Method for evaluating surface deterioration degree for repairing stone cultural relics based on thermal infrared imaging

Abstract

The application discloses a surface deterioration degree evaluation method for stone cultural relics repair based on thermal infrared imaging and relates to the field of image recognition.The method comprises the following steps: controlled thermal excitation is carried out on the surface of a stone cultural relic, and time-series thermal infrared imaging is collected, and true thermal distribution data are obtained after correction; the temperature rise amplitude, heat dissipation compliance degree, heat dissipation lag feature and sudden drop heat dissipation feature are extracted, the surface constraint rupture degree is constructed, and expansion and contraction driving information is obtained; the deterioration driving distribution result is generated in combination with the spatial proximity relationship; the stability degree distribution map is formed and continuously constrained; and the repair priority order is determined according to the stability degree distribution map and the cultural sensitivity, and the deterioration grade result is generated. Through radiation correction and multi-feature extraction on the time-series thermal infrared data under the controlled thermal excitation, the surface constraint rupture degree, the expansion and contraction driving information and the stability degree distribution map are constructed, the repair priority order is determined in combination with the importance of the key positions of the cultural relics, and the grading evaluation of the surface deterioration state of the stone cultural relics and the repair sorting are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image recognition, in particular to a surface deterioration degree evaluation method for stone cultural relics repair based on thermal infrared imaging. BACKGROUND

[0002] Stone cultural relics are prone to surface weathering, salt precipitation bulging, flaky peeling, and structural bulging and other forms of deterioration under long-term natural exposure and complex environmental effects. The formation mechanism is influenced by factors such as environmental temperature and humidity changes, thermal expansion and contraction cycles, salt crystallization expansion, and microbial erosion, which gradually weakens the physical constraints between the surface and the internal matrix of the cultural relics, resulting in the expansion of microcracks, interlayer hollowing, and local structural detachment. Such deterioration processes are usually highly concealed, slow in evolution, and uneven in spatial distribution, making early identification and repair decision-making difficult.

[0003] Existing stone cultural relic deterioration detection methods mostly use contact or semi-contact methods, including ultrasonic detection, knocking echo method, moisture sensitivity test, and optical microscopic imaging. These methods can reflect local physical changes or surface weathering, but have limitations in detection efficiency, spatial coverage, and time resolution, making it difficult to achieve large-area, non-contact, real-time dynamic evaluation of the surface of cultural relics. In addition, traditional image detection methods are mainly based on visible light or near-infrared imaging, which are significantly affected by lighting conditions, surface color and texture, making it difficult to accurately reveal the thermal structure and bonding state of the surface of cultural relics.

[0004] Thermal infrared imaging technology can reflect the thermal conductivity and internal defect distribution of materials through non-contact detection of surface thermal radiation, and has unique advantages in structural diagnosis, material non-destructive testing, and cultural relic protection. However, traditional thermal infrared detection methods mostly remain at the qualitative visualization stage, relying only on temperature field images to observe thermal anomaly areas, lacking quantitative feature extraction and time series behavior analysis. Especially in the context of stone cultural relics, due to factors such as surface roughness, environmental radiation background, local shielding, and temperature drift, single-frame temperature images are difficult to accurately depict thermal diffusion characteristics, resulting in insufficient stability of thermal anomaly recognition results.

[0005] Therefore, it is necessary to propose a surface deterioration degree evaluation method for stone cultural relics repair based on thermal infrared imaging, which can correct radiation, geometry, and multi-parameter features of time-series thermal images under controlled thermal excitation, construct surface constraint rupture degree and expansion and contraction driving information, generate deterioration driving distribution results combined with spatial proximity, and on this basis, establish stability distribution and graded deterioration level results, thereby realizing quantitative evaluation of the deterioration state of the surface of cultural relics and scientific determination of the repair priority order. SUMMARY

[0006] Based on the above-mentioned defects of the prior art, the purpose of the present application is to provide a surface layer degradation degree evaluation method for stone cultural relics restoration based on thermal infrared imaging to solve the above technical problems.

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solutions: a surface layer degradation degree evaluation method for stone cultural relics restoration based on thermal infrared imaging, comprising:

[0008] Setting a detection area on the surface layer of the stone cultural relics, applying a controlled thermal excitation, collecting a continuous thermal infrared imaging sequence in the heating stage and the heat dissipation stage, performing radiation correction and geometric correction on the thermal infrared imaging sequence to form time-series thermal image data consistent with the real thermal distribution of the surface layer of the cultural relics;

[0009] Based on the time-series thermal image data, the temperature rise amplitude, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop heat dissipation feature are extracted position by position to generate local degradation symptom information and construct a surface layer constraint rupture degree;

[0010] In the detection area, the pre-response stage and the post-response stage of the same position are compared and analyzed, and the long-term residual temperature rise, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop heat dissipation feature are associated according to the surface layer constraint rupture degree to obtain expansion and contraction driving information;

[0011] In the detection area, the surface layer constraint rupture degree, the expansion and contraction driving information and the corresponding information of the adjacent position are compared and analyzed based on the spatial proximity relationship to generate a degradation driving distribution result;

[0012] Based on the degradation driving distribution result, a stability representation information is formed, and the stability representation information is subjected to a spatial continuity constraint to obtain a stability distribution map that is continuously distributed in the detection area;

[0013] The stability distribution map is graded, and the importance of the cultural relic decoration details, the inscription information and the key positions of the shape are determined to determine the repair priority and generate a graded degradation level result.

[0014] The present application is further provided, and the formation of the time-series thermal image data consistent with the real thermal distribution of the surface layer of the cultural relics comprises:

[0015] The detection area is demarcated on the surface layer of the stone cultural relics, the controlled thermal excitation is applied in the detection area, the detection area experiences the heat injection and release process in the heating stage and the heat dissipation stage, and the thermal infrared imaging sequence at different time points is collected;

[0016] The radiation brightness signal of the thermal infrared imaging sequence is inverted to obtain apparent temperature information corresponding to time;

[0017] According to the radiation characteristics of the stone cultural relic surface layer material and the environmental radiation background, the apparent temperature information is subjected to radiation correction to generate equivalent surface layer temperature data;

[0018] A geometric mapping relationship of the cultural relic surface layer is established by a structured light scanning mode, the equivalent surface layer temperature data is projected to a reference plane coordinate system to form time-series thermal image data consistent with the actual surface layer geometric structure.

[0019] The application further provides that the feature extraction based on the time-series thermal image data and the surface layer constraint rupture degree generation include:

[0020] The baseline moment and the peak moment of the heating stage are determined for each position in the detection area, and the temperature difference between the peak moment and the baseline moment is used to extract the temperature rise amplitude;

[0021] A preset time interval is selected in the heat dissipation stage, and the heat dissipation compliance degree is determined according to the temperature change rate of the time interval;

[0022] The temperature maintenance ratio of the sustained high temperature stage is selected in the post-peak time period to obtain the heat dissipation lag feature;

[0023] The temperature rapid drop time period is identified in the late heat dissipation stage, and the sudden drop heat dissipation feature is calculated according to the temperature drop intensity of the time period;

[0024] The temperature rise amplitude, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop heat dissipation feature are associated to form the local deterioration symptom information representing the bonding state of the cultural relic surface layer and the matrix.

[0025] The application further provides that the surface layer constraint rupture degree construction includes:

[0026] The temperature rise amplitude and the heat dissipation lag feature in the local deterioration symptom information are extracted to establish the surface layer heat retention relationship, and the heat dissipation compliance degree is set as a constraint parameter;

[0027] The sudden drop heat dissipation feature is amplified by weighting to describe the instantaneous release feature in the heat dissipation stage;

[0028] The product of the temperature rise amplitude and the heat dissipation lag feature is taken as the surface layer heat response cumulative amount, the heat dissipation compliance degree is combined to form the release constraint amount, and the sudden drop heat dissipation feature is combined to form the surface layer constraint rupture degree, representing the constraint relaxation state between the surface layer and the matrix in the detection area.

[0029] The application further provides that the expansion and contraction driving information acquisition includes:

[0030] The time intervals of the early response stage and the late response stage are determined according to the inflection point of the temperature change rate curve after the thermal excitation is completed, the early response stage corresponds to the temperature rapid decay interval, and the late response stage corresponds to the temperature slow decay interval;

[0031] Extracting surface temperature information corresponding to the time in the early response stage and the late response stage respectively, and calculating residual temperature rise according to the difference between the late temperature and the baseline temperature;

[0032] Establishing time period retention rate based on residual temperature rise and temperature rise amplitude, and forming time period consistency retention rate combined with heat dissipation delay characteristics;

[0033] Nonlinearly combining the sudden heat dissipation characteristics and the heat dissipation compliance degree to obtain local release intensity;

[0034] According to the surface constraint rupture degree, the time period consistency retention rate and the local release intensity, weighted combination is carried out to form the expansion and contraction driving information, which represents the time sequence behavior characteristics of the detection position in terms of salting enrichment, bulging expansion and contraction, drum separation or sheet peeling.

[0035] The application further provides that the deterioration driving distribution result generated in the detection area in a spatially adjacent relationship constraint comprises:

[0036] A spatially adjacent set of position points and adjacent position points is established in a geometrically corrected reference coordinate domain, and the adjacent relationship is determined according to the distance threshold between the position points;

[0037] The surface constraint rupture degree and the expansion and contraction driving information of each position point are spatially associated, and the difference ratio of the position point relative to the adjacent set in the surface constraint rupture degree and the expansion and contraction driving information is calculated;

[0038] The difference ratio is weighted and accumulated based on the adjacent relationship weight index to form a difference contrast quantity between positions;

[0039] The difference contrast quantity is combined with the surface constraint rupture degree and the expansion and contraction driving information of the position point to form a deterioration driving distribution result, which represents the spatial difference state of the constraint relaxation trend and the expansion and contraction driving intensity of each position in the detection area.

[0040] The application further provides that the formation of the stability representation information comprises:

[0041] The deterioration driving distribution result is subjected to a normalization process to generate a driving normalization quantity, and a single-point vulnerability quantity is obtained by inversion according to the driving normalization quantity;

[0042] The single-point vulnerability quantity is accumulated in a ratio under the spatially adjacent relationship to construct a neighborhood coupling balance term;

[0043] A stability initial value is formed based on the single-point vulnerability quantity and the neighborhood coupling balance term, the stability initial value and the stability initial value of the adjacent position point are executed to form a neighborhood coordination operator;

[0044] An iterative convergence is implemented under the combination of the stability initial value and the neighborhood coordination operator to generate an iterative stability.

[0045] Corresponding the iterative stability to each position of the reference coordinate domain, a stability representation information continuously distributed in the detection area is obtained.

[0046] The application further provides that the stability representation information is subjected to a spatial continuity constraint, and the stability distribution map is obtained by:

[0047] Difference constraints are performed on the iterative stability of adjacent position points in the detection area, and a coordination index is determined according to the distance between the position points and the surface connectivity;

[0048] The iterative stability of each position point is weighted and smoothed by taking the coordination index as a weighting factor, and a continuous stability field is constructed;

[0049] In the continuous stability field, a convergence judgment is performed according to the variation gradient of the iterative stability, and when the stability difference between adjacent position points is lower than a preset threshold, the spatial continuity state is determined;

[0050] The stability field meeting the continuity state is mapped to the reference coordinate domain of the detection area to form a stability distribution map.

[0051] The application further provides that the stability distribution map is graded, and the grading includes:

[0052] The stability value of each position in the stability distribution map is read, and the interval ranges of a high-risk section, a medium-grade section, a low-grade section and a stable section are determined according to a preset stability grading threshold group;

[0053] The stability value of each position is compared with the grading threshold to establish a stability section index;

[0054] Based on the stability section index, the continuous positions of the same section in the detection area are aggregated to form a stability grading partition;

[0055] The formed stability grading partition is mapped with the geometric coordinates of the cultural relic surface to generate a stability section map.

[0056] The application further provides that the repair priority order is determined by combining the importance of the cultural relic decoration details, the inscription information and the shape key positions, and the combining includes:

[0057] A cultural sensitivity weight is assigned to each position on the stability section map, and a cultural sensitivity matrix is established according to the cultural relic decoration distribution, the inscription density and the structure feature location information;

[0058] The cultural sensitivity matrix and the stability section map are superimposed position by position to calculate a comprehensive grading index;

[0059] According to the correspondence relationship between the comprehensive classification index and the preset grade threshold value, the grade assignment is performed, and a deterioration grade indication quantity is generated;

[0060] According to the descending order relationship between the deterioration grade indication quantity and the cultural sensitivity weight, a position sequence is arranged, a repair priority order is determined, and a graded deterioration grade result is recorded.

[0061] The application provides a surface deterioration degree evaluation method for stone cultural relics repair based on thermal infrared imaging, which sets a detection area on the surface of the stone cultural relics, applies a controlled thermal excitation, collects a continuous thermal infrared imaging sequence in the heating stage and the heat dissipation stage, performs radiation correction and geometric correction on the thermal infrared imaging sequence to form time-series thermal image data consistent with the real thermal distribution of the surface of the cultural relics; based on the time-series thermal image data, the temperature rise amplitude, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop heat dissipation feature are extracted position by position to generate local deterioration symptom information and construct the surface constraint rupture degree; the early response stage and the late response stage of the same position in the detection area are compared and analyzed, the long-period residual temperature rise, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop heat dissipation feature are associated according to the surface constraint rupture degree, and the expansion and contraction driving information is obtained; in the detection area, the surface constraint rupture degree, the expansion and contraction driving information and the corresponding information of the adjacent position of each position are compared and compared in the space adjacent relationship, and the deterioration driving distribution result is generated; based on the deterioration driving distribution result, the stability representation information is formed, the spatial continuity constraint is implemented on the stability representation information, and the stability distribution map in the continuous distribution in the detection area is obtained; the stability distribution map is graded, the importance of the key positions of the cultural relic decoration details, the inscription information and the shape is determined, the repair priority order is determined, the graded deterioration grade result is generated, and the beneficial effects include:

[0062] 1. Non-contact dynamic thermal response analysis: by applying a controlled thermal excitation to the surface of the stone cultural relics and collecting a continuous thermal infrared imaging sequence, the dynamic process in the heating stage and the heat dissipation stage is analyzed in time series, and the real thermal distribution of the surface of the cultural relics is restored by combining radiation correction and geometric correction, which can reflect the internal thermal conduction characteristics and structure combination state without contacting the surface of the cultural relics, providing basic data support for early hidden deterioration identification;

[0063] 2. Quantitative characterization of deterioration mechanism of multi-parameter coupling: by extracting and associating multi-dimensional thermal parameters such as temperature rise amplitude, heat dissipation compliance degree, heat dissipation lag feature and sudden drop heat dissipation feature, the surface constraint rupture degree and the expansion and contraction driving information are constructed, and the deterioration driving distribution result is calculated combined with the space adjacent relationship, which can quantitatively characterize the constraint relaxation, bulging expansion and contraction and peeling tendency of the surface of the cultural relics from the time and space double scales, overcoming the limitations of traditional methods relying on only a single temperature difference or heat dissipation rate;

[0064] 3. Cultural sensitivity driven repair priority assessment: On the basis of stability distribution map, the importance of cultural sensitive details, inscription information and key positions of shape and size is introduced, the cultural sensitivity weight matrix is established and superimposed with the stability index to generate a comprehensive classification index, and the repair priority order is determined according to the comprehensive index, which can output the degradation level results on the basis of the trade-off between technical degradation and cultural value, and provide quantifiable and traceable decision basis for cultural relic repair process.

[0065] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the following specific embodiments of the present application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application will be described. BRIEF DESCRIPTION OF DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description only represent some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor. In the drawings:

[0067] Figure 1 The flow chart of the surface degradation degree evaluation method for stone cultural relics repair based on thermal infrared imaging shown in an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0068] The embodiments of the present application will be described below with reference to the drawings and preferred embodiments, and those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in the specification. The present application can also be implemented or applied by different specific embodiments, and the details in the specification can be modified or changed based on different views and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustration of the present application, and are not intended to limit the protection scope of the present application.

[0069] It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present application in a schematic manner, and only show the components related to the present application in the diagrams, not the number, shape and size of the components when actually implemented. The actual implementation of each component may be a random change in shape, number and proportion, and the layout pattern of the components may also be more complex.

[0070] In the following description, numerous specific details are discussed in order to provide a thorough understanding of embodiments of the present application. However, those skilled in the art will recognize that the embodiments of the present application can be practiced without these specific details. In other instances, well-known structures and devices are not described in detail in order to avoid obscuring embodiments of the present application.

[0071] Embodiment one:

[0072] A surface layer degradation degree evaluation method for stone cultural relics repair based on thermal infrared imaging, as shown in Figure 1 includes:

[0073] A detection area is set for the surface layer of the stone cultural relic, a controlled thermal excitation is applied, a continuous thermal infrared imaging sequence of the heating stage and the heat dissipation stage is collected, radiation correction and geometric correction are performed on the thermal infrared imaging sequence, and time-series thermal image data consistent with the real thermal distribution of the surface layer of the cultural relic is formed;

[0074] Based on the time-series thermal image data, the temperature rise amplitude, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop type heat dissipation feature are extracted position by position, the local degradation symptom information is generated, and the surface layer constraint rupture degree is constructed;

[0075] In the detection area, the early response stage and the late response stage of the same position are compared and analyzed, the long-period residual temperature rise, the heat dissipation compliance degree, the heat dissipation lag feature and the sudden drop type heat dissipation feature are associated according to the surface layer constraint rupture degree, and the expansion and contraction driving information is obtained;

[0076] In the detection area, the surface layer constraint rupture degree, the expansion and contraction driving information and the corresponding information of the adjacent position are compared and analyzed in a space proximity relationship, and the degradation driving distribution result is generated;

[0077] Based on the degradation driving distribution result, the stability representation information is formed, the spatial continuity constraint is implemented on the stability representation information, and the stability distribution map continuously distributed in the detection area is obtained;

[0078] The stability distribution map is graded, the importance of the cultural relic decoration details, the inscription information and the shape key position is determined, the repair priority order is determined, and the graded degradation level result is generated.

[0079] The present application further provides that the time-series thermal image data consistent with the real thermal distribution of the surface layer of the cultural relic includes:

[0080] The detection area is divided on the surface of the stone cultural relics, and controlled thermal excitation is implemented in the detection area. The detection area experiences heat injection and release processes in the heating stage and the heat dissipation stage, and thermal infrared imaging sequences at different time points are collected. Specifically, a rectangular detection area is selected on the front surface of the stone cultural relics to be detected, and the detection area covers the positions where fine bumps, shallow layer of rust and salt deposition exist on the surface of the cultural relics. Controlled thermal excitation is applied to the detection area. The controlled thermal excitation is a low-power area array radiation source. The constant starting radiation intensity is maintained for a preset heating time in a linear increasing manner, so that the detection area experiences the heating process in the heating stage, and enters the heat dissipation stage after the radiation source is turned off. Thermal infrared imaging is collected at a fixed frame rate in the heating stage and the heat dissipation stage, respectively, to obtain a plurality of continuous thermal infrared imaging sequences. Each frame corresponds to a specific time stamp, so as to establish the thermal response process of the same area in the early heating stage, the late heating stage, the early heat dissipation stage and the late heat dissipation stage.

[0081] The radiation brightness signal of the thermal infrared imaging sequence is inverted to obtain the apparent temperature information corresponding to the time. Specifically, the original detection response signal of the thermal infrared imaging sequence is read, and the radiation brightness signal at each position pixel is inverted through the imaging device radiation calibration curve completed in advance to obtain the apparent temperature information corresponding to the time of the pixel. The apparent temperature information still contains the detector body radiation, environmental reflection component and viewing angle coupling factor, and has not yet corresponded to the real thermal state of the surface of the cultural relics.

[0082] According to the radiation characteristics of the surface material of the stone cultural relics and the environmental radiation background, the apparent temperature information is subjected to radiation correction to generate equivalent surface temperature data. Specifically, a control small area with known radiation characteristics is arranged in the detection area, and the apparent temperature information at different times is subjected to position-related radiation correction processing in combination with the emissivity records of the stone body material to be detected, the surface weathering layer and the crystalline salt deposition layer. In the radiation correction process, the environmental reflection component of the low-emissivity area is deducted, and the diffuse radiation component of the high-roughness area is retained, so that each pixel obtains equivalent surface temperature data at the same time. The equivalent surface temperature data is a temperature sequence position by time, which can represent the real thermal distribution evolution of the surface of the cultural relics in the heating and heat dissipation process.

[0083] The geometric mapping relationship of the cultural relic surface layer is established by a structured light scanning mode, equivalent surface layer temperature data is projected to a reference plane coordinate system, time sequence thermal image data consistent with the actual surface layer geometric structure is formed, and specifically, the cultural relic surface layer is measured with high precision in three dimensions by the structured light scanning mode, three-dimensional surface coordinate information corresponding to the detection area is obtained, and the geometric mapping relationship of the cultural relic surface layer is established. The equivalent surface layer temperature data is projected to a reference plane coordinate system according to the geometric mapping relationship, and the equivalent surface layer temperature data collected at different time points is position-aligned in the same reference coordinate system, so that the same physical position has one-to-one temperature record at different time frames. Through the above steps, time sequence thermal image data consistent with the actual surface layer geometric structure is obtained. The time sequence thermal image data depicts the equivalent surface layer temperature distribution of the detection area at each time, continuously covers the heating stage and the heat dissipation stage in the time dimension, and maintains a corresponding relationship with the real surface layer position of the cultural relic in the spatial dimension.

[0084] The application is further provided that the feature extraction based on the time sequence thermal image data and the surface layer constraint rupture degree generation include:

[0085] The baseline time and the peak time of the heating stage are determined for each position in the detection area, and the temperature difference between the peak time and the baseline time is used to extract the temperature rise amplitude; specifically, the baseline time and the peak time of the heating stage are determined for discrete position points in the detection area divided on the cultural relic surface layer. The baseline time is the initial time before the start of the controlled thermal excitation, and records the initial temperature of the position. The peak time is the time when the position reaches the highest temperature in the heating stage, and records the peak temperature of the position. The temperature rise amplitude is obtained by the difference between the peak temperature and the initial temperature, and the corresponding temperature rise amplitude distribution is given for all positions in the detection area. The temperature rise amplitude reflects the heat absorption degree and temperature rise amplitude of the position to the thermal excitation, and the area showing the tendency of bulging, lifting and shell separation usually records a significant high value;

[0086] A preset time interval is selected in the heat dissipation stage, and the heat dissipation compliance degree is determined according to the temperature change rate of the time interval; specifically, after the controlled thermal excitation stops, a preset time interval is selected for each position in the heat dissipation stage, and the temperature reduction rate of the time interval with time is taken as the heat dissipation compliance degree. A fixed time length at the beginning of the heat dissipation stage is selected as an analysis window, and the temperature reduction rate in the window is calculated. The position with a higher reduction rate is considered as a position where heat can be released smoothly, and the position with a lower reduction rate is considered as a position where heat is retained, and the heat dissipation compliance degree represents whether the position shows a state of smooth heat discharge or heat retention;

[0087] In the post-peak time period, the temperature maintaining ratio of the high-temperature stage is selected to obtain the heat dissipation delay characteristic; specifically, after the peak moment, the temperature curve of each position is continuously tracked, a high-temperature period is selected, and the maintaining ratio of the high-temperature period is calculated, that is, the temperature rise relative to the baseline moment is maintained to what extent. When a position maintains a temperature level obviously higher than the baseline for a long time after the heat source stops, and the temperature drop process is obviously delayed, the heat dissipation delay characteristic of the position is recorded as a high value. The characteristic reflects whether there is a near-surface isolation layer or loose layer in the position, which forms a heat-retaining shell structure in the local area, thereby delaying heat dissipation in time;

[0088] In the late heat dissipation stage, the temperature rapid drop time period is identified, and the sudden drop heat dissipation characteristic is calculated according to the temperature drop strength of the time period; specifically, in the late stage of heat dissipation, whether a significant temperature rapid drop interval occurs along the time sequence for each position is checked, that is, a stage in which a significant drop is formed in a short duration. The temperature drop strength in the interval is calculated and recorded as the sudden drop heat dissipation characteristic. If the position appears a short-time rapid drop in the late stage of heat dissipation, it usually corresponds to the release of local heat through newly formed or significantly opened release path, which usually corresponds to the physical state of surface-matrix interface cracking, drum boundary formation, and sheet layer separation;

[0089] The temperature rise amplitude, heat dissipation compliance degree, heat dissipation delay characteristic and sudden drop heat dissipation characteristic are associated to form local deterioration sign information representing the combination state of the surface layer and the matrix of cultural relics, and specifically, the temperature rise amplitude, heat dissipation compliance degree, heat dissipation delay characteristic and sudden drop heat dissipation characteristic are aggregated at the same position to obtain the local deterioration sign information of the position, and the four characteristic values are recorded for each position, and the joint performance of the four characteristic values is taken as the local deterioration sign description of the position. When a position simultaneously shows high temperature rise amplitude, low heat dissipation compliance degree, high heat dissipation delay characteristic and sudden drop heat dissipation characteristic, the position is identified as having a significant abnormal trend of surface layer structure. The local deterioration sign information represents whether the constraint relationship between the surface layer and the internal matrix of cultural relics tends to be relaxed, drummed up or separated.

[0090] The application further provides that the construction of the surface layer constraint rupture degree comprises:

[0091] The temperature rise amplitude and heat dissipation lag feature in the extracted local deterioration sign information are used to establish a surface layer heat retention relationship, and the heat dissipation compliance degree is set as a constraint parameter. Specifically, in the detection area of the stone cultural relic surface layer, for any position, firstly, the temperature rise amplitude and heat dissipation lag feature of the position are associated. The temperature rise amplitude represents the rising amplitude of the position from the initial temperature to the peak temperature during the controlled thermal excitation, and the heat dissipation lag feature represents whether the position remains in a high temperature state close to the peak value for a long time after the heat source is turned off. The temperature rise amplitude and heat dissipation lag feature of the position are respectively set to a quantitative scale, and after corresponding to each other, a surface layer heat retention relationship value is obtained. The surface layer heat retention relationship value reflects whether the position has a heat accumulation behavior similar to a bulging thin layer or a shell layer, that is, the position not only rises significantly after being heated, but also maintains a high temperature during the initial temperature drop. When a certain position has a large temperature rise amplitude and maintains a temperature higher than the initial temperature for a long time during the initial heat dissipation, the surface layer heat retention relationship value of the position is marked as high. When the temperature rise amplitude of a certain position is general or the temperature drops rapidly after heat dissipation, the surface layer heat retention relationship value of the position is marked as low. At the same time, the heat dissipation compliance degree is introduced as a constraint parameter in the association process. The heat dissipation compliance degree represents whether the temperature of the position drops smoothly. If the temperature of a certain position drops significantly slowly during the heat dissipation stage, it is considered that the heat dissipation compliance degree is low. For the position with low heat dissipation compliance degree, the surface layer heat retention relationship value is not weakened during calculation. For the position with high heat dissipation compliance degree, the surface layer heat retention relationship value is inhibited during calculation, so that even if the position has a high temperature rise in a short time, it is not incorrectly identified as a significant heat retention area. Through processing, a surface layer heat retention relationship result of each position can be obtained, reflecting the heat absorption intensity and retention persistence of the position, and whether the position has a significant heat dissipation obstacle has been considered.

[0092] The sudden drop heat dissipation feature is implemented to amplify the instantaneous release feature of the heat dissipation phase by weighting. Specifically, in the later stage of the heat dissipation phase, the temperature time series of each position is continuously tracked to identify whether there is a significant rapid cooling section. If the temperature suddenly drops from a higher temperature to a lower temperature in a short time, the sudden drop heat dissipation feature of the position is considered to be significant. This phenomenon is usually related to the appearance of a rapid release channel at the bonding interface between the surface layer and the substrate, such as the sudden opening of the drum-off boundary, the formation of a heat dissipation gap at the edge of the sheet-shaped peeling, etc. The sudden drop heat dissipation feature is set as a separate risk amplification factor. In processing, it is not simply averaged with other features, but is retained as an independent multiplicative enhancement marker. That is, when a position exhibits significant sudden drop heat dissipation behavior, the position is assigned a higher instantaneous release marker; when no such sudden drop behavior is observed, the position is assigned a lower or near-zero instantaneous release marker. The significance of this weighted amplification is that even if a position is identified as having a heat retention relationship in the first step, if no significant instantaneous release occurs, the position is considered to mainly exhibit slow bulging heat retention. Once the sudden drop heat dissipation behavior occurs, it will be explicitly marked as having potential cracking or interface rapid release behavior.

[0093] The product of the temperature rise amplitude and the heat dissipation hysteresis characteristic is taken as a surface layer thermal response cumulative amount, and the heat dissipation compliance is combined to form a release constraint amount, which is combined with the sudden drop heat dissipation characteristic to form a surface layer constraint rupture degree, which represents the constraint relaxation state between the surface layer and the base in the detection area. Specifically, for each position, the temperature rise amplitude and the heat dissipation hysteresis characteristic are taken together as the surface layer thermal response cumulative amount. The surface layer thermal response cumulative amount represents whether the position has experienced a large temperature rise and maintained the rise, which is manifested as the position forming a shallow thermal shell with a certain independence after being heated. After obtaining the surface layer thermal response cumulative amount, the heat dissipation compliance is introduced to form the release constraint amount. The release constraint amount describes whether the heat dissipation of the position is inhibited in the heat dissipation stage. If the heat dissipation compliance is low, that is, the cooling process is slow, the release constraint amount is given a higher weight, indicating that the position has a blockage in the cooling process; if the heat dissipation compliance is high, that is, the cooling is smooth, the release constraint amount is given a lower weight, indicating that the position does not show significant limited heat dissipation. The surface layer thermal response cumulative amount and the release constraint amount are jointly considered as the overall thermal response intensity of the position in the whole heating-heat dissipation period, that is, whether the position has a thermal behavior performance that can be regarded as a loose shell. By nonlinearly superimposing the surface layer thermal response cumulative amount, the release constraint amount and the sudden drop heat dissipation characteristic, the surface layer constraint rupture degree of each position is obtained. If a position simultaneously satisfies the following conditions: having a significant surface layer thermal retention relationship result, which is manifested as a significant release constraint amount, and having a high level of instantaneous release marker, the surface layer constraint rupture degree of the position is marked as high; if only a slight thermal retention relationship is shown and no obvious instantaneous release marker is shown, the surface layer constraint rupture degree of the position is marked as low; if it is between the two, it is marked as an intermediate level. The surface layer constraint rupture degree is given in units of position, indicating whether the surface layer of the position has no longer responded to thermal excitation as a whole with the base, but shows a behavior that can be regarded as an independent layer, a bulging layer or an early detached layer.

[0094] The application further provides that the obtaining the expansion and contraction driving information comprises:

[0095] The time interval of the early response stage and the late response stage is determined according to the inflection point of the temperature change rate curve after the end of the thermal excitation, the early response stage corresponds to the temperature rapid decay interval, and the late response stage corresponds to the temperature slow decay interval; specifically, after completing the controlled thermal excitation, the temperature decay curve with time is extracted from the time sequence thermal image data of the region. For each detection position, the change rate of the temperature after the peak value starts to decrease is determined, and the inflection point time when the decrease rate changes from obviously fast to obviously slow is found. The time interval before the inflection point is marked as the early response stage, and the time interval after the inflection point is marked as the late response stage. The early response stage corresponds to the temperature rapid decay interval, which is manifested as the position rapidly cooling after the heat source is removed; the late response stage corresponds to the temperature slow decay interval, which is manifested as the position entering a slow falling state after experiencing the initial rapid cooling. Through the division, the early behavior and the late behavior of the same detection position in the heat dissipation process are distinguished;

[0096] The surface temperature information corresponding to the time is extracted in the early response stage and the late response stage, and the residual temperature rise is calculated according to the difference between the late temperature and the baseline temperature; specifically, after completing the time interval division, the surface temperature information corresponding to the representative time of the same detection position is read from the early response stage and the late response stage respectively. At the same time, the baseline temperature information of the detection position before the start of the thermal excitation is obtained. The temperature value of the late response stage is compared with the baseline temperature, the difference between the two is obtained, and is recorded as the residual temperature rise of the detection position. The residual temperature rise represents the temperature rise degree that the position still maintains in the late stage after experiencing the initial rapid cooling. If a position is still significantly higher than the baseline temperature in the late response stage, it can be determined that the position continuously retains the heating residue for a long period of time and does not completely fall back to the initial state;

[0097] The time period retention rate is established based on the residual temperature rise and the temperature rise amplitude, and the time period consistency retention rate is formed in combination with the heat dissipation delay characteristic; specifically, in order to depict whether the detection position maintains a significant heat delay behavior in a long period of time, first, the residual temperature rise is compared with the temperature rise amplitude obtained in the heating stage to obtain a retention ratio, which measures the proportion of the temperature rise that the position still maintains in the late stage to the initial temperature rise. The retention ratio reflects whether the position behaves in a state of continuous bulging and heat retention. The retention ratio is combined with the heat dissipation delay characteristic to obtain the time period consistency retention rate of the position. The time period consistency retention rate indicates whether the position maintains the characteristics of slow descent in the form of bulging and shell in a long observation period, rather than only showing occasional heat delay in a short period of time. In comparison, if a position is still significantly hot in the late response stage and does not significantly cool down immediately after the peak value, the time period consistency retention rate of the position is recorded as high; if the position initially has a large temperature rise, but quickly falls back in the initial heat dissipation period, the time period consistency retention rate of the position is recorded as low;

[0098] The local release intensity is obtained by nonlinearly combining the sudden drop heat dissipation feature and the heat dissipation compliance degree; specifically, in the later stage of heat dissipation, for the same detection position, it is checked whether there is a sharp cooling behavior in a short time, that is, there is a significant and prominent temperature drop section between adjacent sampling time points, which corresponds to the sudden drop heat dissipation feature. At the same time, the sudden drop heat dissipation feature is adjusted in combination with the heat dissipation compliance degree, and the local release intensity of the position is obtained by nonlinearly combining the sudden drop heat dissipation feature and the heat dissipation compliance degree. The local release intensity unifies the description of the instantaneous rapid cooling in the later stage and the overall difficult heat dissipation in the early stage. When a certain position continuously remains hot in the first half of the heat dissipation but suddenly releases rapidly in the later stage, the local release intensity of the position is recorded as high; when a certain position presents smooth cooling from beginning to end without significant instantaneous release, the local release intensity of the position is recorded as low;

[0099] According to the surface constraint rupture degree, the time period consistency retention rate and the local release intensity, the expansion and contraction driving information is formed, which represents the time sequence behavior characteristics of the detection position in aspects of salting-out enrichment, bulging expansion and contraction, drum separation or sheet peeling, and specifically, the surface constraint rupture degree, the time period consistency retention rate and the local release intensity are weighted and combined to obtain the expansion and contraction driving information of the position, and the logic of the weighted combination is that: when a certain detection position has obvious surface constraint rupture degree, high time period consistency retention rate and high local release intensity at the same time, the high value of the expansion and contraction driving information is recorded; when only weak surface constraint rupture degree is possessed, or although temperature rise is presented, but no continuous heat retention and no obvious instantaneous release behavior are presented, the low value of the expansion and contraction driving information is recorded; if between the two cases, the intermediate level is recorded. The expansion and contraction driving information is generated in sequence in units of detection positions, and represents the time sequence behavior characteristics of the detection position in aspects of salting-out enrichment, bulging expansion and contraction, drum separation boundary formation and sheet peeling initiation.

[0100] The application further provides that the generating the deterioration driving distribution result in the detection area under the constraint of spatial proximity relationship comprises:

[0101] The spatial neighbor set of the position point and the adjacent position point is established in the geometrically corrected reference coordinate field, and the neighbor relationship is determined according to the distance threshold between the position points; specifically, the detection area of the cultural relic surface layer is geometrically corrected to obtain the corresponding coordinate distribution of the area in the reference coordinate system. The reference coordinate system is taken as a unified positioning plane, and the detection area is discretized into a series of position points. For each position point, a number of spatial adjacent position points are selected from around the position point according to a preset distance threshold, forming a spatial neighbor set of the position point, and the distance threshold is set as a fixed distance range matched with the local structure scale of the cultural relic, and all position points within a limited distance from the position point are selected as the spatial neighbor set of the position point. The set is used to represent the directly connected region of the position point on the surface morphology, thereby limiting the comparison range within the region having actual structure continuity, without crossing the obvious morphological fracture, peeling boundary or contour edge;

[0102] The surface layer constraint fracture degree and the swelling and shrinking driving information of each position point are spatially associated, and the difference ratio of the position point relative to the neighbor set in the surface layer constraint fracture degree and the swelling and shrinking driving information is calculated; specifically, for each position point in the detection area, the surface layer constraint fracture degree and the swelling and shrinking driving information of the position point are read respectively. At the same time, the surface layer constraint fracture degree and the swelling and shrinking driving information of each adjacent position point in the spatial neighbor set of the position point are read. For the current position point, the adjacent position points are compared one by one, and two types of difference ratios are calculated: the first type of difference ratio measures the deviation degree of the position point in the surface layer constraint fracture degree relative to the adjacent position points. If the surface layer constraint fracture degree of the current position point is significantly higher than that of the adjacent position points, the difference ratio is recorded as high; if it is basically consistent or lower than that of the adjacent position points, the difference ratio is recorded as medium or low. The second type of difference ratio measures the deviation degree of the position point in the swelling and shrinking driving information. If the swelling and shrinking driving information of the current position point is significantly higher than that of the adjacent position points, the difference ratio is recorded as high; if the difference is not obvious, the difference ratio is recorded as low. Through the above process, a group of relative difference ratios can be obtained on each position point, which indicates whether the position point shows an abnormally prominent constraint relaxation trend and swelling and shrinking driving feature in the local region;

[0103] The difference ratio is weighted and accumulated based on the proximity weight index to form a position difference contrast quantity; specifically, after obtaining the difference ratio for each adjacent position point, the difference ratios are weighted and accumulated according to the importance of the adjacent relationship to obtain the difference contrast quantity of the position point, and higher weight is given to position points that are closer in space, and lower weight is given to position points that have obvious geometric transitions, discontinuous surface materials or sudden changes in carving levels. The weighted results are accumulated to obtain the difference contrast quantity of the position point. The difference contrast quantity indicates whether the position point behaves as a significant abnormal point in its local neighborhood. If the position point exhibits higher surface constraint rupture degree and stronger swelling and shrinking driving information compared to almost all adjacent position points, the difference contrast quantity of the position point is recorded as high; otherwise, if the behavior of the position point is consistent with the overall neighborhood, the difference contrast quantity is recorded as low;

[0104] The difference contrast quantity is combined with the surface constraint rupture degree and the swelling and shrinking driving information of the position point to form a deterioration driving distribution result, which represents the spatial difference state of each position in the detection area in terms of constraint relaxation trend and swelling and shrinking driving intensity. Specifically, after obtaining the difference contrast quantity, the difference contrast quantity is combined with the surface constraint rupture degree and the swelling and shrinking driving information of the position point to generate a deterioration driving distribution result on the position point. The combination process follows the following logic: if the surface constraint rupture degree of a position point itself is high, it indicates that the surface and the matrix have shown a significant combination relaxation trend; if the swelling and shrinking driving information of the position point itself is high, it indicates that there are signs of bulging, bulging, salt precipitation driving swelling or sheet peeling starting to form; if the difference contrast quantity of the position point is also high, it indicates that the above-mentioned state not only exists in the point itself, but also has local outstanding discreteness relative to the surrounding area, rather than being uniformly distributed in the whole area. When the above three conditions are met at the same time, the deterioration driving distribution result of the position point is marked as high; when only one or two of them are met, it is marked as intermediate; when none of the three is significant, it is marked as low. In this way, a distribution image type data result is generated in the reference coordinate domain of the detection area. The result is defined as a deterioration driving distribution result, which represents the spatial difference state of each position in the detection area in terms of constraint relaxation trend and swelling and shrinking driving intensity.

[0105] The present application further provides that the formation of stability representation information includes:

[0106] The deterioration driving distribution result is normalized to generate a driving normalization quantity, and a single-point vulnerability quantity is obtained by inversion according to the driving normalization quantity; specifically, for each position point of the detection area, the deterioration driving distribution result of the position point is read. In order to compare the deterioration driving distribution results of different position points on the same scale, the result is normalized. The deterioration driving distribution result of each position point is compressed to be within a limited interval, so as to avoid individual extreme high values dominating the overall judgment in subsequent processing. The quantity obtained by the compression is the driving normalization quantity. Based on the driving normalization quantity, the single-point vulnerability quantity of the position point is calculated. The single-point vulnerability quantity can be understood as a reverse representation of the driving normalization quantity: the higher the driving normalization quantity of a certain position point, the lower the single-point vulnerability quantity of the position point; when the driving normalization quantity is low, the single-point vulnerability quantity is high, and the single-point vulnerability quantity separates the strong driving region from the weak driving region: the strong driving region often corresponds to a potential unstable point, and the weak driving region tends to be a more stable region. At this time, the single-point vulnerability quantity provides an initial local scale index for subsequent stability calculation;

[0107] The single-point vulnerability quantity is accumulated by ratio under the spatial neighborhood relationship to construct a neighborhood coupling balance term; specifically, after obtaining the single-point vulnerability quantity of each position point, the position point and its adjacent position points are compared by using the spatial neighborhood relationship. For each position point, the single-point vulnerability quantity proportional relationship between the position point and its adjacent position points is investigated one by one, and the proportions are accumulated according to the importance of the spatial neighborhood relationship to form a neighborhood coupling balance term. For regions that are geometrically adjacent, continuous ornamentation regions, and regions within the same stone bedding plane, a higher accumulation weight is given; for regions that cross the carving boundary line, the fracture contour or the material quality mutation, a lower accumulation weight is given. The neighborhood coupling balance term indicates that if the single-point vulnerability quantity of a certain position point is consistent with most of the surrounding position points, the position point is in a coordinated state in the neighborhood coupling balance term; if a certain position point is significantly different from the surrounding position points, the position point is in an unbalanced state in the neighborhood coupling balance term. The stability judgment is no longer dependent on a single point;

[0108] The stability initial value is obtained based on the single-point vulnerability quantity and the neighborhood coupling balance term, and the stability initial value is coordinated and aggregated with the stability initial values of neighboring position points to form a neighborhood coordination operator; specifically, after the single-point vulnerability quantity and the neighborhood coupling balance term are obtained, the two are combined for each position point to obtain the stability initial value. If the single-point vulnerability quantity of a position point is high and the neighborhood coupling balance term shows that the position point is not significantly different from the surrounding area, the stability initial value of the position point is recorded as a high level; if the single-point vulnerability quantity of a position point is low, or although there is a certain vulnerability, but it is shown in the neighborhood coupling balance term that it is out of balance, that is, it is significantly different from the surrounding area, the stability initial value of the position point is recorded as a low level. The stability initial value is still an uncorrected stability index at this stage, and has not yet reflected the global coupling effect between position points, but only embodies the single-point and direct neighborhood relationship; after the stability initial values of the position points are obtained, the stability initial values of neighboring position points of each position point are further read. The stability initial values of these neighboring position points are coordinated and aggregated to form the neighborhood coordination operator of the position point. The neighborhood coordination operator represents the local collective stability tendency of the small range area where the position point is located. If the stability initial values of the neighboring position points are generally high, the neighborhood coordination operator of the position point is given a high level; if the stability initial values of the neighboring position points are generally low, the neighborhood coordination operator of the position point is given a low level.

[0109] The iteration convergence is implemented under the combination of the stability initial value and the neighborhood coordination operator to generate an iterative stability; specifically, the stability initial value and the corresponding neighborhood coordination operator are jointly processed to obtain a new stability value, and the value is taken as the iterative stability. Based on the stability initial value, the iterative stability is modified according to the neighborhood coordination operator: in the area where the neighborhood coordination operator is high, the iterative stability is improved; in the area where the neighborhood coordination operator is low, the iterative stability is inhibited. The obtained iterative stability is brought back again, and the same joint processing can be repeatedly performed until the iterative stabilities between adjacent position points no longer change significantly. The repeated process realizes the transmission of stability in space, so that the stabilities in the same continuous area gradually tend to be consistent, while avoiding that a single abnormal point is excessively enlarged in the final result;

[0110] The iterative stability is corresponded to each position in the reference coordinate domain to obtain stability representation information that is continuously distributed in the detection area; specifically, when the iteration process reaches convergence, the stability value of each position point after the final iteration is recorded. The final stability value is the stability representation information of the position point. At this time, the stability representation information has integrated three factors: the performance of the position point itself in the degradation driving; the consistency relationship between the position point and its direct neighborhood; and the overall coordination level of the small range area where the position point is located; the stability representation information is no longer a thermal response feature at a certain moment, but a stability evaluation index under the spatial consistency constraint.

[0111] The application is further configured to implement spatial continuity constraint on the stability representation information, and obtain the stability distribution map including:

[0112] Difference constraint is performed on the iterative stabilities of adjacent position points in the detection area, and a coordination index is determined according to the distance between the position points and the surface connectivity. Specifically, the detection area is discretized into a series of position points in the reference coordinate system after geometric correction. For each position point, the iterative stability of the position point is read, and the iterative stabilities of adjacent position points having spatial proximity with the position point are read. The difference between the iterative stabilities of the position point and each adjacent position point is compared in turn, and the geometric distance between the two points and whether the two points are located in the same continuous surface area are considered at the same time. If the two points are close to each other and are located in the same continuous stone surface, the difference between the two points is considered to be strictly constrained; if there are obvious engraved grooves, surface fracture seams, buried damage boundaries or material transition zones between the two points, the difference between the two points is not strictly constrained. The above comparison process obtains a coordination index for each adjacent point pair. The coordination index indicates whether the two points have a continuous relationship that should maintain a close stability level: the higher the continuity of the area, the higher the coordination index; the geometric fracture and material discontinuity are assigned a lower coordination index.

[0113] The iterative stabilities of the position points are weighted and smoothed using the coordination index as a weighting factor to construct a continuous stability field. Specifically, after obtaining the coordination index, the iterative stability of each position point is weighted and smoothed. The processing order of the weighted smoothing is as follows: for a certain position point, the iterative stabilities of its adjacent position points are collected; for each adjacent position point, the aforementioned coordination index is used as a weighting factor to assign the stability influence strength of the position point; the weighted stability quantity and the iterative stability of the position point itself are integrated together to obtain the smoothed stability value of the position point. The smoothing process indicates that if the iterative stability of a certain position point is too different from the surrounding adjacent position points, and these adjacent position points have high surface connectivity with the position point, the smoothed stability value of the position point will be pulled to the neighborhood level; if there is a clear geometric boundary between the position point and the surrounding area, the value of the position point will not be forcibly pulled, thereby avoiding false smoothing across the real physical boundary. After the smoothing process is performed on all position points, a continuous stability field covering the entire detection area is obtained. The continuous stability field no longer presents a large number of isolated sharp jump points, but forms a gradually changing stability distribution on the same actual connected surface;

[0114] The convergence judgment is performed according to the variation gradient of the iterative stability in the continuous stability field, and the spatial continuity state is determined when the stability difference between adjacent position points is lower than a preset threshold; specifically, after the continuous stability field is obtained, spatial consistency judgment is performed. The judgment is based on the stability difference between adjacent position points, and each pair of spatial adjacent position points is checked, and it is confirmed whether the stability difference of the two is lower than a preset threshold. When the difference between a position point and a plurality of adjacent position points is less than the threshold, it is indicated that the position point is in the spatial continuity state, that is, there is no significant inconsistency between the stability level of the position point and its neighborhood, and the smoothing of the region is sufficient. When there is still a local region with a significant difference exceeding the threshold, the weighted smoothing process can be repeated on the local region, and the un-converged region is modified again, until the stability difference between adjacent points is within the threshold range at most position points. Through the difference constraint and convergence judgment process, a continuous stability field in space can be obtained, and non-real abrupt changes caused by single-frame noise, individual abnormal measuring points or local temperature fluctuations are avoided;

[0115] The stability field meeting the continuity state is mapped to the reference coordinate domain of the detection region to form a stability distribution map, specifically, after the convergence judgment is completed, the final continuous stability field is mapped back to the reference coordinate domain point by point to restore the spatial position relationship of the stability on the surface of the cultural relic. The reference coordinate domain and the real surface geometry of the cultural relic are in one-to-one correspondence, which is established by geometric correction in the previous step. Therefore, the final stability value can be given to each specific surface position of the detection region. The mapping process forms a stability distribution map. The stability distribution map expresses the stability value of each position point in a spatially continuous form within the detection region, and obtains a stability classification base map.

[0116] The application further provides that the classification of the stability distribution map comprises:

[0117] The stability value of each position in the stability distribution map is read, and the interval ranges of the high-risk section, the medium-grade section, the low-grade section, and the stable section are determined according to the preset stability grading threshold group. Specifically, the stability value of the stability distribution map is read position by position. The stability value is a spatially continuous stability representation obtained in the previous step, representing the holding state of the position in the current detection period. A stability grading threshold group is set. The threshold group is given by the previous records in the cultural relic protection field, the historical detection archives of similar material samples, and the on-site expert evaluation, and is divided into four grade intervals, corresponding to the high-risk section, the medium-grade section, the low-grade section, and the stable section. The high-risk section corresponds to the lowest stability range, and is a characteristic region with potential drumming, bulging, flaking, salt-induced swelling, and other significant risk signs. The medium-grade section corresponds to the transition range between the high-risk section and the low-grade section. The low-grade section corresponds to the region of slight degradation or local stress concentration. The stable section corresponds to the region that does not show significant abnormalities. By setting the threshold group, any position point in the stability distribution map can be classified into a specific grade interval in the subsequent step;

[0118] The stability value of each position is compared with the grading threshold to establish a stability section index. Specifically, after the threshold group is set, the stability value of each position point in the stability distribution map is compared with the threshold group one by one, and a stability section index is assigned to the position point. If the stability value of a position point is lower than the upper threshold of the high-risk section, the position point is marked with a high-risk section index. When the stability value is between the demarcation threshold of the high-risk section and the medium-grade section, it is marked as a medium-grade section index. When the stability value is between the demarcation threshold of the medium-grade section and the low-grade section, it is marked as a low-grade section index. When the stability value is higher than the lower threshold of the stable section, it is marked as a stable section index. Through this assignment process, the stability distribution map is converted from a continuous quantity field to four types of discrete classification markers, forming a set of stability section indices for each position point;

[0119] Based on the stability section index, the continuous positions in the same section within the detection area are aggregated to form a stability grading partition. Specifically, after obtaining the stability section index of each position point, spatial aggregation processing is performed on the cultural relic detection area. It is detected whether there are several adjacent position points in the reference coordinate domain that are marked with the same section index at the same time. If so, these adjacent position points are considered as a stability grading partition;

[0120] The stability grading subarea is mapped with the geometric coordinates of the surface layer of the cultural relic to generate a stability section map. Specifically, after obtaining each subarea, the spatial range of each stability grading subarea is geometrically positioned, and the spatial range is remapped to the actual geometric coordinates of the surface layer of the cultural relic. The correspondence between the reference coordinate domain of the cultural relic detection area and the real surface of the cultural relic has been established in the geometric correction step in the preamble, so each stability grading subarea can be directly labeled back to the surface of the cultural relic to obtain a stability section map covering the actual surface of the cultural relic.

[0121] The application further provides that the determination of the repair priority order according to the importance of the details of the cultural relic decoration, the engraved information and the key positions of the shape includes:

[0122] A cultural sensitivity weight is assigned to each position on the stability section map, and a cultural sensitivity matrix is established according to the decoration distribution, the engraved density and the structure feature position information of the cultural relic. Specifically, for the stability section map of the surface layer of the cultural relic, the distribution of cultural elements is identified subarea by subarea, including the position of the decoration details, the position of the engraved content and the key position of the structure shape. A cultural sensitivity weight is assigned to each specific spatial position to represent the importance of the position in information carrying. The weight is assigned to each position point in the detection area to form a cultural sensitivity matrix. The matrix maintains the same coordinate correspondence with the stability section map in space, i.e. each position point has both a stability section mark and a cultural sensitivity weight;

[0123] The cultural sensitivity matrix and the stability section map are superimposed position by position to calculate a comprehensive grading index. Specifically, for each position point, two pieces of information are considered: the stability section level of the stability subarea where the position point is located, and the cultural sensitivity weight of the position point. A high weight indicates that the position carries irreplaceable engraved, symbolic decoration or key shape features. By superimposing the above two factors, a comprehensive grading index is calculated for each position point. This comprehensive grading index reflects double sensitivity: not only the physical stability risk, but also the cultural information value. Even if a position is only in the medium section in terms of stability, if the position carries rare engraved, the comprehensive grading index of the position can be higher than that of the surrounding area. Conversely, even if a region is a high-risk section, if it is only a repetitive background decoration and the cultural sensitivity weight is low, the comprehensive grading index of the region can not be promoted to the highest level.

[0124] According to the correspondence between the comprehensive classification index and the preset grade threshold value, the grade assignment is performed to generate the deterioration grade indication quantity; specifically, after obtaining the comprehensive classification index, a grade threshold value group is set for the region. The comprehensive classification index is divided into several intervals, and each interval corresponds to an explicit deterioration grade indication quantity mark. According to the following rules, the grade assignment is performed: the comprehensive classification index in the highest interval corresponds to the highest grade indication quantity, which is marked as the grade of urgent priority treatment; the comprehensive classification index in the second interval corresponds to the second grade indication quantity, which is marked as the grade of key tracking; the comprehensive classification index in the middle interval corresponds to the general grade indication quantity, which is marked as the grade of routine maintenance; and the comprehensive classification index in the lowest interval corresponds to the low grade indication quantity, which is marked as the grade of routine maintenance. Through the assignment process, each position point is assigned a deterioration grade indication quantity. The deterioration grade indication quantity takes into account both cultural attributes and structural state, rather than only ranking according to thermal stability or only ranking according to visual decorative value;

[0125] According to the descending order relationship between the deterioration grade indication quantity and the cultural sensitivity weight, the position sequence is arranged to determine the repair priority order, and is recorded as the graded deterioration grade result; specifically, all position points assigned with the highest grade indication quantity are extracted, and are arranged in descending order of cultural sensitivity weight within these position points to obtain a first sequence. The first sequence corresponds to the region to be intervened in priority, for example, a micro region where the drum-off, salt drum bump and flaky scale boundary appear in the inscription strokes of the stele. All position points assigned with the second grade indication quantity are extracted, and are also arranged in descending order of cultural sensitivity weight to obtain a second sequence. The second sequence corresponds to the region that needs to be consolidated and recorded in the future, for example, the shallow bump region at the corner of the decorative belt and the transition edge of the load-bearing part. Similarly, the third sequence and the fourth sequence are formed by the middle grade indication quantity and the low grade indication quantity respectively. After arranging the sequences in turn, a complete repair priority order list is obtained. The repair priority order list is registered together with the spatial coordinate information of each position point, the stability section mark, the cultural sensitivity weight and the corresponding deterioration grade indication quantity to form the graded deterioration grade result. The result can be directly used as the basis for repair intervention planning, reinforcement material allocation order and monitoring point review frequency arrangement, and can be archived in the protection technology file of the cultural relic body for long-term preservation.

[0126] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging, characterized in that, include: A detection area is set on the surface of the stone cultural relic, controlled thermal excitation is applied, and continuous thermal infrared imaging sequences of the heating and heat dissipation stages are collected. The thermal infrared imaging sequences are then radiometrically and geometrically corrected to form time-series thermal image data that are consistent with the actual thermal distribution on the surface of the cultural relic. Based on time-series thermal image data, the temperature rise amplitude, heat dissipation compliance, heat dissipation hysteresis characteristics, and sudden heat dissipation characteristics are extracted at each location to generate local degradation symptom information and construct the surface constraint fracture degree. Comparative analysis was conducted on the early and late response stages at the same location within the detection area. Based on the surface constraint fracture degree, the long-term residual temperature rise, heat dissipation compliance, heat dissipation hysteresis characteristics, and sudden heat dissipation characteristics were correlated to obtain expansion and contraction driving information. Within the detection area, spatial proximity is used as a constraint. The surface constraint fracture degree and expansion / contraction driving information at each location are compared with the corresponding information at adjacent locations to generate the degradation driving distribution results. Based on the degradation-driven distribution results, stability representation information is formed. Spatial continuity constraints are applied to the stability representation information to obtain a stability distribution map that is continuously distributed within the detection area. The stability distribution map is graded, and the priority of restoration is determined by combining the details of the cultural relic's decoration, inscription information, and the importance of key positions in its form, thus generating a graded deterioration level result.

2. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 1, characterized in that, The time-series thermal image data that corresponds to the actual thermal distribution on the surface of the cultural relics includes: A detection area was delineated on the surface of the stone artifact, and controlled thermal excitation was applied to the detection area so that the detection area experienced heat injection and release processes during the heating and heat dissipation phases, and thermal infrared imaging sequences at different time points were collected. The radiance signal of the thermal infrared imaging sequence is inverted to obtain the apparent temperature information corresponding to time. Based on the radiation characteristics of the surface material of stone cultural relics and the background of environmental radiation, radiation correction is performed on the apparent temperature information to generate equivalent surface temperature data; By establishing the geometric mapping relationship of the surface of cultural relics through structured light scanning, the equivalent surface temperature data is projected onto the reference plane coordinate system to form time-series thermal image data consistent with the actual surface geometry.

3. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 2, characterized in that, Based on features extracted from time-series thermal image data, the surface constraint fracture degree is generated, including: Within the detection area, the baseline and peak times of the heating stage are determined for each location, and the temperature rise is extracted based on the temperature difference between the peak time and the baseline time. During the heat dissipation phase, a preset time interval is selected, and the degree of heat dissipation compliance is determined based on the temperature change rate within the time interval. By selecting the temperature maintenance ratio during the sustained high-temperature phase after the peak, the heat dissipation hysteresis characteristic can be obtained. In the later stages of heat dissipation, identify the period of rapid temperature drop and calculate the characteristics of sudden heat dissipation based on the intensity of temperature drop during that period. By correlating the temperature rise, heat dissipation compliance, heat dissipation lag characteristics, and sudden heat dissipation characteristics, local deterioration signs information representing the bonding state between the surface and the substrate of cultural relics can be formed.

4. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 3, characterized in that, The construction of surface constraint fracture degree includes: By extracting the temperature rise amplitude and heat dissipation hysteresis characteristics from the local degradation symptom information, a surface heat retention relationship is established, and the degree of heat dissipation compliance is set as a constraint parameter. We apply weighted amplification to the sudden drop in heat dissipation characteristics to describe the instantaneous release features during the heat dissipation phase; The product of the temperature rise and the heat dissipation hysteresis characteristic is used as the cumulative amount of surface thermal response. Combined with the heat dissipation compliance process, it forms the constraint release amount. Combined with the sudden drop heat dissipation characteristic, it forms the surface constraint rupture degree, which represents the constraint relaxation state between the surface and the substrate in the detection area.

5. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 1, characterized in that, Obtaining expansion / contraction drive information includes: After the thermal excitation ends, the time intervals of the early response stage and the late response stage are determined based on the inflection point of the temperature change rate curve. The early response stage corresponds to the rapid temperature decay interval, and the late response stage corresponds to the slow temperature decay interval. Surface temperature information at corresponding times is extracted during the early response phase and the late response phase, and the residual temperature rise is calculated based on the difference between the late temperature and the baseline temperature. Establish a time-period retention rate based on residual heating amount and heating amplitude, and form a time-period consistency retention rate by combining heat dissipation hysteresis characteristics; By nonlinearly combining the sudden drop heat dissipation characteristics with the degree of heat dissipation compliance, the intensity of local release is obtained; Based on the weighted combination of surface constraint fracture degree, time-term consistency retention rate and local release intensity, expansion and contraction driving information is formed, which represents the temporal behavior characteristics of the detection location in terms of salt precipitation enrichment, bulging expansion and contraction, bulging separation or sheet-like peeling.

6. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 5, characterized in that, Within the detection area, the degradation-driven distribution results generated under spatial proximity constraints include: In the geometrically corrected reference coordinate domain, a spatial proximity set of the location point and its adjacent location points is established, and the proximity relationship is determined based on the distance threshold between the location points. Spatial correlation is performed between the surface constraint fracture degree and the expansion and contraction driving information at each location point, and the difference ratio of the surface constraint fracture degree and expansion and contraction driving information between the location point and the neighboring set is calculated. A weighted cumulative sum based on the proximity index is applied to the difference ratio to form a comparison of differences between locations; The difference comparison quantity is combined with the surface constraint fracture degree and expansion and contraction driving information at that location point to form the degradation driving distribution result, which represents the spatial difference state of each location in the detection area in terms of constraint relaxation trend and expansion and contraction driving intensity.

7. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 6, characterized in that, The information forming the stability representation includes: The degradation-driven distribution results are normalized to generate a driving normalization quantity, and the single-point vulnerability quantity is obtained by inversion based on the driving normalization quantity. The ratios of single-point vulnerability quantities are accumulated under spatial proximity to construct a neighborhood coupling balance term. An initial stability value is formed based on the single-point vulnerability quantity and the neighborhood coupling balance term. The initial stability value is then coordinated and aggregated with the initial stability values ​​of neighboring points to form a neighborhood coordination operator. Iterative convergence is achieved by combining the initial stability value with the neighborhood coordination operator, thereby generating iterative stability. By mapping the iterative stability to various positions in the reference coordinate domain, stability representation information that is continuously distributed within the detection region is obtained.

8. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 7, characterized in that, Applying spatial continuity constraints to the stability representation information yields a stability distribution map, including: Within the detection area, difference constraints are applied to the iterative stability of adjacent locations, and a coordination index is determined based on the distance between locations and surface connectivity. A continuous stability field is constructed by weighting the iterative stability of each location point using the coordination index as a weighting factor. In a continuous stability field, convergence judgment is performed based on the gradient of the change in iterative stability. When the stability difference between adjacent points is lower than a preset threshold, the spatial continuity state is determined. The stability field that satisfies the continuous state is mapped to the reference coordinate domain of the detection area to form a stability distribution map.

9. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 1, characterized in that, The classification of stability distribution maps includes: The stability value at each location in the stability distribution map is read, and the range of high-risk, medium-level, low-level, and stable sections is determined according to the preset stability classification threshold group. The stability values ​​at each location are compared with the grading thresholds to establish stability segment indicators; Based on the stability segment index, consecutive positions of the same segment within the detection area are aggregated to form a stability classification partition. The resulting stability grading and partitioning are mapped to the geometric coordinates of the artifact's surface to generate a stability segment map.

10. The method for evaluating the degree of surface deterioration in the restoration of stone cultural relics based on thermal infrared imaging according to claim 9, characterized in that, The priority order for restoration is determined by considering the details of the artifact's decoration, inscription information, and the importance of key positions in its form, including: On the stability segment map, cultural sensitivity weights are assigned to each location, and a cultural sensitivity matrix is ​​established based on the distribution of cultural relic patterns, inscription density, and structural feature location information. The cultural sensitivity matrix and the stability zone map are overlaid position by position to calculate the comprehensive grading index; Based on the correspondence between the comprehensive grading index and the preset grading threshold, grading assignment is performed to generate a degradation grading indicator. Arrange the position sequence according to the descending relationship between the degradation level indicator and the cultural sensitivity weight, determine the priority order of repair, and record it as the degradation level result.

Images