School of Information and Communication Engineering, School of Computer Science and Cyberspace Security, State Key Laboratory of Marine Resource Utilization in South China Sea, Hainan University, Haikou, China.
Environ Monit Assess. 2020 Jul 8;192(8):493. doi: 10.1007/s10661-020-08409-9.
Aquaculture is an important part of agricultural economy. In the past, major farming accidents often occurred due to subjective experience. There are many factors affecting the water quality of aquaculture. Maintaining an ecological environment with good water quality is the most critical link to ensure the production efficiency and quality of aquaculture. With the continuous development of science and technology, intelligence and informatization in aquaculture has become a new trend. Smart aquaculture cannot only realize real-time monitoring, prediction, warning, and risk control of the physical and chemical factors of the aquaculture environment but can also conduct real-time monitoring of the characteristics and behaviors of the fish, which infers the changes of the aquaculture ecological environment. In this paper, the research achievements over past two decades both are summarized from four aspects: water quality factor acquisition and pre-processing, water quality factor prediction, morphological characteristics, and behavioral characteristic recognition of fish and the mechanism between fish behavior and water quality factors. The advantages and disadvantages of existing research routes, algorithm models, and research methods in smart aquaculture are summarized. The work in this paper can provide a well-organized and summative knowledge reference for further study on the dynamic mechanism between the changes of water quality factors and the fish body characteristics and behavior. Meanwhile, the work can also provide valuable reference for promoting the smart, ecological, and efficient development of aquaculture.
水产养殖是农业经济的重要组成部分。过去，由于主观经验，经常发生重大养殖事故。影响水产养殖水质的因素很多。保持良好水质的生态环境是确保水产养殖生产效率和质量的最关键环节。随着科学技术的不断发展，水产养殖的智能化和信息化已成为新趋势。智能水产养殖不仅可以实现对水产养殖环境理化因素的实时监测、预测、预警和风险控制，还可以对鱼类的特征和行为进行实时监测，推断水产养殖生态环境的变化。本文从水质因子获取与预处理、水质因子预测、鱼类形态特征和行为特征识别及其与水质因子的相互作用机制四个方面总结了过去二十年来的研究成果。总结了智能水产养殖中现有研究路线、算法模型和研究方法的优缺点。本文的工作可以为进一步研究水质因子变化与鱼类特征和行为之间的动态机制提供系统的、总结性的知识参考。同时，也为促进水产养殖的智能化、生态化和高效化发展提供了有价值的参考。
