基于图像识别的碎屑流颗粒分布特征及碎屑流与房屋相互作用探究

由于崩滑-碎屑流具有高隐蔽性、发生时间短暂以及较低的可预见性，较难直观观测崩滑-碎屑流发生的过程。为了对碎屑动力特性及对房屋的破坏情况进行研究，拟对碎屑流堆积体及其粒径分布进行分析。以普洒村崩塌-碎屑流为例，使用PCAS系统对碎屑流堆积体图像进行颗粒识别、统计，并通过量纲分析法，分析堆积体颗粒与房屋破坏之间的定量联系。分析结果发现:与现场筛分统计相比，图像识别堆积体颗粒的方法得到的数据更全面、详细，并且能节省大量的人力、物力资源。另外，对房屋与碎屑流之间的相互作用进行了探究，发现房屋对碎屑流颗粒有"拦粗排细"的作用。同时，利用图像识别得到的颗粒粒径数据对推导出房屋破坏的判别公式，判别效果较好，能够在地质灾害防治、预测领域发挥一定作用。 

Particle distribution characteristics of rock avalanche and the interaction between rock avalanche and houses based on image recognition

Due to the high concealment, short-lived time and low predictability of the rock avalanche, it is difficult to visually observe the process of it. In order to study the dynamic characteristics and the catastrophicto houses of rock avalanche, this paper analyzes the rock avalanche congeries and its particle size distribution.Taking the rock avalanche in Pusa Village as an example, this paper uses the PCAS system to identify congeries particles by image.And through the dimensional analysis method, the quantitative relationship between the congeries of particles and the damage of the building is analyzed.The author found that: compared with on-site screening statistics, the data obtained by the method of image recognition of congeries particles is more comprehensive and detailed, and can save a lot of manpower and material resources.In addition, the interaction between the house and the rock avalanche is explored, and it is found that the house has the effect of "blocking coarse and fine discharge" on the rock avalanche particles. At the same time, using the particle size data obtained by image recognition to derive the discrimination formula for building damage, the discriminating effect is good, laying the foundation for follow-up research.