Research On Diagnosis Of Mariculture Diseases Based On Big Data Technology

The scale of my country's marine aquaculture industry is large,and the problem of the spread of aquatic biological diseases has occurred from time to time,causing serious economic losses to farmers.In the past,the disease mainly relied on manual on-site diagnosis.Farmers can easily delay the disease based on personal experience and blind medication.The lack of professional and technical personnel also prevents the timely diagnosis and treatment of farming biological diseases.Numerous authoritative institutions,scientific research institutes,and fishing units each collect a large amount of mariculture disease data.Many of the valuable information contained in them can be fully excavated and utilized with the help of big data,deep learning,and other technologies.Provide scientific diagnosis and treatment methods to reduce breeding risks and promote the construction and development of marine aquaculture.This paper proposes a big data platform for marine aquaculture disease diagnosis based on Hadoop technology,and based on HDFS,Map Reduce,HBase and other core components,it sorts out the disease data processing process under the framework of the big data platform,extensively collects disease data and digs in massive amounts of disease data The information contained inside.In view of the difficulty of designing accurate and complete image classification features with traditional image recognition technology,this paper develops a set of disease diagnosis models based on Goog Le Net convolutional neural network,using convolution integral solution,normalization method,h-Swish activation function The Inception structure is improved,which saves network parameters and improves recognition accuracy while increasing training speed.In order to optimize the model fitting process,this paper uses batch normalization to prevent gradient explosion,uses Dropout and data enhancement to prevent overfitting,and uses transfer learning for pre-training to reduce the extraction time of low-level features,and finally achieves rapid and accurate diagnosis of marine aquaculture diseases.According to the business requirements of disease diagnosis,this article makes a general plan for the functional modules of the marine aquaculture disease diagnosis big data platform,and builds on the platform's specific functions based on advanced software development technology to realize the application of big data technology in marine aquaculture disease diagnosis.The mariculture disease diagnosis big data platform can make full use of massive disease data resources,analyze and effectively use the potential value of the disease data,provide scientific and thoughtful disease diagnosis services for the mariculture industry,and promote the good and fast development of the mariculture industry.