Method for automatically monitoring construction progress based on machine vision

Abstract

The invention discloses a method for automatically monitoring the construction progress based on machine vision, and the method comprises the steps: dividing a construction region and a non-construction region in a construction project general arrangement drawing of a construction site, and using a building individual in a dynamic region as a single recognition object; rasterizing the dynamic region and a steady region to form a binary raster image carrying monitoring region range information; performing three-dimensional reconstruction on the construction site scene based on the reference images of the construction site and the surrounding scene; performing masking processing on the construction region in a reference photo, extracting the non-construction region in the reference photo, taking the extracted non-construction region as a registration primitive, performing conventional monitoring measurement on the construction site, and obtaining construction site photos of a to-be-monitored conventional monitoring batch; and performing three-dimensional reconstruction of the construction scene again in combination with the registration primitives to generate three-dimensional pointcloud, DSM and DOM, extracting the elevation of the working surface of each building individual from the monitored DSM or three-dimensional point cloud, and realizing automatic monitoring of the construction progress through elevation comparison.

Description

technical field [0001] The invention relates to the technical field of machine vision, in particular to a method for automatically monitoring construction progress based on machine vision. Background technique [0002] Construction progress monitoring is very necessary for project management. At present, the commonly used manual inspection method is time-consuming and labor-intensive, and the records are prone to errors. UAVs are a good monitoring tool. UAVs carry cameras for monitoring and can obtain a good monitoring perspective. At present, there are also application examples of using UAVs for construction site monitoring, but most of them are through manual interpretation of aerial photos. Identifying the specific buildings and counting the construction floors in a crowded construction site still has the problems of labor-intensive and error-prone. [0003] In recent years, relatively advanced methods for automatic analysis of construction site images have also emerged...

AI Technical Summary

Problems solved by technology

Strictly speaking, for subsequent monitoring batches, it is also necessary to specify the parts that are consistent with the appearance of the previous batch, otherwise it may lead to registration failure; since there is currently no better way to automatically determine the parts with the same appearance, Therefore, using the ICP algorithm to register with the 3D reconstruction model will reduce the automation level of data processing; the "image control-free" photogrammetry technology that has emerged in recent years can well avoid the problem of specifying image control points each time, but these "free image control" The photogrammetry technology of "image control" uses photographic equipment that can obtain high-precision GNSS coordinates such as CORS and RTK (such as drones equipped with RTK), and will put forward more technical requirements for networks, links, base stations, etc. , and the cost is higher

Images