Research On The Intelligent Feeding System Of Ornamental Fish

The ornamental fish industry,as a fresh blood of leisure fishery in China,plays an important role in leisure fisheries with the continuous development of economy.But the breeding conditions of ornamental fish are very demanding literally.The professional knowledge level of the feeders falls far behind,lacking the experience of scientific feeding.Meanwhile,the breeders are too busy to take care of the ornamental fish in time,which will easily lead to the death of the ornamental fish and cause economic losses.At present,most research on the intelligent feeding system of ornamental fish stay on monitoring the living environment of ornamental fish by using Internet of Things technology,which requires users to judge whether to feed or not,and can not automatically and accurately feed ornamental fish.This article is based on computer vision and the IOT technology,with the Raspberry Pi as the core,combined with a variety of sensor devices and target detection methods,not only to monitor the living environment of ornamental fish,feeding behavior,activity behavior,but also intelligent control of feeding,oxygen supply and other behaviors to achieve the purpose of intelligent feeding ornamental fish.The following are the main research content of this paper.1.In IOT,the temperature sensor,TDS sensor and other sensor modules are used to measure the water temperature,water purity,PH value,water level and other indicators in the aquarium in real time,and raspberries are used to shoot them at the same time,which also achieves the goal of remotely monitoring the fish dynamics in the aquarium.2.In the aspect of target detection model,the system can use PaddleX and other tools provided by PaddlePaddle to train the migration learning model of the constructed data set,and obtain an ornamental fish target detection model based on the YOLOv3 model and the PPYOLO model,and construct an ornamental fish detection system combining local detection with cloud service detection.3.With regard to the intelligent feeding system,aiming at the feeding process of ornamental fish,such as feeding,oxygen supply,vital signs judgment,water purity,water temperature,etc.,the corresponding automatic control management algorithms and processes are designed,successfully achieve the intelligent automatic feeding control and the vital signs judging.4.In terms of control system,the local control system developed by PyQt based on the Raspberry Pi hardware provides users with convenient and fast equipment control and clear fish tank indicator display.The remote control system is developed based on Spring Boot to help users remotely control the fish tank.This system uses the IOT technology and computer vision technology to provide data support for the intelligent feeding system,reaching the goal of the intelligent and automatic feeding of ornamental fish,and it is proved to have powerful functions.In the actual operation of the experiment,the system runs stably and can satisfy the automatic feeding without the user.It reduces the difficulty of feeding ornamental fish,solves the unscientific feeding method of the feeder,and also provides great help to the feeder from the perspective of feeding.