Monitoring method of dangerous rock collapse based on infrared thermal imaging technology

Abstract

The invention relates to the field of geological disaster monitoring, and discloses a dangerous rock collapse monitoring method based on infrared thermal imaging technology. The method comprises the following steps: arranging an infrared thermal imager in a dangerous rock target monitoring area to monitor and capture a thermal image of the target monitoring area, and then dividing the area into a plurality of grid units. The temperature data of each grid unit is extracted and transmitted to a data processing center for data preprocessing. The reference temperature of the temperature data of each grid unit is analyzed. The reference temperature is compared with the actual temperature of each grid unit to screen out grid units with abnormal temperatures. Cluster analysis is performed on adjacent grid units with abnormal temperatures screened out to identify whether there is a collapse risk area in the target monitoring area. A random forest model is then used to predict whether there is a collapse risk in the target monitoring area. The method is conducive to ensuring the continuity of monitoring data, detecting potential risks abnormally and quickly, and improving the collapse risk identification rate and the accuracy of risk identification results.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disaster monitoring, and more specifically to a dangerous rock collapse monitoring method based on infrared thermal imaging technology. Background Art

[0002] At present, the monitoring methods for dangerous rock collapse mainly include buried sensors, radar monitoring and video monitoring. The traditional collapse monitoring method mainly uses traditional buried sensors to carry out wiring and on-site construction in dangerous rocks. However, the traditional collapse monitoring has the following main disadvantages:

[0003] Traditional methods of collapse monitoring have many problems such as complex sensor installation, high cost, risk of data loss, and recognition rate. Traditional collapse monitoring mainly involves burying sensors through a large amount of wiring, which is difficult and dangerous, especially in remote areas or areas with complex terrain, where sensor installation and maintenance are more difficult. Although radar monitoring has a wide coverage range, it is costly and has a large monitoring granularity. Video monitoring is limited by visual conditions and cannot work in severe weather and low-light environments, making it difficult to meet the needs of emergency monitoring for rapid data acquisition. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a dangerous rock collapse monitoring method based on infrared thermal imaging technology to solve the problems existing in the above-mentioned background technology.

[0005] The present invention provides the following technical solution: a dangerous rock collapse monitoring method based on infrared thermal imaging technology, comprising:

[0006] S1: used to set the dangerous rock target monitoring area, and to monitor the temperature changes in the dangerous rock target monitoring area in real time by arranging infrared thermal imagers in the dangerous rock target monitoring area;

[0007] S2: regularly capture thermal images of the target monitoring area through an infrared thermal imager, divide the captured thermal images into a number of grid units, extract temperature data of each grid unit from the thermal images, and transmit the temperature data of each grid unit to S3;

[0008] S3: pre-process the temperature data of each grid unit separately through the data processing center, and transmit the pre-processed temperature data of each grid unit to S4;

[0009] S4: By performing feature analysis on the temperature data of each grid unit, a reference temperature is obtained, and then a temperature deviation index is calculated based on a comparison between the reference temperature and the actual temperature of each grid unit. The grid units with abnormal temperatures are screened out based on the temperature deviation value, and the screening analysis results are transmitted to S5;

[0010] S5: By performing cluster analysis on the adjacent grid cells with screened temperature anomalies, a collapse risk identification coefficient is obtained, thereby identifying whether there is a collapse risk area in the target monitoring area, and transmitting the identification result to S6;

[0011] S6: Calculate the collapse risk prediction coefficient through the random forest model, and predict whether the collapse risk occurs in the target monitoring area through the collapse risk prediction coefficient;

[0012] S7: Feedback the collapse risk prediction results and automatically generate a regional risk detection report and send it to the management terminal.

[0013] Preferably, the S1 selects a suitable dangerous rock target monitoring area, and selects several key points in the dangerous rock target monitoring area that are susceptible to collapse to arrange infrared thermal imagers, the selected key points include rock cracks and faults, areas with unstable rock structures, and locations where collapse or landslides have occurred before, and then adjusts the shooting angle and focal length of the infrared thermal imager to ensure that the infrared thermal imager clearly captures the thermal image of the target monitoring area, thereby comprehensively covering the entire monitoring area, and starting the adjusted infrared thermal imager to monitor the temperature data of the rock surface in real time.

[0014] Preferably, S2 uses an infrared thermal imager to regularly capture thermal images of the dangerous rock target monitoring area, then divides the thermal images of the dangerous rock target monitoring area into regions, divides the monitoring area into a number of grid units, and numbers each grid unit as 1, 2, 3...n, and then extracts the temperature data in each grid unit, and at the same time, automatically transmits the extracted temperature data of each grid unit to the data processing center through the data transmission interface at the collection time point.

[0015] Preferably, S3 performs denoising processing on the collected temperature data of each grid unit through a denoising algorithm, including median filtering, Gaussian filtering, non-local mean filtering and wavelet transform on the temperature value of each grid unit, and then checks the processing result of the temperature data of each grid unit.

[0016] Preferably, S4 performs feature analysis on the temperature data of each grid unit collected for a period of time and calculates the reference temperature. The specific calculation formula is: Among them, T ij represents the temperature value of the jth acquisition time point of the i-th grid unit, and N represents the number of acquisition time points;

[0017] The temperature deviation index is analyzed and calculated by comparing the actual temperature of each grid cell with the reference temperature using the Z-score.

[0018] The specific analysis method of the temperature deviation index is:

[0019] Step S411: By analyzing the temperature data of each grid unit collected over a period of time, the temperature standard deviation of each grid unit is calculated as Indicates the reference temperature;

[0020] Step S412: By reference temperature and the standard deviation of the temperature of each grid cell σ i Analyze and calculate the temperature deviation index T i Represents the current temperature value of the i-th grid cell;

[0021] By setting a temperature deviation threshold θ, the temperature deviation index z i Compare with the temperature deviation threshold θ to determine whether the temperature of each grid unit is abnormal; if the temperature deviation index z i ≤ temperature deviation threshold θ, indicating that there is no abnormality in the temperature of the grid unit, and the temperature of the grid unit will continue to be monitored; if the temperature deviation index z i >temperature deviation threshold θ, indicating that the temperature in the grid unit is abnormal, all grid units with temperature anomalies are screened out, and the screening results of all grid units with temperature anomalies are transmitted to S5.

[0022] Preferably, S5 uses a spatial clustering algorithm to cluster adjacent grid cells with abnormal temperatures based on the screening analysis results, and analyzes and calculates a collapse risk identification coefficient using a DBSCAN algorithm. The specific calculation formula of the collapse risk identification coefficient is: Among them, M core Indicates the number of core points in the cluster, M total represents the number of all points in the cluster, σ temp represents the standard deviation of the temperature of all points in the cluster, Z avg Represents the average temperature deviation index of all points in the cluster;

[0023] The collapse risk identification coefficient R is compared with the preset collapse risk threshold λ to identify whether there is a collapse risk area in the target monitoring area. If the collapse risk identification coefficient ≤ R preset collapse risk threshold λ, it indicates that there is no collapse risk in the target monitoring area, and the temperature in the target monitoring area continues to be monitored. If the collapse risk identification coefficient is greater than R preset collapse risk threshold λ, it indicates that there is a collapse risk in the target monitoring area, and the identification analysis result is transmitted to S6.

[0024] Preferably, the S6 uses a supervised learning method, selects a random forest model to construct a prediction model, trains the random forest model through time-space temperature data and known collapse event data, and then inputs the time-space temperature data into the trained random forest model to predict the collapse risk of the target monitoring area, and automatically outputs the probability of collapse of each sample through the random forest model, so as to calculate the collapse risk prediction coefficient by the formula C=P(y=1|x), where x represents the input variable, and y represents the binary variable, i.e., whether a collapse event has occurred, y=1 indicates that a collapse has occurred, and y=0 indicates that no collapse has occurred.

[0025] Technical effects and advantages of the present invention:

[0026] The present invention arranges an infrared thermal imager in the dangerous rock target monitoring area to monitor the temperature change in the dangerous rock target monitoring area in real time, and regularly captures the thermal image of the target monitoring area through the infrared thermal imager, divides the captured thermal image into a number of grid units, extracts the temperature data of each grid unit from the thermal image, and transmits it to the data processing center for data preprocessing, and then performs feature analysis on the processed temperature data of each grid unit to obtain a reference temperature, and then compares and analyzes the reference temperature with the actual temperature of each grid unit, thereby screening out grid units with abnormal temperatures, and performing cluster analysis on adjacent grid units with abnormal temperatures screened out to identify whether there is a collapse risk area in the target monitoring area, and then predicts the collapse of the target area through a random forest model. Whether there is a risk of collapse in the target monitoring area, the use of infrared thermal imaging technology avoids the complicated installation process and realizes the convenient deployment of equipment. It is not affected by light and weather conditions, so as to monitor the changes of dangerous rocks around the clock. It has relatively low effective cost and high monitoring accuracy. At the same time, it has a long service life and low maintenance costs, which is conducive to reducing the risk of loss during data transmission and ensuring the continuity of monitoring data. The analysis of temperature field changes can identify subtle anomalies, which provides a more accurate data basis for the long-term monitoring and analysis of geological disaster risks. It can quickly detect potential risks through temperature field anomalies, improve the collapse risk identification rate and the accuracy of risk identification results, thereby reducing interference with natural landforms and helping to maintain the ecological balance of the monitored area. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The figure is a flow chart of the dangerous rock collapse monitoring method based on infrared thermal imaging technology of the present invention. DETAILED DESCRIPTION

[0028] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following embodiments are merely illustrative. The dangerous rock collapse monitoring method based on infrared thermal imaging technology involved in the present invention is not limited to the various structures recorded in the following embodiments. All other implementations obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0029] like Figure 1 The embodiment shown provides a dangerous rock collapse monitoring method based on infrared thermal imaging technology, including:

[0030] S1: It is used to set the dangerous rock target monitoring area, and to monitor the temperature changes in the dangerous rock target monitoring area in real time by arranging infrared thermal imagers in the dangerous rock target monitoring area.

[0031] In this embodiment, the S1 selects a suitable dangerous rock target monitoring area, and selects several key points in the dangerous rock target monitoring area that are susceptible to collapse to arrange infrared thermal imagers, the selected key points include rock cracks and faults, areas with unstable rock structures, and locations where collapse or landslides have occurred before, and then adjusts the shooting angle and focal length of the infrared thermal imager to ensure that the infrared thermal imager clearly captures the thermal image of the target monitoring area, thereby fully covering the entire monitoring area, and starting the adjusted infrared thermal imager to monitor the temperature data of the rock surface in real time.

[0032] S2: Regularly capture thermal images of the target monitoring area through an infrared thermal imager, divide the captured thermal images into a number of grid units, extract temperature data of each grid unit from the thermal images, and transmit the temperature data of each grid unit to S3.

[0033] In this embodiment, the S2 uses an infrared thermal imager to regularly capture thermal images of the dangerous rock target monitoring area, and then divides the thermal images of the dangerous rock target monitoring area into regions, dividing the monitoring area into a number of grid units, and numbering each grid unit as 1, 2, 3...n, and then extracting the temperature data in each grid unit. At the same time, the extracted temperature data of each grid unit is automatically transmitted to the data processing center through the data transmission interface at the collection time point.

[0034] It should be specifically noted that the selection of the grid size for the dangerous rock target monitoring area depends on the spatial scope of on-site monitoring, and requires a balance between processing speed and monitoring accuracy. When monitoring relatively small risk areas, a smaller grid can be set, which can provide higher resolution.

[0035] S3: The temperature data of each grid unit is preprocessed separately by the data processing center, and the preprocessed temperature data of each grid unit is transmitted to S4.

[0036] In this embodiment, S3 performs denoising on the collected temperature data of each grid unit through a denoising algorithm, including median filtering, Gaussian filtering, non-local mean filtering and wavelet transform on the temperature value of each grid unit, and then checks the processing results of the temperature data of each grid unit.

[0037] S4: By performing characteristic analysis on the temperature data of each grid unit, a reference temperature is obtained, and then a temperature deviation index is calculated based on a comparison between the reference temperature and the actual temperature of each grid unit. The grid units with abnormal temperatures are screened out based on the temperature deviation value, and the screening analysis results are transmitted to S5.

[0038] In this embodiment, S4 performs feature analysis on the temperature data of each grid unit collected for a period of time and calculates the reference temperature. The specific calculation formula is: Among them, T ij represents the temperature value of the jth acquisition time point of the i-th grid unit, and N represents the number of acquisition time points;

[0039] The temperature deviation index is analyzed and calculated by comparing the actual temperature of each grid cell with the reference temperature using the Z-score.

[0040] The specific analysis method of the temperature deviation index is:

[0041] Step S411: By analyzing the temperature data of each grid unit collected over a period of time, the temperature standard deviation of each grid unit is calculated as Indicates the reference temperature;

[0042] Step S412: By reference temperature and the standard deviation of the temperature of each grid cell σ i Analyze and calculate the temperature deviation index T i Represents the current temperature value of the i-th grid cell;

[0043] By setting a temperature deviation threshold θ, the temperature deviation index z i Compare with the temperature deviation threshold θ to determine whether the temperature of each grid unit is abnormal; if the temperature deviation index z i ≤ temperature deviation threshold θ, indicating that there is no abnormality in the temperature of the grid unit, and the temperature of the grid unit will continue to be monitored; if the temperature deviation index z i>temperature deviation threshold θ, indicating that the temperature in the grid unit is abnormal, all grid units with temperature anomalies are screened out, and the screening results of all grid units with temperature anomalies are transmitted to S5.

[0044] S5: By performing cluster analysis on adjacent grid cells with screened temperature anomalies, a collapse risk identification coefficient is obtained to identify whether there is a collapse risk area in the target monitoring area, and the identification result is transmitted to S6.

[0045] In this embodiment, S5 uses a spatial clustering algorithm to cluster adjacent grid cells with abnormal temperatures based on the screening analysis results, and analyzes and calculates the collapse risk identification coefficient through the DBSCAN algorithm. The specific calculation formula of the collapse risk identification coefficient is: Among them, M core Indicates the number of core points in the cluster, M total represents the number of all points in the cluster, σ temp represents the standard deviation of the temperature of all points in the cluster, Z avg Represents the average temperature deviation index of all points in the cluster;

[0046] The collapse risk identification coefficient R is compared with the preset collapse risk threshold λ to identify whether there is a collapse risk area in the target monitoring area. If the collapse risk identification coefficient ≤ R preset collapse risk threshold λ, it indicates that there is no collapse risk in the target monitoring area, and the temperature in the target monitoring area continues to be monitored. If the collapse risk identification coefficient is greater than R preset collapse risk threshold λ, it indicates that there is a collapse risk in the target monitoring area, and the identification analysis result is transmitted to S6.

[0047] S6: The collapse risk prediction coefficient is calculated through the random forest model, and the collapse risk prediction coefficient is used to predict whether the target monitoring area has a collapse risk.

[0048] In this embodiment, S6 uses a supervised learning method and selects a random forest model to construct a prediction model. The random forest model is trained through time-space temperature data and known collapse event data. The time-space temperature data is then input into the trained random forest model to predict the collapse risk of the target monitoring area. The probability of collapse of each sample is automatically output through the random forest model, so that the formula for calculating the collapse risk prediction coefficient is C=P(y=1|x), where x represents the input variable and y represents a binary variable, i.e., whether a collapse event has occurred. y=1 indicates that a collapse has occurred, and y=0 indicates that no collapse has occurred.

[0049] S7: Feedback the collapse risk prediction results and automatically generate a regional risk detection report and send it to the management terminal.

[0050] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0051] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A dangerous rock collapse monitoring method based on infrared thermal imaging technology, characterized in that: include: S1: used to set the dangerous rock target monitoring area, and to monitor the temperature changes in the dangerous rock target monitoring area in real time by arranging infrared thermal imagers in the dangerous rock target monitoring area; S2: regularly capture thermal images of the target monitoring area through an infrared thermal imager, divide the captured thermal images into a number of grid units, extract temperature data of each grid unit from the thermal images, and transmit the temperature data of each grid unit to S3; S3: pre-process the temperature data of each grid unit separately through the data processing center, and transmit the pre-processed temperature data of each grid unit to S4; S4: By performing feature analysis on the temperature data of each grid unit, a reference temperature is obtained, and then a temperature deviation index is calculated based on a comparison between the reference temperature and the actual temperature of each grid unit. The grid units with abnormal temperatures are screened out based on the temperature deviation value, and the screening analysis results are transmitted to S5; S5: By performing cluster analysis on the adjacent grid cells with screened temperature anomalies, a collapse risk identification coefficient is obtained, thereby identifying whether there is a collapse risk area in the target monitoring area, and transmitting the identification result to S6; S6: Calculate the collapse risk prediction coefficient through the random forest model, and predict whether the collapse risk occurs in the target monitoring area through the collapse risk prediction coefficient; S7: Feedback the collapse risk prediction results and automatically generate a regional risk detection report and send it to the management terminal.

2. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: The S1 selects a suitable dangerous rock target monitoring area, and selects several key points in the dangerous rock target monitoring area that are susceptible to collapse to arrange infrared thermal imagers, the selected key points include rock cracks and faults, areas with unstable rock structures, and locations where collapse or landslides have occurred before, and then adjusts the shooting angle and focal length of the infrared thermal imager to ensure that the infrared thermal imager clearly captures the thermal image of the target monitoring area, thereby fully covering the entire monitoring area, and starting the adjusted infrared thermal imager to monitor the temperature data of the rock surface in real time.

3. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: The S2 uses an infrared thermal imager to regularly capture thermal images of the dangerous rock target monitoring area, and then divides the thermal images of the dangerous rock target monitoring area into regions, divides the monitoring area into a number of grid units, and numbers each grid unit as 1, 2, 3...n, and then extracts the temperature data in each grid unit. At the same time, the extracted temperature data of each grid unit is automatically transmitted to the data processing center through the data transmission interface at the collection time point.

4. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: The S3 performs denoising processing on the collected temperature data of each grid unit through a denoising algorithm, including median filtering, Gaussian filtering, non-local mean filtering and wavelet transform on the temperature value of each grid unit, and then checks the processing result of the temperature data of each grid unit.

5. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: S4 performs feature analysis on the temperature data of each grid unit collected for a period of time and calculates the reference temperature. The specific calculation formula is: Among them, T ij represents the temperature value of the jth acquisition time point of the i-th grid unit, and N represents the number of acquisition time points; The temperature deviation index is analyzed and calculated by comparing the actual temperature of each grid cell with the reference temperature using the Z-score. The specific analysis method of the temperature deviation index is: Step S411: By analyzing the temperature data of each grid unit collected over a period of time, the temperature standard deviation of each grid unit is calculated as Indicates the reference temperature; Step S412: By reference temperature and the standard deviation of temperature for each grid cell σ i Analyze and calculate the temperature deviation index T i Represents the current temperature value of the i-th grid cell; By setting a temperature deviation threshold θ, the temperature deviation index z i Compare with the temperature deviation threshold θ to determine whether the temperature of each grid unit is abnormal; if the temperature deviation index z i ≤ temperature deviation threshold θ, indicating that there is no abnormality in the temperature of the grid unit, the temperature of the grid unit will continue to be monitored; if the temperature deviation index z i >temperature deviation threshold θ, indicating that the temperature in the grid unit is abnormal, all grid units with temperature anomalies are screened out, and the screening results of all grid units with temperature anomalies are transmitted to S5.

6. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: Based on the screening analysis results, S5 uses a spatial clustering algorithm to cluster adjacent grid cells with abnormal temperatures, and uses the DBSCAN algorithm to analyze and calculate the collapse risk identification coefficient. The specific calculation formula of the collapse risk identification coefficient is: Among them, M core Indicates the number of core points in the cluster, M total represents the number of all points in the cluster, σ temp represents the standard deviation of the temperature of all points in the cluster, Z avg Represents the average temperature deviation index of all points in the cluster; The collapse risk identification coefficient R is compared with the preset collapse risk threshold λ to identify whether there is a collapse risk area in the target monitoring area. If the collapse risk identification coefficient ≤ R preset collapse risk threshold λ, it indicates that there is no collapse risk in the target monitoring area, and the temperature in the target monitoring area continues to be monitored. If the collapse risk identification coefficient is greater than R preset collapse risk threshold λ, it indicates that there is a collapse risk in the target monitoring area, and the identification analysis result is transmitted to S6.

7. The method for monitoring dangerous rock collapse based on infrared thermal imaging technology according to claim 1 is characterized in that: The S6 uses a supervised learning method, selects a random forest model to construct a prediction model, trains the random forest model through time-space temperature data and known collapse event data, and then inputs the time-space temperature data into the trained random forest model to predict the collapse risk of the target monitoring area, and automatically outputs the probability of collapse of each sample through the random forest model, thereby calculating the collapse risk prediction coefficient with a formula of C=P(y=1|x), where x represents the input variable, and y represents a binary variable, i.e., whether a collapse event has occurred, y=1 indicates that a collapse has occurred, and y=0 indicates that no collapse has occurred.

Images