Research On Real-time Surveillance Technology Of Red Seabream Iridovirus Of Oplegnathus Punctatus Based On Video Calculation

Oplegnathus puntatus is a new fish species in internal mariculture industry in recent years.It has high ornamental value,economic value and edible value.oplegnathus puntatus culture has great development prospects.However,oplegnathus punctatus has extremely strict requirements on the production and breeding environment.Diseases produced in breeding,especially the red seabream iridovirus disease and other viral diseases,spread rapidly and have high mortality rate,which seriously restrict its development.In this paper,the research on the real-time monitoring technology of oplegnathus punctatus disease is realized by video calculation method.On the basis of reducing the cost of manpower monitoring,the disease situation of oplegnathus punctatus can be identified and preliminarily diagnosed by monitoring the body surface characteristics and behavior characteristics of oplegnathus punctatus in real time.The research content of this article is mainly divided into the following four parts:Oplegnathus punctatus datasets were created.A series of data sets of oplegnathus punctatus video and living environment parameters from health to suffering from red seabream iridovirus disease to death were created.The data sets were constructed under the guidance of professionals to ensure the comprehensiveness,accuracy and reliability of experimental data.The data set used in this paper includes oplegnathus puntatus individual identification dataset,oplegnathus puntatus abnormal characterization dataset,oplegnathus puntatus abnormal swimming behavior dataset.The method proposed in this paper is verified on the three original datasetsAn individual identification network model based on convolution neural network is proposed.Firstly,the similarity algorithm is used to preprocess the video images such as similarity deletion on the oplegnathus punctatus individual recognition data set collected from the experiment.Secondly,through the initial target location and segmentation,II?Net backbone convolutional neural network and three modules of the improved full connection layer based on genetic algorithm,a convolutional neural network model framework for individual identification of oplegnathus punctatus is constructed.Finally,through the experimental comparison and analysis with several neural networks,the validity and practicability of this convoluted neural network for oplegnathus punctatus individual identification are verified.This paper presents a method to identify the abnormal characterization of oplegnathus punctatus with red seabream iridovirus disease.According to the actual situation of the body surface characteristics of oplegnathus puntatus with red sSeabream iridovirus disease,the abnormal characterization indexes of the diseased oplegnathus puntatus are designated as the shape characteristics such as the opening and closing size of the fish mouth,the fish body color changes and other fish body characteristics.Firstly,on the basis of individual identification,the disease of a single fish is tracked and monitored to obtain body surface characteristic data information from health to red seabream iridovirus disease to death of oplegnathus punctatus.Secondly,through the oplegnathus punctatus body surface feature extraction module and the oplegnathus punctatus suffering from red seabream iridovirus disease identification module based on VGG and GoogleNet network model improvement,the oplegnathus punctatus body surface feature identification network model for red seabream iridovirus disease is constructed.Finally,the neural network model is used to train,verify and test the data set,so as to achieve the purpose of identifying different periods of disease of oplegnathus puntatus by monitoring the body surface features of oplegnathus puntatus.An abnormal swimming behavior identification method for patients with red seabream iridovirus disease oplegnathus punctatus is proposed.According to the actual swimming behavior of oplegnathus puntatus suffering from red seabream iridovirus disease,the abnormal swimming behavior index of oplegnathus puntatus is defined as rollover.First,select the appropriate fish school video image for data preprocessing,such as screening video frames and marking them to obtain the data set to be identified.Secondly,through the feature network module and the area generation network module,the abnormal swimming behavior identification model of the red seabream iridovirus disease oplegnathus punctatus is constructed.Finally,the abnormal swimming behaviors in oplegnathus punctatus swimming behavior data set are detected and identified to achieve the purpose of monitoring and identifying abnormal swimming behaviors.