基于颗粒识别分析系统的碎屑流堆积物颗粒识别和统计方法研究

Particle identification and statistical methods of a rock avalanche accumulation body based on the particle analysis system

碎屑流堆积物颗粒识别和统计是碎屑流灾害的研究重点。文章基于图像处理孔隙(颗粒)及裂隙图像识别与分析系统(PCAS),以贵州纳雍普洒村崩塌-碎屑流为例,结合纳雍崩塌堆积物粒径实测结果,通过阐释识别过程中阈值、孔喉封闭半径、最小孔隙面积的参数意义,研究PCAS软件在碎屑流颗粒识别与统计中的应用,并提出了颗粒识别时这些参数的选取方法。分析结果表明:(1)PCAS能自动准确地识别碎屑流堆积物颗粒与孔隙,相比人工计数更精细,所识别堆积物各区小颗粒比重较大,0~2 m颗粒粒径各区占比都在50%以上;(2)当阈值为170(像素)时能获得精细的二值图像,颗粒与孔隙得到了准确地区分;(3)不同参数取值下获得堆积物颗粒粒径分布结果不同,碎屑流堆积物颗粒识别宜采用较大的孔喉封闭半径和较小孔隙面积,当二者比值为3/30(像素)时能更好地反应颗粒粒径分布情况;(4)PCAS具有较高的可行性,统计结果显示,各粒径含量变化趋势与人工统计相近,两种统计方法各粒径占比、分布规律基本吻合,说明利用PCAS可以实现对崩塌碎屑流颗粒粒径分布的高效便捷分析。

Particle identification and statistics of rock avalanche deposits are the focus of researches on rock avalanche disasters.Based on the image processing PCAS particle recognition system,taking the collapse-rock avalanche of the Pusa Village in Nayong of Guizhou Province as an example,combined with the measured results of the particle size of the Nayong collapse,the parameter values of threshold(T),pore throat closing radius(r)and minimum pore area(S0)during the identification process are explained.The application of PCAS software to the identification and statistics of rock avalanche particles is studied,and the selection methods of these parameters are proposed.The analysis results show that(1)PCAS can automatically and accurately identify debris flow accumulation particles and pores,which are more precise than manual counting.Small particles in each area of the recognized accumulation have a large proportion,and the proportion of 0~2 m particles in each area is more than 50%.(2)When the threshold value is 170(pixels),a fine binary image can be obtained,and the particles and the pores are accurately distinguished.(3)The particle size distribution results of the deposits obtained under different parameter values are different,and the particle identification of the rock avalanche deposits should adopt larger r and S0.When r/S0=3/30(pixels),the particle size distribution can be better reflected.(4)PCAS is of the high feasibility and the statistical results show that the variation trend of each particle size content is similar to that of artificial statistics.The proportion and distribution of particle sizes in the two statistical methods are basically consistent with each other,which is of great significance for the efficient and convenient analysis of particle size distribution of collapse-rock avalanche.