model parameters and the amount of calculation. Experimental results showed that the classification accuracy for four types of feeding intensity was 92.68%, which is 4.88% higher compared with that of the classical 3D ResNet network. The parameters were decreased by 46.08% and the GFLOPs decreased by 44.10%. The proposed network improved the training and recognition speed and reduced the hardware equipment requirements, which can provide a theoretical basis for subsequent feeding decisions.

In aquaculture, accurate and automatic quantification of fish behaviour can provide useful data input for production management and decision‐making. In recent years, with the focus on fish welfare, it has become urgent to study and use nondestructive quantitative methods of fish behaviour in aquaculture. In this paper, based on the literature of the past 30 years, nonintrusive and automatic quantitative methods for fish behaviour are analysed. Firstly, several important fish behaviours in aquaculture are listed, and the quantification of fish behaviour is summarized in four stages: detection, tracking, feature extraction and behaviour recognition. Then, nonintrusive methods of fish behaviour quantification, through machine vision, acoustics and sensors, and their advantages and disadvantages are also compared and discussed in detail. It is concluded that the combination of multiple methods and deep learning is a key technology for fish behaviour quantification, which has gradually become a popular focus of research and application in recent years. This review can be used as a reference to improve fish behaviour quantification in future, so as to create a more effective and economic technical method.