Building envelope state monitoring method

Abstract

The invention relates to a building envelope state monitoring method, which comprises the following steps of providing an unmanned aerial vehicle, and loading a camera and a position sensor on the unmanned aerial vehicle; the unmanned aerial vehicle is controlled to inspect the building facade, and picture data and corresponding position information of the building facade are collected through the camera and the position sensor; the collected picture data are preprocessed to form data fragments, and the data fragments and the corresponding position information are stored; analyzing the building facade in each data segment and detecting whether the building facade has an abnormal area, if not, outputting the data segment and marking the data segment as normal, and if the data segment has an abnormal area, outputting the data segment and marking the data segment as abnormal; and feeding back the data fragments marked as abnormal and the corresponding position information to a maintenance management department.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of building maintenance and detection, in particular to a building envelope state monitoring method. BACKGROUND

[0002] With the continuous advancement of urban renewal and building adaptive reconstruction, the maintenance and repair of large building facades have become important problems to be solved. Traditionally, the detection of building facades mainly relies on manual inspection or simple photographic equipment for visual inspection. This method has the following disadvantages:

[0003] Low efficiency: manual inspection has a long cycle and limited coverage, which cannot meet the real-time and high-frequency detection requirements of large-scale buildings.

[0004] Easily affected by subjective factors: different personnel are affected by experience and judgment ability during inspection, which may cause misjudgment or missed detection, affecting the accuracy of maintenance decisions.

[0005] It is difficult to accurately estimate the cost in advance: since a large amount of manpower and equipment are required for detection work, it is difficult to quantify the cost in advance, and the maintenance cost is uncontrollable. SUMMARY

[0006] The purpose of the present application is to overcome the defects of the prior art and provide a building envelope state monitoring method. The unmanned aerial vehicle patrols the building envelope, thereby detecting the damaged area of the building facade. When damage or aging is detected in the photographed picture, the picture is transmitted to the maintenance management department for warning.

[0007] The technical solution for achieving the above-mentioned purpose is a building envelope state monitoring method, comprising the following steps:

[0008] Providing an unmanned aerial vehicle, loading a camera and a position sensor on the unmanned aerial vehicle;

[0009] Controlling the unmanned aerial vehicle to inspect the building facade, and collecting picture data and corresponding position information of the building facade through the camera and the position sensor;

[0010] Pretreating the collected picture data to form data segments, storing the data segments and the corresponding position information;

[0011] Analyzing the building facade in each data segment and detecting whether there is an abnormal area in the building facade. If there is no abnormal area, the data segment is output and marked as normal. If there is an abnormal area, the data segment is output and marked as abnormal;

[0012] The data segment marked as abnormal and the corresponding position information are fed back to a maintenance management department.

[0013] Further, when controlling the unmanned aerial vehicle to inspect the building facade, a flight path is planned according to the building arrangement on site, and the building facade is inspected according to the flight path.

[0014] Further, when setting the patrol route, a timing patrol instruction is set for the unmanned aerial vehicle according to actual needs on site.

[0015] Further, when preprocessing the picture data, the collected picture data is denoised, compressed and format-converted through an EdgeX Foundry edge computing framework.

[0016] Further, when preprocessing the picture data, the position information is data-filtered and integrated through the EdgeX Foundry edge computing framework.

[0017] Further, after the preprocessing of the picture data is completed, the preprocessed data segment and the corresponding position information are stored by using a Minio distributed object storage system.

[0018] Further, after the analysis and detection of the data segment are completed, a building maintenance database and a comparison program are provided, the data segment marked as abnormal is imported into the comparison program, the comparison program matches the data segment marked as abnormal with building damages in the building maintenance database, if the matching is successful, a maintenance suggestion in the building maintenance database is output, and the maintenance suggestion is fed back to the maintenance management part, if the matching fails, no maintenance suggestion is output, and no maintenance suggestion is fed back to the maintenance management part.

[0019] Further, after the unmanned aerial vehicle collects the picture data of the building facade and the corresponding position information, the picture data and the corresponding position information are transmitted by using an NB-IoT protocol.

[0020] Compared with the prior art, the present application has the following beneficial effects:

[0021] The building envelope is patrolled by the unmanned aerial vehicle, so that the damage area of the facade of the building is detected, when there is damage or aging in the photographed picture, the picture is transmitted to the maintenance management department for warning. DETAILED DESCRIPTION

[0022] The present application will be further described below in conjunction with specific embodiments.

[0023] The application discloses a building envelope state monitoring method, comprising the following steps: providing a UAV, loading a camera and a position sensor on the UAV; controlling the UAV to patrol a building facade, and collecting picture data and corresponding position information of the building facade through the camera and the position sensor; preprocessing the collected picture data to form data segments, storing the data segments and the corresponding position information; analyzing the building facade in each data segment and detecting whether the building facade has an abnormal area, if not, outputting the data segment and marking it as normal, if yes, outputting the data segment and marking it as abnormal; feeding back the data segment marked as abnormal and the corresponding position information to a maintenance management department.

[0024] In the application, when the building envelope needs to be detected, the UAV is started to carry the camera and the position sensor, the UAV is controlled to fly along the edge of the building, and the camera and the position sensor on the UAV record pictures and position information of the building envelope. Preferably, the UAV is provided with a wireless signal transmission device, the wireless signal transmission device is connected with the camera and the position sensor through wires, a working base station is arranged outside the building, the wireless signal transmission device is connected with the working base station through wireless signal communication, the camera and the position sensor can transmit the recorded picture data and position information to the working base station through the wireless signal transmission device, so that the picture data and the position information can be processed more timely, the picture data and the position information are preprocessed, preprocessing includes data cleaning, format conversion, noise reduction and preliminary analysis, data transmission delay is reduced, cloud processing pressure is reduced, when the data segments are analyzed, a picture of a perfect building facade is provided for comparison, the perfect picture and the preprocessed data segments are analyzed, whether the data segments have abnormalities such as damage and aging is analyzed and judged, preferably, the perfect picture is a picture taken by the UAV when the building is perfect, the angle and the position of the UAV when the picture is taken are recorded, when the UAV is used to patrol the building envelope subsequently, the UAV is used to take pictures at the same angle and position as when the perfect building facade is taken, the picture of the perfect building facade is compared with the picture taken when the building is patrolled, whether the data segments have abnormalities is analyzed and judged, the data segments marked as normal and abnormal are classified, and the data segments marked as abnormal are sent to the maintenance management department, then the maintenance management department can confirm corresponding positions through the position information on the data segments, so that subsequent maintenance work is facilitated.

[0025] Further, when controlling the unmanned aerial vehicle to inspect the building facade, a route is planned according to the arrangement of the building on site, and the building facade is inspected according to the inspection route. In use, the inspection route can be planned according to the distribution environment of the building on site, and then a corresponding operation script can be written for the unmanned aerial vehicle, so that the unmanned aerial vehicle inspects along the inspection route, thereby reducing the workload of construction personnel.

[0026] Further, when setting the patrol route, a timing patrol instruction is set for the unmanned aerial vehicle according to the actual needs on site. Preferably, a script for the unmanned aerial vehicle to work at a timing can be written according to the actual requirements on site.

[0027] Further, when preprocessing the picture data, the collected picture data is denoised, compressed and format-converted through the EdgeX Foundry edge computing framework. By denoising, compressing and format-converting the picture data, the data transmission delay is reduced, and the cloud processing pressure is reduced.

[0028] Further, when preprocessing the picture data, the position information is filtered and integrated through the EdgeX Foundry edge computing framework. By filtering and integrating the position information, the data transmission delay is reduced, and the cloud processing pressure is reduced.

[0029] Further, after the preprocessing of the picture data is completed, the preprocessed data segments and the corresponding position information are stored by using the Minio distributed object storage system. By using the Minio distributed object storage system to efficiently manage the preprocessed data segments and the position information, the safe storage and convenient retrieval of massive data are ensured, and the metadata information of the inspection is recorded for subsequent data tracing and maintenance management.

[0030] Further, after the analysis and detection of the data segments are completed, a building maintenance database and a comparison program are provided, the data segments marked as abnormal are imported into the comparison program, the comparison program matches the data segments marked as abnormal with the building damages in the building maintenance database, if the matching is successful, a maintenance suggestion in the building maintenance database is output, and the maintenance suggestion is fed back to the maintenance management part, if the matching fails, no maintenance suggestion is output, and no maintenance suggestion is fed back to the maintenance management part. Preferably, TensorFlow deep learning algorithm is used to batch recognize the image data stored in Minio, a large number of building facade image data marked as “normal”, “repair” and “replace” are used in the training stage of the TensorFlow deep learning algorithm to establish a high-precision image recognition model, and in the detection stage, the preprocessed image data is batched, inferred and automatically classified to automatically determine the “normal” or “abnormal” of the building facade, preferably, the TensorFlow deep learning algorithm can be deepened to further detect the “abnormal” state, thereby classifying the state such as “repair” or “replace”, and marking the specific damage area; according to the recognition result, combining the building maintenance database and the preset rules, the corresponding repair or replacement suggestion is automatically generated to support the subsequent decision and maintenance.

[0031] Further, after the unmanned aerial vehicle collects the picture data of the building facade and the corresponding position information, the NB-IoT protocol is used to transmit the picture data and the corresponding position information.

[0032] The use process of the building envelope state monitoring method is described below.

[0033] When the building envelope needs to be detected, the unmanned aerial vehicle is started to carry a camera and a position sensor, the unmanned aerial vehicle is controlled to fly along the edge of the building, and the camera and the position sensor on the unmanned aerial vehicle record photos of the building envelope and position information, preferably, the unmanned aerial vehicle is provided with a wireless signal transmission device, the wireless signal transmission device is connected with the camera and the position sensor through wires, a working base station is arranged outside the building, the wireless signal transmission device is connected with the working base station through wireless signal communication, the camera and the position sensor can transmit recorded picture data and position information to the working base station through the wireless signal transmission device, so that the picture data and the position information can be processed more timely, the picture data and the position information are preprocessed, preprocessing includes data cleaning, format conversion, noise reduction and preliminary analysis, data transmission delay is reduced, cloud processing pressure is reduced, when data segments are analyzed, photos of a perfect building facade are provided for comparison, the perfect photos and the preprocessed data segments are analyzed, whether the data segments are abnormal, such as damaged and aged, is analyzed and judged, the data segments marked as normal and abnormal are classified, and the data segments marked as abnormal are sent to a maintenance management department, then the maintenance management department can go to the corresponding place according to the position information on the data segments to confirm, so that subsequent maintenance work is facilitated.

[0034] The above describes the present application in detail in combination with the embodiments, and those skilled in the art can make various changes to the present application according to the above description. Therefore, some details in the embodiments should not constitute a limitation on the present application, and the present application will be limited by the scope defined in the appended claims.

Claims

1. A method for monitoring the condition of a building envelope, characterized in that, Includes the following steps: Provide a drone, on which a camera and a position sensor are mounted; The drone is controlled to inspect the building facade and collect image data and corresponding location information of the building facade through the camera and position sensor; The collected image data is preprocessed to form data fragments, and the data fragments and corresponding location information are stored. The building facade in each data segment is analyzed and the presence of abnormal areas is detected. If no abnormal areas are found, the data segment is output and marked as normal. If abnormal areas are found, the data segment is output and marked as abnormal. The data segments marked as abnormal and the corresponding location information are fed back to the maintenance management department.

2. The method for monitoring the condition of a building envelope according to claim 1, characterized in that: When controlling the drone to inspect the building facade, the inspection route is planned according to the building layout on site, and the building facade is inspected according to the inspection route.

3. The method for monitoring the condition of a building envelope according to claim 2, characterized in that: When setting the patrol route, a timed patrol command is set for the drone according to the actual needs on site.

4. The method for monitoring the condition of a building envelope according to claim 1, characterized in that: in When preprocessing the image data, the EdgeX Foundry edge computing framework is used to perform noise reduction, compression, and format conversion on the acquired image data.

5. The method for monitoring the condition of a building envelope according to claim 1, characterized in that: in When preprocessing the image data, the location information is filtered and integrated using the EdgeX Foundry edge computing framework.

6. The method for monitoring the condition of a building envelope according to claim 1, characterized in that: in After the image data preprocessing is completed, the Minio distributed object storage system is used to store the preprocessed data fragments and their corresponding location information.

7. A method for monitoring the condition of a building envelope according to claim 1, characterized in that: in After the data fragments are analyzed and detected, a building maintenance database and a comparison program are provided. The data fragments marked as abnormal are imported into the comparison program, which matches the data fragments marked as abnormal with the building damage in the building maintenance database. If the match is successful, a maintenance suggestion in the building maintenance database is output and fed back to the maintenance management department. If the match fails, no maintenance suggestion is output and fed back to the maintenance management department.

8. The method for monitoring the condition of a building envelope according to claim 1, characterized in that: After the drone collects image data and corresponding location information of the building facade, the image data and corresponding location information are transmitted using the NB-IoT protocol.