Quantitative Calculation Method Of Sediment Grain Size Based On Digital Image Processing

Quantitative analysis of sediment grain size has always been one of the emphases of geology and river hydraulics.River water carries sediments of different grain sizes to provide a lot of information.It reflects the hydrodynamic conditions and indirectly indicates a series of river changes,such as river velocity,discharge,upstream sediment loss,etc.At the same time,it provides necessary basic data for environmental monitoring and mineral resources exploration.The traditional grain size analysis method is to collect samples in the field and send them to the laboratory for instrument analysis.In the process of systematically studying the grain size variation of sediments in river direction,a lot of work is necessary.A large number of samples were collected in the field.Large-scale experimental measurements and analysis were carried out in the laboratory.This method not only consumes a lot of material and manpower,but also requires a lot of experimenter's energy.Therefore,how to quickly and conveniently quantitatively analyze the grain size of sediments has been studied by geologists,geographers and hydraulics.In view of the shortcomings of the previous detection methods,this paper adopts the detection method based on digital image processing.Image segmentation is an important branch of image processing,and its application is very extensive.Superpixel segmentation is ideal in generating the compactness of superpixel and contour preservation.With the research and development of superpixel segmentation method,it is widely used in various fields.In this paper,the application of super-pixel segmentation method in gravel particle size calculation is studied,with emphasis on the in-depth analysis of SLIC superpixel segmentation method.The main contents and innovations of this paper are as follows:Firstly,digital camera is always portable to take along in various field environment.It can photo rocks target whenever and wherever possible.Therefore,the detection method of digital image processing can reduce the problems of field investigation and material collection and transportation,only need to use digital camera for sample shooting.The texture of gravel image has a great influence on the segmentation effect of some segmentation methods in this paper.To ensure a better segmentation effect,the surface fuzzy algorithm is used to process the initial collected image properly in this paper.Secondly,two image segmentation methods are implemented in this paper.Both of them belong to super-pixel segmentation methods.One is based on SLIC super-pixel image segmentation and the other is based on normalized cut image segmentation.The former performs better in segmentation efficiency,and the latter has better robustness.According to the actual engineering needs,a grain size measurement algorithm based on SLIC superpixel segmentation and merging method is selected,and its segmentation results are used for subsequent experiments and analysis.At the same time,three mainstream image segmentation algorithms are used for gravel image segmentation,and different algorithms are compared and analyzed.the performance of this algorithm is excellent.The algorithm of this paper performs well.Finally,the quantitative calculation method of sediment grain size based on segmentation results is studied.According to the actual needs,the compilation of manual sediment measurement software and automatic sediment measurement algorithm has been completed.Through the analysis of image segmentation results,the quantitative calculation of grain size is carried out from the measurement strategies and methods of particle size.The results are compared with those obtained by manual measurement and other methods.At the same time,the whole process of image preprocessing to measurement and statistics is programmed,which makes the quantitative analysis of gravel particle size more convenient and improves work efficiency.The results show that the detection method adopted in this paper has less error under the condition of guaranteeing faster speed.This method reduces human error and improves the automation of detection.The program that integrates the detection method has the advantages of simple operation,accuracy and efficiency.