Rock damage and fracture precursor warning method based on infrared thermal image texture characteristics

Abstract

The present invention discloses a rock damage and fracture precursor warning method based on infrared thermal image texture features, and relates to the field of monitoring and early warning technology for fracture and instability of coal rock mass in mines. The influence of time background noise on the infrared radiation temperature of the bearing rock is eliminated by referring to the rock, and the spatial background noise of the infrared radiation temperature matrix of the bearing rock is eliminated by using the Ostu threshold segmentation and adaptive median filtering method. The instantaneous damage infrared temperature response matrix of the bearing rock is calculated, and the infrared radiation information response characteristics caused by the crack development of the bearing rock are effectively extracted; the grayscale co-occurrence matrix texture eigenvalues ​​of the differential damage infrared temperature response matrix are calculated, and the energy texture feature deviation value is calculated to determine the lower and upper thresholds of the rock damage and fracture precursors. It can accurately and effectively identify the initial precursors and critical precursors of rock damage and destruction, has good engineering application value, and is of great significance to the prevention and control of rock engineering and natural geological disasters.

Description

Technical Field

[0001] The invention relates to the technical field of monitoring and early warning of cracking and instability of coal rock masses in mines, and in particular to a rock damage and cracking precursor early warning method based on texture features of infrared thermal images. Background Art

[0002] Under the action of external loads, the internal microcracks of coal and rock will gradually initiate, nucleate, and expand to form macro cracks, which will then cause rock failure and instability. The damage and destruction of rock materials is a thermodynamic process, that is, the process of energy transfer and transformation between rocks and the outside world, such as elastic energy, damage energy, thermal energy, acoustic energy, etc. Among them, the dislocation proliferation and heat dissipation generated at the crack tip during crack propagation, the friction heat generated by internal microcracks and particles, and the heat energy released by friction between fragments during local damage will cause changes in the infrared radiation information on the rock surface. Therefore, the infrared radiation information on the rock surface is closely related to the evolution process of its internal microcracks, which makes it possible to use infrared radiation to identify the precursors of rock failure. Compared with other methods, infrared radiation monitoring technology has the advantages of real-time, non-destructive and non-contact, and the operation method is simple. It is an ideal means to identify the precursors of rock failure and instability.

[0003] To this end, the existing technologies include average rock infrared radiation temperature (AIRT), variance of differential infrared radiation (VSMIT), high temperature point proportional factor (HTPSF), infrared radiation b value (b IR -value), as well as infrared radiation characteristic parameters such as the critical slowing-down autocorrelation coefficient, are used to identify the precursors of load-bearing rock failure. However, since the surface infrared radiation temperature caused by microcracks inside the rock is relatively weak, the above characteristic parameters contain a lot of interference information in the calculation process. Therefore, they are only sensitive to the changes in infrared information caused by macro cracks on the rock surface, and the timeliness of precursor identification for brittle rocks is difficult to guarantee. In addition, the existing rock damage early warning methods often rely on manual judgment of the precursor moment, and there is still a lack of a unified and effective precursor threshold, which is prone to missed judgments and misjudgments. Therefore, it is urgent to develop an accurate and effective method for dynamic identification of infrared radiation of rock damage precursors. Summary of the invention

[0004] In view of the shortcomings in the prior art, a rock damage and fracture precursor warning method based on infrared thermal image texture features is provided, which is sensitive to the precursor response existing in the rock damage and fracture infrared radiation warning method and can perform real-time dynamic identification.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0006] A rock damage and fracture precursor warning method based on infrared thermal image texture features comprises the following steps:

[0007] S1: Infrared radiation time background noise correction during the damage and destruction process of bearing rock: Set a reference rock next to the bearing rock, use an infrared thermal imager to simultaneously detect the infrared radiation data of the bearing rock and the reference rock, use median filtering to eliminate the spatial noise of the infrared radiation data of the bearing rock and the reference rock, and obtain the original infrared temperature matrix of the bearing rock and the original infrared temperature matrix of the reference rock arranged in time series;

[0008] The time background noise of the infrared radiation information of the rock is removed by using the average infrared radiation temperature of the reference rock to obtain the infrared temperature matrix of the rock after the correction of the time background noise at all times; the average infrared radiation temperature is the temperature mean of each frame in the monitoring process;

[0009] S2: Subtract the infrared temperature matrix of the bearing rock at the initial moment from the infrared temperature matrix of the bearing rock at all moments, and take the absolute value to obtain the infrared temperature increment matrix of the bearing rock;

[0010] S3: Using the threshold segmentation method, the infrared radiation response information increment of the bearing rock caused by damage and fracture is separated from the infrared temperature increment matrix of the bearing rock to obtain the infrared temperature increment matrix of the bearing rock damage;

[0011] S4: Eliminate the residual spatial background noise of the infrared temperature increment matrix of the bearing rock damage, and calculate the damage infrared temperature response matrix;

[0012] S5: Perform a forward first-order difference operation on the infrared temperature response matrix of the bearing rock damage to obtain a differential damage infrared temperature response matrix. Then, the differential damage infrared temperature response matrix is ​​segmented with 0°C as the threshold to obtain an instantaneous damage infrared temperature response matrix.

[0013] S6: Calculate the energy texture eigenvalue of the instantaneous damage infrared temperature response matrix of the coal rock through the gray level co-occurrence matrix of the instantaneous damage infrared temperature response matrix;

[0014] S7: cumulatively summing the energy texture characteristic values ​​of the bearing rock to obtain a cumulative energy texture characteristic value, and then performing a difference operation between the cumulative energy texture characteristic value and time to calculate a cumulative energy texture characteristic deviation value of the bearing rock;

[0015] S8: Derivative the cumulative energy texture characteristic deviation value of the bearing rock to obtain the energy texture characteristic deviation value, and by comparing the energy texture characteristic deviation in the rock damage and fracture process with the precursor threshold, the moment when the upper limit of the precursor threshold is exceeded for the first time is judged as the precursor of rock failure, thereby realizing the identification of the bearing rock damage and fracture precursor.

[0016] Furthermore, the method for correcting the temporal background noise of the infrared radiation information of the rock in step S1 is as follows:

[0017] Calculate the average infrared radiation temperature of the reference rock using the following formula:

[0018]

[0019] Where AIRT r OITM is the average infrared radiation temperature of the rock. r is the element of the i-th row and j-th column of the original infrared temperature matrix of the reference rock, m and n are the number of rows and columns of the matrix respectively, and t is any time; the original infrared temperature matrix of the bearing rock is corrected for time background noise using the average infrared radiation temperature of the reference rock to obtain the infrared temperature matrix of the bearing rock after time background noise correction, which is recorded as the infrared temperature matrix of the bearing rock DITM s , the formula is as follows:

[0020] DITM(t)=OITM s (t)-AIRT r (t)

[0021] Where DITM is the infrared temperature matrix of the bearing rock after correction of the temporal background noise.

[0022] Furthermore, the calculation formula of the infrared temperature increment matrix of the bearing rock in step S2 is:

[0023] ΔDITM(t)=|DITM(t)-DITM(t 0 )|

[0024] Where ΔDITM is the infrared temperature increment matrix of the bearing rock, t 0 (t) is the initial moment.

[0025] Furthermore, the calculation method of the damage infrared temperature increment matrix in step S3 is: the Otsu method is used to calculate the segmentation threshold of the DITM damage infrared response information at each moment, and the threshold is used to perform threshold segmentation on the infrared temperature matrix of the bearing rock to obtain the infrared radiation response information increment of the bearing rock caused by the damage. The calculation formula is as follows:

[0026]

[0027] Where DTCM is the damage infrared temperature increment matrix, and Th is the segmentation threshold of the damage infrared response information.

[0028] Taking a step further, the method for eliminating the residual spatial background noise of the infrared temperature increment matrix of the bearing rock damage in step S4 is: using an adaptive median filter to filter the infrared temperature increment matrix of the bearing rock damage to remove the residual spatial background noise. The formula is as follows:

[0029] ΔDTCM(t)=admedian(DTCM(t),9)

[0030] Where ΔDTCM is the damage infrared temperature increment matrix after adaptive median filtering, admedian is the adaptive median filter, and “9” is the maximum window value of the adaptive median filter.

[0031] Furthermore, the calculation method of the damage infrared temperature response matrix in step S4 is: taking 0°C as the threshold, binarizing the damage infrared temperature increment matrix after the adaptive median filter, and multiplying it with the damage infrared temperature increment matrix to obtain the damage infrared temperature response matrix, denoted as DIRTM, and the formula is as follows:

[0032]

[0033] IRDM(t)=BDTCM(t)×DTCM(t)

[0034] Where BDTCM is the damage infrared temperature increment matrix after binarization; IRDM is the bearing rock damage infrared temperature response matrix.

[0035] Further, the calculation formula of the instantaneous damage infrared temperature response matrix in step S5 is as follows:

[0036] ΔIRDM(t)=IRDM(t)-IRDM(t-1 / f)

[0037]

[0038] Where ΔIRDM is the differential damage infrared temperature response matrix, f is the monitoring frequency, and IIRDM is the instantaneous damage infrared temperature response matrix.

[0039] Further, the calculation method of the texture feature of the instantaneous damage infrared temperature response matrix in step S6 is: calculating the grayscale co-occurrence matrix of the instantaneous damage infrared temperature response matrix, and extracting the energy texture feature value of the instantaneous damage infrared temperature response matrix of the coal rock, the formula is as follows:

[0040] E IIRDM (t)=∑p(h,k,t) 2

[0041] In the formula, E IIRDM Gray-level co-occurrence matrix energy texture eigenvalue, p is the gray-level co-occurrence matrix of the differential damage infrared temperature response matrix, h and k are the elements of the hth row and kth column of the gray-level co-occurrence matrix.

[0042] Furthermore, in step S7, the calculation method of the cumulative energy texture characteristic deviation value of the bearing rock is: performing a difference operation on the cumulative energy texture characteristic value and its corresponding time to obtain the cumulative energy texture characteristic deviation value, and the formula is as follows:

[0043]

[0044] Where DCE IIRDM (t) is the cumulative energy texture feature deviation value.

[0045] Furthermore, based on artificial experience, 0.01 is set as the lower threshold of the precursor, and 0.1 is set as the upper threshold of the precursor. The moment when the lower threshold is exceeded for the first time is the precursor of initial rock failure, and the moment when the upper threshold is exceeded for the first time is the precursor of critical rock failure.

[0046] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0047] 1) The present invention successively uses the average infrared radiation temperature of the reference sample, Otsu threshold segmentation, and adaptive median filtering methods to eliminate the influence of time and space interference factors on the rock infrared radiation temperature matrix, and obtains the differential damage infrared temperature response matrix, which effectively extracts the infrared radiation response information caused by the evolution of rock cracks and solves the problem of key infrared radiation information being covered. 2) The present invention uses the energy texture feature deviation value of the instantaneous damage infrared temperature response matrix as a characteristic parameter for identifying rock damage precursors. Compared with other characteristic parameters, the energy texture feature deviation value can accurately evaluate the rock crack evolution process, has a good sensitivity to the response of rock fracture precursors, and can effectively avoid missed judgments and misjudgments. 3) This method determines the lower threshold value (0.01) and the upper threshold value (0.1) of rock loss and fracture precursors. The threshold value has good universality and can accurately and effectively identify the initial and critical precursors of rock failure. For the initial precursors, methods such as strengthening support and rock reinforcement can be used to avoid disasters. For the critical precursors, the staff and equipment should be evacuated immediately to reduce casualties and economic losses. It has good engineering application value and is of great significance to the prevention and control of rock engineering and natural geological disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flow chart of the rock damage and fracture precursor warning method based on the texture characteristics of infrared thermal images of the present invention;

[0049] Figure 2 The energy texture characteristic value and stress variation curve of the load-bearing rock in the embodiment of the present invention;

[0050] Figure 3 The cumulative energy texture characteristic value and stress variation curve of the load-bearing rock in the embodiment of the present invention;

[0051] Figure 4 It is the curve of cumulative energy texture characteristic deviation value and stress variation with time of the load-bearing rock in the embodiment of the present invention. DETAILED DESCRIPTION

[0052] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings:

[0053] like Figure 1 As shown, an infrared radiation identification method for the starting point of coal rock crack initiation and stable expansion includes the following steps:

[0054] S1: Infrared radiation time background noise correction during the damage and failure process of the bearing rock: A reference rock is set 10 cm away from the bearing rock at the same level, and an infrared thermal imager is used to detect the infrared radiation data of the bearing and reference rocks. The median filter is used to eliminate the spatial noise of the infrared radiation data to obtain the original infrared temperature matrix of the bearing rock (OITM s ) and the reference rock original infrared temperature matrix (OITM r ). After that, the average infrared radiation temperature of the reference rock is calculated using the following formula:

[0055]

[0056] Where AIRT r OITM is the average infrared radiation temperature of the rock. r is the element in the i-th row and j-th column of the original infrared temperature matrix of the reference rock, m and n are the number of rows and columns of the matrix respectively, and t is the time. The original infrared temperature matrix of the bearing rock is corrected for the time background noise using the average infrared radiation temperature of the reference rock to obtain the infrared temperature matrix of the bearing rock after the time background noise correction, which is recorded as the bearing rock infrared temperature matrix (DITM), and the formula is as follows:

[0057] DITM(t)=OITM s (t)-AIRT r (t)

[0058] Where DITM is the infrared temperature matrix of the bearing rock after correction of the temporal background noise.

[0059] S2: Calculate the infrared temperature increment matrix of the bearing rock: subtract the DITM at the initial moment from the DITM at each moment, and take the absolute value to obtain the infrared temperature increment matrix of the bearing rock. The formula is as follows:

[0060] ΔDITM(t)=|DITM(t)-DITM(t 0 )|

[0061] Where ΔDITM is the infrared temperature increment matrix of the bearing rock, t 0 is the initial moment.

[0062] S3: Separation of infrared radiation response information of bearing rock damage: The Otsu method is used to calculate the segmentation threshold of the DITM damage infrared response information at each moment, and the threshold is used to perform threshold segmentation on the bearing rock infrared temperature matrix to obtain the incremental infrared radiation response information of the bearing rock caused by the damage. The calculation formula is as follows

[0063]

[0064] Where DTCM is the damage infrared temperature increment matrix, and Th is the segmentation threshold of the damage infrared response information.

[0065] S4: Calculate the infrared temperature response matrix of the bearing rock damage: Use the adaptive median filter to filter the bearing rock damage infrared temperature increment matrix to remove the residual spatial background noise. Then, use 0℃ as the threshold, binarize the filtered temperature matrix, and multiply it with the damage infrared temperature increment matrix to obtain the damage infrared temperature response matrix, denoted as DIRTM, the formula is as follows:

[0066] ΔDTCM(t)=admedian(DTCM(t),9)

[0067]

[0068] IRDM(t)=BDTCM(t)×DTCM(t)

[0069] Wherein, ΔDTCM is the damage infrared temperature increment matrix after adaptive median filtering, admedian is the adaptive median filter, and “9” is the maximum window value of the adaptive median filter; BDTCM is the damage infrared temperature increment matrix after binarization processing; IRDM is the bearing rock damage infrared temperature response matrix.

[0070] S5: Calculate the instantaneous damage infrared temperature response matrix: Perform a forward first-order difference operation on the bearing rock damage infrared temperature response matrix to obtain the differential damage infrared temperature response matrix. Then, use 0°C as the threshold to segment the differential damage infrared temperature response matrix to obtain the instantaneous damage infrared temperature response matrix. The formula is as follows:

[0071] ΔIRDM(t)=IRDM(t)-IRDM(t-1)

[0072]

[0073] Where ΔIRDM is the differential damage infrared temperature response matrix; IIRDM is the instantaneous damage infrared temperature response matrix.

[0074] S6: Calculate the texture features of the infrared temperature response matrix of the instantaneous damage of the bearing coal rock: Calculate the grayscale co-occurrence matrix of the instantaneous damage infrared temperature response matrix and extract its energy texture feature value. The formula is as follows:

[0075] E IIRDM (t)=∑p(h,k,t) 2

[0076] In the formula, E IIRDM The gray-level co-occurrence matrix energy texture eigenvalue, p is the gray-level co-occurrence matrix of the instantaneous damage infrared temperature response matrix, and h and k are the elements of the h-row and k-column of the gray-level co-occurrence matrix. Figure 2 As shown in the figure, with the increase of rock stress, microcracks inside the rock gradually develop. The heat energy released by the crack development causes the infrared radiation temperature field on the rock surface to change from uniform to non-uniform. Its energy texture characteristic value starts to decrease from horizontal, and the reduction rate increases sharply as the cracks expand unstably.

[0077] S7: Calculate the cumulative energy texture characteristic deviation value of the bearing rock: perform cumulative summation of the energy texture of the bearing rock to obtain the cumulative energy texture characteristic value; perform a difference operation between the cumulative energy texture characteristic value and its corresponding time to obtain the cumulative energy texture characteristic deviation value, and the formula is as follows:

[0078]

[0079] Where DCE IIRDM (t) is the cumulative energy texture feature deviation value. Figure 3 As shown in Figure 2, during initial loading, the cumulative energy texture increases nearly linearly with time (y=x), but gradually deviates from the reference line when the crack begins to expand. Figure 4 As shown in the figure, with the unstable crack expansion, the cumulative energy texture feature deviation value begins to decrease exponentially, which provides a basis for the identification of rock failure precursors.

[0080] S8: Precursor warning of damage and fracture of bearing rock: Derivative operation is performed on the accumulated energy texture characteristic deviation value to obtain the energy texture characteristic deviation value. Afterwards, 0.01 is used as the precursor lower threshold and 0.1 is used as the precursor upper threshold to identify the moment when the energy texture characteristic deviation value exceeds the threshold. Among them, the moment when the lower threshold is exceeded for the first time is the precursor of initial rock damage, and the moment when the upper threshold is exceeded for the first time is the precursor of critical rock damage. Figure 4 As shown, the initial precursor time and critical precursor time of the rock sample identification are 405s and 543s, respectively, which are 243s and 125s ahead of the failure time, respectively.

Claims

1. A rock damage and fracture precursor warning method based on infrared thermal image texture features, characterized in that: The steps include: S1: Infrared radiation time background noise correction during the damage and destruction process of bearing rock: Set a reference rock next to the bearing rock, use an infrared thermal imager to simultaneously detect the infrared radiation data of the bearing rock and the reference rock, use median filtering to eliminate the spatial noise of the infrared radiation data of the bearing rock and the reference rock, and obtain the original infrared temperature matrix of the bearing rock and the original infrared temperature matrix of the reference rock arranged in time series; The time background noise of the infrared radiation information of the rock is removed by using the average infrared radiation temperature of the reference rock to obtain the infrared temperature matrix of the rock after the correction of the time background noise at all times; the average infrared radiation temperature is the temperature mean of each frame in the monitoring process; S2: Subtract the infrared temperature matrix of the bearing rock at the initial moment from the infrared temperature matrix of the bearing rock at all moments, and take the absolute value to obtain the infrared temperature increment matrix of the bearing rock; S3: Using the threshold segmentation method, the infrared radiation response information increment of the bearing rock caused by damage and fracture is separated from the infrared temperature increment matrix of the bearing rock to obtain the infrared temperature increment matrix of the bearing rock damage; S4: Eliminate the residual spatial background noise of the infrared temperature increment matrix of the bearing rock damage, and calculate the damage infrared temperature response matrix; S5: Perform a forward first-order difference operation on the infrared temperature response matrix of the bearing rock damage to obtain a differential damage infrared temperature response matrix. Then, the differential damage infrared temperature response matrix is ​​segmented with 0°C as the threshold to obtain an instantaneous damage infrared temperature response matrix. S6: Calculate the energy texture eigenvalue of the instantaneous damage infrared temperature response matrix of the coal rock through the gray level co-occurrence matrix of the instantaneous damage infrared temperature response matrix; S7: cumulatively summing the energy texture characteristic values ​​of the bearing rock to obtain a cumulative energy texture characteristic value, and then performing a difference operation between the cumulative energy texture characteristic value and time to calculate a cumulative energy texture characteristic deviation value of the bearing rock; S8: Derivative the cumulative energy texture characteristic deviation value of the bearing rock to obtain the energy texture characteristic deviation value, and by comparing the energy texture characteristic deviation in the rock damage and fracture process with the precursor threshold, the moment when the upper limit of the precursor threshold is exceeded for the first time is judged as the precursor of rock failure, thereby realizing the identification of the bearing rock damage and fracture precursor.

2. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: The method for correcting the temporal background noise of the infrared radiation information of rocks in step S1 is as follows: Calculate the average infrared radiation temperature of the reference rock using the following formula: Where AIRT r OITM is the average infrared radiation temperature of the rock. r is the element of the i-th row and j-th column of the original infrared temperature matrix of the reference rock, m and n are the number of rows and columns of the matrix respectively, and t is any time; the original infrared temperature matrix of the bearing rock is corrected for time background noise using the average infrared radiation temperature of the reference rock to obtain the infrared temperature matrix of the bearing rock after time background noise correction, and the formula is as follows: DITM(t)=OITM s (t)-ART r (t) Where DITM is the infrared temperature matrix of the bearing rock after correction of the temporal background noise.

3. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 2 is characterized in that: The calculation formula of the infrared temperature increment matrix of the bearing rock in step S2 is: ΔDITM(t)=|DITM(t)-DITM(t0)| Where DITM is the infrared temperature matrix of the bearing rock after correction of the temporal background noise, ΔDITM is the infrared temperature increment matrix of the bearing rock, t0 is the initial time, and t is any time.

4. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 3 is characterized in that: The calculation method of the damage infrared temperature increment matrix in step S3 is: the Otsu method is used to calculate the segmentation threshold of the DITM damage infrared response information at each moment, and the threshold is used to perform threshold segmentation on the infrared temperature matrix of the bearing rock to obtain the infrared radiation response information increment of the bearing rock caused by the damage. The calculation formula is as follows: Where DTCM is the damage infrared temperature increment matrix, Th is the segmentation threshold of the damage infrared response information, and t is any time.

5. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: The method for eliminating the residual spatial background noise of the infrared temperature increment matrix of the bearing rock damage in step S4 is: using an adaptive median filter to filter the infrared temperature increment matrix of the bearing rock damage to remove the residual spatial background noise. The formula is as follows: ΔDTCM(t)=admedian(DTCM(t),9) Where DTCM is the damage infrared temperature increment matrix, ΔDTCM is the damage infrared temperature increment matrix after adaptive median filtering, admedian is the adaptive median filter, "9" is the maximum window value of the adaptive median filter, and t is any time.

6. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: The calculation method of the damage infrared temperature response matrix in step S4 is: taking 0°C as the threshold, binarizing the damage infrared temperature increment matrix after the adaptive median filter, and multiplying it with the damage infrared temperature increment matrix to obtain the damage infrared temperature response matrix, which is recorded as IRDM. The formula is as follows: IRDM(t)=BDTCM(t)×DTCM(t) Where DTCM is the damage infrared temperature increment matrix, ΔDTCM is the damage infrared temperature increment matrix after adaptive median filtering, t is any time, BDTCM is the damage infrared temperature increment matrix after binarization processing; IRDM is the bearing rock damage infrared temperature response matrix.

7. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: The calculation formula of the instantaneous damage infrared temperature response matrix in step S5 is as follows: ΔIRDM(t)=IRDM(t)-IRDM(t-1 / f) Where ΔIRDM is the differential damage infrared temperature response matrix, f is the monitoring frequency, IIRDM is the instantaneous damage infrared temperature response matrix, IRDM is the bearing rock damage infrared temperature response matrix, and t is any time.

8. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: Step S6: The method for calculating the texture features of the instantaneous damage infrared temperature response matrix is ​​as follows: calculating the grayscale co-occurrence matrix of the instantaneous damage infrared temperature response matrix, and extracting the energy texture feature value of the instantaneous damage infrared temperature response matrix of the coal rock, the formula is as follows: E IIRDM (t)=∑p(h,k,t) 2 In the formula, E IIRDM Gray-level co-occurrence matrix energy texture eigenvalue, p is the gray-level co-occurrence matrix of the differential damage infrared temperature response matrix, h and k are the elements of the hth row and kth column of the gray-level co-occurrence matrix.

9. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 1 is characterized in that: Step S7: The calculation method of the cumulative energy texture characteristic deviation value of the bearing rock is: performing a difference operation between the cumulative energy texture characteristic value and its corresponding time to obtain the cumulative energy texture characteristic deviation value, and the formula is as follows: In the formula, E IIRDM Gray-level co-occurrence matrix energy texture eigenvalue, DCE IIRDM (t) is the cumulative energy texture feature deviation value, and t is any time.

10. The rock damage and fracture precursor warning method based on infrared thermal image texture features according to claim 9 is characterized by: Through experience, 0.01 is set as the lower threshold of precursor, and 0.1 is set as the upper threshold of precursor. The moment when the lower threshold is exceeded for the first time is the precursor of initial rock failure, and the moment when the upper threshold is exceeded for the first time is the precursor of critical rock failure.

Images