Research on water quality monitoring of rivers and lakes based on unbalanced image
Water quality monitoring is of great significance to the ecological construction of rivers and lakes,but the traditional water quality monitoring methods of rivers and lakes have the problems of difficult monito-ring methods and high monitoring costs.In order to make water quality monitoring more intelligent and con-venient,this paper improves the VGG16 convolutional neural network to analyze river and lake images for water quality monitoring by cost-sensitive cross-entropy function method based on river and lake images with unbalanced characteristics,and compares it with unbalanced data processing methods such as random un-dersampling and image enhancement.After a large number of experiments,the results show that the accura-cy,precision,recall and F1 value of the method combining VGG16 convolutional neural network with cost cross entropy function are higher than other methods,which can reach 0.91,0.92,0.91 and 0.92,re-spectively.It is proved that this method can effectively classify the water quality of river and lake unbal-anced images.
water quality monitoringunbalanced data setcost sensitiveconvolutional neural networkVGG16
万方数据
维普
