Research Of Acquisition Device About Shipyard Energy Consumption Data Based On Image Recognition Technology

Shipbuilding industry is a large energy consumer,and the cost of energy consumption is relatively high.Ship manufacturing process is scattered,energy consumption management is rough,manual meter reading and statistical methods are still used,which can not achieve effective supervision,and energy consumption waste is serious.Against the background of increasing energy shortage,the shipbuilding industry urgently needs to find a breakthrough in energy saving and emission reduction,reduce manufacturing costs and improve the competitiveness of shipbuilding industry.In this paper,image recognition technology is applied to automatically identify energy consumption meters,read energy consumption data and transmit them to energy consumption data center for centralized management.Through the analysis and statistics of the collected energy consumption data,the effective supervision of energy consumption in shipbuilding industry can be realized,so as to reduce the energy consumption cost of shipbuilding industry,which has far-reaching significance for promoting the sustainable development of shipbuilding industry.The main research contents of this paper include the following aspects: collecting image information by camera,transmitting image information by wired or wireless way.After the image enters the computer,according to the characteristics of the digital mechanical table,the method of color model extraction and gray scale transformation is used to preprocess the digital image;applying Canny edge.The edge detection algorithm obliquely corrects the pre-processed image;then uses binarization operation to extract the region with higher brightness value in the image;secondly,obliquely corrects the image after binarization;thirdly,accurately locates the digital area by horizontal and vertical projection of the corrected image;and then,the digital area is located through the digital area.After the precise location of the image,horizontal and vertical projections are implemented in order to achieve digital segmentation.When extracting digital features,structural features are extracted by threading method and statistical features are extracted by coarse mesh method.In the process of digital recognition,the pattern recognition method which combines the digital threading method and the pixel statistics method is adopted.The system of image recognition software is completed,and the recognition system is tested.There are data Misreading in the experiment,but it does not affect the recognition rate of the number.After improving the reason,the ideal state can be achieved.Experiments show that it is feasible to use energy consumption data acquisition device based on image recognition technology to collect energy-saving monitoring data.The system can collect energy consumption data of each branch and high-energy-consuming equipment at fixed time and point.Moreover,because of the growth law of existing meters,it can make mistakes in this way.The recognition rate of the erroneous data is very high.It can basically read the instrumentation digits by camera,reduce the interference of noise to the digits to the greatest extent,improve the accuracy of the digit recognition,and adapt to the monitoring environment of shipyards and industrial enterprises.