Intelligent Recognition Of Coding On The Hull Structure In Natural Scene Based On Deep Learning

"Made in China 2025" emphasizes accelerating the deep integration of the new generation of information technology and manufacturing,and speeding up the intelligent transformation of production equipment in the shipbuilding and machinery industries.Deep learning is to learn the internal rules and presentation level of sample data.Its goal is to enable machines to have analytical learning capabilities like humans,and to recognize data such as text,images and sounds.In order to adapt to the modern shipbuilding mode,the ship’s production mode and management mode must be modernized and efficient.During the construction of the ship,the hull components were manually coded.The hull parts are coded using a combination of numbers and letters to completely and accurately express the position,type,serial number,processing method and other information of the hull parts and components.Therefore,this paper explores the intelligent identification of coding on the hull parts in the process of ship production.In this paper,the convolutional neural network is used to identify the coding on the hull component combined with the image processing algorithm.Firstly,the image with the hull construction coding in the natural scene is preprocessed,including noise reduction and edge enhancement.Then the text detection algorithm is used for coding positioning,at the same time,the advantages and disadvantages of various detection algorithms are compared.After locating the encoded,projection method is used to segment a single character and a single character data set for training is made.Finally,the data set is used for convolution neural network supervised training,neural network model reference LeNet-5 convolutional neural network structure.At the same time,in order to avoid the learning shortage caused by too little data,the migration learning method is used to pre train the convolutional neural network with EMNIST data set in advance,The experimental results show that the method in this paper can accurately identify the coding on the hull component,recognition accuracy reaches 88.57%.Subsequently the identified coding is used to retrieve the model of the hull segment and component from the database,and combine AR technology to realize model display and process information display,and further improve the intelligent level of shipbuilding and construction process.This paper realizes the application of deep learning technology to the identification of coding on the hull parts,which is an exploration of the intelligence of the ship production and construction process,and get through the data flow of ship design and ship construction site,which is conducive to the modernization and efficiency of the ship production process,and will further improve the intelligence level of the ship production and construction process.