Aggregate shape classification method based on machine learning

Abstract

The invention discloses an aggregate shape classification method based on machine learning. The aggregate shape classification method comprises the steps that 1, dividing an aggregate sample into sixaggregate shapes; collecting an aggregate image of each aggregate; 2, pretreating; 3, obtaining the value of the aggregate morphological characteristic parameter corresponding to each aggregate; 4, taking the types of the aggregates in the aggregate samples and the morphological characteristic parameters of the aggregates as a training set, and performing training by utilizing an XGBoost algorithm; 5, collecting an aggregate image of the aggregate to be detected and preprocessing the aggregate image; 6, acquiring aggregate morphological characteristic parameter values corresponding to the to-be-detected aggregates; and 7, substituting the aggregate morphological characteristic parameters of the aggregate to be detected into the aggregate shape classification model to obtain a shape classification result. According to the method, the problem that aggregate features are difficult to extract is solved, the defect that in the prior art, the classification result is not accurate enough isovercome, aggregate shape classification is conducted through machine learning, existing aggregate morphological features and aggregate shape categories are learned, and the classification result is high in efficiency and accuracy.

Description

technical field [0001] The invention belongs to the field of road engineering and relates to a method for classifying aggregate shapes based on machine learning. Background technique [0002] The morphological characteristics of aggregates mainly include four aspects: shape, size, angularity, and surface texture. Among them, the difference in the shape of the aggregate will lead to the difference in the water retention and cohesion of the concrete, which will affect the deformation of the asphalt pavement. Aggregates with good shapes such as multi-edged corners can effectively reduce the deformation of the pavement. Optimal amount of asphalt can also effectively reduce the deformation of the mixture at high temperature; the difference in the shape of the aggregate will also affect the fatigue resistance of the asphalt pavement, and aggregates with good shapes such as multi-edged edges can increase the hardness of the asphalt mixture, thereby Improve the fatigue resistance o...

AI Technical Summary

Problems solved by technology

[0007] Among them, the first two problems are often encountered in the collection of aggregate images, which will cause difficulties in the extraction of aggregate features. Generally, they are solved by constructing morphological features of aggregate images with obvious visual resolution capabilities. The morphological characteristic parameters of the category (such as the basic characteristic parameters and shape characteristic parameters of the aggregate) describe the aggregate, but this method has high requirements for the placement of aggregate particles, which usually affects the accuracy of the aggregate classification results. degree; and the third problem is still no effective solution

Images