CN-121998920-A - Method for searching and identifying road diseases by using persistent coherence

Abstract

The invention discloses a method for searching and identifying road diseases by using persistent coherence, and relates to the technical field of road defect detection. A topology data analysis method based on continuous coherence utilizes the persistent coherence to find road diseases. The method comprises the following steps of rescaling an image, finding out a topological structure by using persistent coherence, bonding a near-position area, and analyzing the shapes and the sizes of different segmentation areas to detect and identify the road diseases. The topology fingerprint generation method based on continuous coherence provides a novel and powerful technical path for solving the problem of extracting the essential characteristics of the two-dimensional object, and can more accurately find and identify the road diseases.

Inventors

• ZHENG JIAXI
• SHI WEIBIN

Assignees

• 伟乐科技集团有限公司

Dates

Publication Date

20260508

Application Date

20260109

Claims (1)

1. 1. A method for locating and identifying road conditions using persistent co-ordination, comprising the steps of: step 1, rescaling the image; The method comprises the steps of obtaining a pavement three-dimensional picture by using laser scanning, and firstly reducing the image according to convolution; Step 2, finding out a topological structure by using persistent coherence; Calculating the three-dimensional image slice, and only considering 0-dimensional coherence and 1-dimensional coherence; Finding out a describing part of which the topological structure is a three-dimensional image by using a standard lasting coherent method, namely, which points on a road are selected, wherein the rising value is 1, and the falling value is 0, and the describing part records the topological structure of the road disease; Step 3, bonding the near-position area; Because a road disease may be composed of a plurality of topological structures, namely, in the step 2, the road disease is covered by a plurality of depiction parts, a depth threshold is set in consideration of a position which is provided with a large number of persistent congruence and is close to the position, only the region with the persistent value larger than the depth threshold is considered, and then the region at the position is adhered to form a road disease score; step 4, analyzing the shapes and the sizes of different segmentation areas to detect and identify the road diseases; different road diseases have different topological property characteristics, the persistent graphs of the same defect are similar to each other, and the persistent harmony of the defects of different types is analyzed and used as the fingerprint of the road; step 5, after the fingerprints are sleeved, different categories can be distinguished, and then the road diseases corresponding to each group are separated; taking into account the descriptive part of all the near-adhering regions in step 3 and their persistent coherent information, these associations are separated according to an unsupervised artificial intelligence model, and then it is known which kind of road disease is in the new image with the association learned by the unsupervised artificial intelligence.

Description

Method for searching and identifying road diseases by using persistent coherence Technical Field The invention relates to the technical field of road defect detection. Background Various types of damage to the road can occur under the influence of long-term use and natural factors. These diseases not only affect the driving comfort, but also threaten the traffic safety. The following are several common road diseases and their descriptions: 1. cracks. Cracks are one of the most common and earliest occurring diseases of asphalt pavement, and are mainly classified into longitudinal cracks, transverse cracks and net-shaped cracks (or called cracks and blocky cracks) according to the form and the cause of the cracks. The longitudinal crack is basically parallel to the central line of the road and takes a strip shape, and is mainly caused by uneven settlement of roadbed or base layer, improper treatment of construction joints or different compactness of the joint part of the new roadbed and the old roadbed when the old roadbed is widened, and in addition, the longitudinal crack is caused by shrinkage of materials under the condition of temperature change. The transverse crack is mainly a temperature shrinkage crack, when the temperature is reduced, the asphalt surface layer and the base layer material shrink to generate larger tensile stress, when the stress exceeds the tensile strength of the material, the pavement is pulled apart, and the crack usually occurs at regular and equidistant intervals. A mesh fracture is a series of alternating fractures that divide the pavement into polygonal, tortoiseshell or mesh-like pieces. Depending on the size of the pieces, they can be classified into fine mesh and wide mesh. Network cracks are a typical manifestation of fatigue damage to pavement structures, and usually due to insufficient strength of pavement structure layers (such as base layers), fatigue damage is generated under the repeated action of traffic load, and network cracks appear in local areas, which means that the structural bearing capacity of the area is severely reduced. 2. And (5) deformation class. Such diseases are mainly manifested by changes in road surface shape, affecting the stability of the driving, including too deep ruts, too high congestion, and too deep subsidence. The ruts are longitudinal and strip-shaped grooves formed on the lane wheel belts, and the pavement images sink to form two channels when seen in the cross section, and the ruts can be classified into compact ruts, flowable ruts and structural ruts. The bulge is formed by local bulge of the road surface, is usually associated with rutting, and is caused by shearing flow and accumulation of an asphalt surface layer under the combined action of horizontal force (such as braking and starting) and high temperature of the wheel. Subsidence is a depression formed after obvious subsidence occurs in a local area of a road surface, and is mainly caused by uneven soil quality of a roadbed, insufficient compaction or hollowed out of the roadbed caused by leakage, excavation and the like of an underground pipeline, thereby causing uneven subsidence. 3. Loosening and pit slots. Such diseases are directly manifested as loss of pavement material, forming voids. Pit grooves are pits of various shapes formed after the pavement local materials are scattered, the depth is usually more than 2 cm, and the pit grooves can form instant threat to the driving safety as a result of further worsening of other diseases (such as cracks and looseness). Loosening is the falling of aggregate particles on the surface layer of the pavement, and the scattering of coarse and fine aggregates is mainly caused by poor adhesion (peeling) between asphalt and aggregate, or too little asphalt consumption in asphalt mixture and aging of asphalt. 4. Surface damage class. Such damage can be classified into oiling and polishing, wherein oiling is the appearance of a layer of free asphalt on the road surface during high temperature seasons, which makes the road surface a smooth, reflective black mirror. The polishing is that the edges and corners of the aggregate on the surface layer of the road surface are polished by the wheels, the road surface is smooth, and the anti-skid performance is greatly reduced. The common road diseases are usually related to each other, for example, cracks can develop into pits if not treated in time, water is easy to accumulate in ruts, and road surface damage is accelerated. Therefore, it is important to find and take the correct maintenance measures in time. The conventional method for detecting road defects of a road is to observe with naked eyes, however, the method has the following disadvantages: Firstly, the subjectivity is strong, the standards are difficult to unify, and different detection personnel can have differences in definition, severity and type of diseases. For example, the same crack, one may be judged as "medium" and another may