Water Turbidity Detection Based On LMBP Neural Network

With the development of social and economic development,people’s demand for quality of life is also getting higher and higher.Among them,water,as the source of life,has an extremely important position in people’s daily drinking and environmental protection.In the water quality evaluation index,turbidity is also a major indicator,because turbidity is caused by particles floating in the water body,which mainly includes silt,colloidal organic matter and inorganic chemicals,cells,germs,etc.,and these impurities may cause the production and spread of diseases,so the detection requirements for turbidity are becoming more and more stringent.Conventional turbidity measurement technology according to the basic principle of turbidity,mainly divided into horizontal transmission method,vertical scattering method,ratio analysis method,etc.,these methods have complex operation,high cost,low accuracy and other defects,this paper combined with neural network analysis and image processing technology,in the measurement of water turbidity,through the neural network of the powerful processing and analysis ability of image data,a water turbidity detection method based on the combination of digital image and LM-BP neural network is proposed.The main contents of this article are as follows:(1)For different turbidity measurement methods,analyze the characteristics of various methods.At the same time,according to the definition of turbidity and the basic principle of measurement,the water quality turbidity measurement technology using image recognition technology is given,which requires relatively little measurement conditions and the cost of the instrument is relatively cheap.(2)Based on different image information extraction algorithms of images,with reference light source intensity as the main information center,an image texture information extraction algorithm based on gray gradient symbiosis matrix is proposed,and after the algorithm extracts the image features,the turbidity liquid features of different states of different turbidity have obvious differences.(3)The neural network required for the analysis of turbidity images is studied,the main structure is the reverse transmission mode,and according to the design requirements,experimental detection is carried out to determine the neural network structure parameters such as the input layer,output layer,the number of hidden layer nodes and the learning rate,and the model structure is designed.Complete the specific scheme design of software and hardware related to the system.Experiments have proved that compared with the conventional turbidity recognition technology,the turbidity detection technology based on LMBP neural network proposed in this paper has achieved good measurement effect and has certain social application value.Figure 40 Table 6 Reference 92...