Construction progress monitoring device and method based on neural network

Abstract

The invention relates to a construction progress monitoring method based on a neural network, comprising: step 1: setting up a plurality of cameras around the construction area, step 2: using a drone to guide and set the camera pose, step 3: after setting After the camera poses, you can take pictures at a certain frequency and upload them to the management platform. The management platform uses the neural network for deep learning to identify the construction progress. The monitoring method proposed by the invention can simultaneously take into account training time and recognition accuracy. And it can automatically adjust the camera pose as the construction progresses, and carry out automatic progress monitoring.

Description

technical field [0001] The invention relates to the field of construction progress monitoring, in particular to a construction progress monitoring device and method based on a neural network system. Background technique [0002] At present, the management of construction progress mainly relies on manual records, which is time-consuming and labor-intensive. There are also people who propose a method of monitoring construction site equipment to monitor construction progress. For example, monitor the serial number and grouting pressure of the bolt currently installed by the bolter rig. This approach requires a large number of device-related sensors. The construction site usually has more equipment, so more types of sensors are required. Moreover, interconnection transformation of equipment and sensors is required, which is costly and time-consuming. And the construction environment is harsh, and the sensor is prone to failure. [0003] At present, some people have proposed...

AI Technical Summary

Problems solved by technology

Moreover, interconnection transformation of equipment and sensors is required, which is costly and time-consuming

And the construction environment is harsh, the sensor is prone to failure

[0003] At present, some people have proposed the method of using neural network to monitor the construction progress, but in the neural network recognition, not only did not optimize the shooting image settings (neither hardware optimization nor software optimization), but also did not optimize the neural network structure. , so that the normal neural network takes a long time to train, and the recognition accuracy is average

Images