A method for enhancing joint images of underground water-sealed tunnel faces
Through technical means such as contrast adjustment, lighting equalization and nonlinear transformation, the palm surface joint image of the groundwater sealing library is optimized, which solves the problem of low image quality under the influence of light changes and improves image details and recognition accuracy.
Patent Information
- Application Number
- CN202411148397.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-08-21
AI Technical Summary
The palm surface joint image of the groundwater sealing library is affected by the changes in light, and the image quality is low, making it difficult to meet the subsequent analysis needs.
By acquiring palm surface joint images under different lighting conditions, contrast adjustment, lighting equalization adjustment, nonlinear transformation and adaptive parameter adjustment are performed, and the image processing algorithm is optimized to improve image quality using structural similarity index and constraint parameter search methods.
It effectively improves the quality of images collected from the palm surface of the groundwater sealing library, enhances image details, improves the accuracy of joint recognition, and reduces the interference of light changes on joint recognition.
Smart Images

Figure CN118657699B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of image processing, and in particular relates to an enhancement method for joint images of an underground water-sealed cavern face. Background Art
[0002] With the continuous advancement of underground water-sealed cavern projects, it is of great significance to study and analyze the development of tunnel face joints to judge the surrounding rock conditions for underground engineering construction. However, in practical applications, the images of tunnel face joints of underground water-sealed caverns are often affected by factors such as underground environment, instruments and equipment, and image acquisition technology, resulting in low image quality and difficulty in meeting the needs of subsequent analysis.
[0003] Therefore, an underground water-sealed cavern face joint image enhancement algorithm is urgently needed to effectively solve the problem of cavern surface illumination changes and thus ensure the image quality of the joint image. Summary of the invention
[0004] The present invention aims to solve at least one of the technical problems existing in the prior art and provide a new technical solution for an enhancement method of joint images of a tunnel face of an underground water-sealed cavern.
[0005] According to one aspect of the present invention, a method for enhancing joint images of an underground water-sealed tunnel face is provided, comprising the following steps:
[0006] Step S100, obtaining a first joint image of the tunnel face of an underground water-sealed cavern under a first lighting condition as an object to be processed;
[0007] Step S200, acquiring a second joint image of the underground water-sealed cavern face as a target object under a second lighting condition; wherein the illumination of the first lighting condition is less than the illumination of the second lighting condition;
[0008] Step S300, adjusting the contrast of the first joint image to obtain a first sub-image;
[0009] Step S400, performing illumination balance adjustment on the first sub-image to obtain a second sub-image;
[0010] Step S500, performing nonlinear transformation and adaptive parameter adjustment on the second sub-image to obtain a third sub-image;
[0011] Step S600, comparing and calculating the structural similarity index of the pixels of the third sub-image and the second joint image;
[0012] Step S700, using the constraint parameter search method to search the parameter space involved in the improved algorithm in steps S300~S500, adjusting the parameter space according to the search results and repeating steps S300~S600 until the structural similarity index between the object to be processed and the target object is less than the target threshold.
[0013] Optionally, in step S300, the contrast is adjusted by using the following method:
[0014] The regularization term is introduced to constrain the smoothness of the first sub-image to maintain the detail information; the regularization term adopts the total variation regularization method, the goal is to minimize the total variation of the image, and the calculation formula is as follows:
[0015] min G(x, y)=|grad(G(x,y))|+λ*∑|G(x,y)-F(x,y)|;
[0016] In the above formula, G(x, y) represents the adjusted image pixel value; F(x, y) represents the first joint image pixel value; grad represents the gradient operator, and λ represents the regularization parameter.
[0017] Optionally, in step S400, the following method is used to perform illumination balance adjustment:
[0018] The local contrast information is introduced for equalization adjustment. The calculation formula is as follows:
[0019] G(x,y)=F(x,y)+λ*C(x,y) * (F(x,y) -μ(x,y));
[0020] In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, C(x,y) represents the local contrast, μ(x,y) represents the local mean of the image, and λ is the regularization parameter.
[0021] Optionally, in step S500, the following method is used to perform nonlinear transformation and adaptive parameter adjustment:
[0022] The nonlinear transformation function is implemented by nonlinearly mapping the image pixels;
[0023] Adaptive adjustment parameters dynamically adjust the parameters in the algorithm according to the local characteristics and statistical information of the image. The calculation formula is as follows:
[0024] λ(x,y) =μ(x,y) + k *σ(x,y);
[0025] In the above formula, μ(x, y) represents the local mean of the image, σ(x, y) represents the local standard deviation of the image, and k is the adjustment factor.
[0026] Optionally, the following formula is used for nonlinear mapping:
[0027] G(x,y) = A*log(1+F(x,y));
[0028] In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, and A is the scaling parameter of the transformation function.
[0029] Optionally, in step S600, the structural similarity index is calculated using the following method:
[0030] SSIM(x,y)=(2*μx*μy+σx*σy) / (μx 2 +μy 2 +σx 2 +σy 2 );
[0031] In the above formula, μx and μy are the mean values of images x and y, respectively, and σx and σy are the standard deviations of images x and y, respectively.
[0032] Optionally, the illumination of the first lighting condition ranges from 10 to 50 lux.
[0033] Optionally, the illumination of the second lighting condition ranges from 100 to 500 lux.
[0034] Optionally, the constraint parameter search method includes random search, simulated annealing, particle swarm optimization and genetic algorithm.
[0035] Optionally, the target threshold value ranges from 0.9 to 0.995.
[0036] A technical effect of the present invention is:
[0037] In the embodiment of the present application, an image enhancement algorithm is used to process the illumination change problem of the joint image of the underground water-sealed cavern face, which can effectively improve the quality of the underground water-sealed cavern face acquisition image, enhance the detail information in the image, and thus improve the accuracy of joint recognition. Moreover, by sequentially adjusting the contrast, illumination balance, nonlinear transformation, and adaptive parameter adjustment of the first joint image, and comparing and calculating the structural similarity index of the pixel points of the third sub-image and the second joint image, until the structural similarity index of the object to be processed and the target object is less than the target threshold, the image can be made clearer and brighter, and the interference of illumination changes on joint recognition can be reduced.
[0038] Therefore, the enhancement method for the joint image of the underground water-sealed cavern face provided in this application is of great significance to the fields of underground energy storage, underground space development, etc., and can provide more accurate and reliable image data support for related research and practice. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 A schematic flow chart of a method for enhancing joint images of an underground water-sealed cavern face according to an embodiment of the present invention;
[0040] Figure 2 is the image obtained in step S100 of the present invention;
[0041] Figure 3 is the image obtained in step S200 of the present invention;
[0042] Figure 4 is the image acquired in step S300 of the present invention;
[0043] Figure 5 is the image acquired in step S400 of the present invention;
[0044] Figure 6 It is the image acquired in step S500 of the present invention. DETAILED DESCRIPTION
[0045] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present application.
[0046] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and should not be construed as limitations on the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.
[0047] The term "first" or "second" in the specification and claims of this application may include one or more of the features explicitly or implicitly. In the description of this application, unless otherwise specified, "plurality" means two or more. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally means that the objects connected before and after are in an "or" relationship.
[0048] According to one aspect of the present invention, see Figure 1, provides an enhancement method for the joint image of the underground water-sealed cavern face, which improves the contrast adjustment and illumination equalization effects by introducing nonlinear transformation functions and adaptive adjustment parameters; at the same time, by combining with the joint feature analysis of the underground water-sealed cavern face, accurate joint identification is achieved.
[0049] Specifically, the method for enhancing the joint image of the underground water-sealed cavern face comprises the following steps:
[0050] Step S100, obtaining a first joint image of the tunnel face of an underground water-sealed cavern under a first lighting condition as an object to be processed;
[0051] Step S200, acquiring a second joint image of the underground water-sealed cavern face as a target object under a second lighting condition; wherein the illumination of the first lighting condition is less than the illumination of the second lighting condition;
[0052] Step S300, adjusting the contrast of the first joint image to obtain a first sub-image;
[0053] Step S400, performing illumination balance adjustment on the first sub-image to obtain a second sub-image;
[0054] Step S500, performing nonlinear transformation and adaptive parameter adjustment on the second sub-image to obtain a third sub-image;
[0055] Step S600, comparing and calculating the structural similarity index of the pixels of the third sub-image and the second joint image;
[0056] Step S700, using the constraint parameter search method to search the parameter space involved in the improved algorithm in steps S300~S500, adjusting the parameter space according to the search results and repeating steps S300~S600 until the structural similarity index of the object to be processed and the target object is less than the target threshold. The purpose of the parameter space search is to find the best parameter combination so that the image enhancement algorithm can produce the best image quality, that is, while maintaining the image details, minimize the structural difference between the image to be processed and the target image.
[0057] It should be noted that parameter space refers to the set of all adjustable parameters in the algorithm, which together determine the behavior and output of the algorithm. Searching parameter space means that the algorithm attempts to find the optimal solution in parameter space, that is, those parameter values that can maximize image quality indicators (such as SSIM) through constrained parameter search methods (such as random search, simulated annealing, particle swarm optimization, and genetic algorithms). Adjusting parameter space means that based on the search results, the algorithm may need to adjust the parameter space, which may include changing the range of parameters, adding new parameters, or modifying the weights of existing parameters.
[0058] Further, once the parameter space is adjusted, the algorithm will repeat the image processing steps (S300 to S600), repeating the steps to apply the new parameters and re-evaluate the image quality. This process will be repeated until the structural similarity index between the image to be processed and the target image is less than a preset target threshold, which means that the image quality has reached or exceeded the required standard.
[0059] In simple terms, the search and adjustment of parameter space is to optimize the performance of the image enhancement algorithm and ensure that the final image output can meet specific quality requirements. This process is iterative and requires continuous evaluation and adjustment until satisfactory results are achieved.
[0060] In the embodiment of the present application, an image enhancement algorithm is used to process the illumination change problem of the joint image of the underground water-sealed cavern face, which can effectively improve the quality of the underground water-sealed cavern face acquisition image, enhance the detail information in the image, and thus improve the accuracy of joint recognition. Moreover, by sequentially adjusting the contrast, illumination balance, nonlinear transformation, and adaptive parameter adjustment of the first joint image, and comparing and calculating the structural similarity index of the pixel points of the third sub-image and the second joint image, until the structural similarity index of the object to be processed and the target object is less than the target threshold, the image can be made clearer and brighter, and the interference of illumination changes on joint recognition can be reduced.
[0061] Therefore, the enhancement method for the joint image of the underground water-sealed cavern face provided in this application is of great significance to the fields of underground energy storage, underground space development, etc., and can provide more accurate and reliable image data support for related research and practice.
[0062] Optionally, in step S300, the contrast is adjusted by using the following method:
[0063] The regularization term is introduced to constrain the smoothness of the first sub-image to maintain the detail information. The regularization term adopts the total variation regularization method, and the goal is to minimize the total variation of the image. The calculation formula is as follows:
[0064] min G(x, y)=|grad(G(x,y))|+λ*∑|G(x,y)-F(x,y)|;
[0065] In the above formula, G(x, y) represents the adjusted image pixel value; F(x, y) represents the first joint image pixel value; grad represents the gradient operator, and λ represents the regularization parameter.
[0066] In the above implementation, in the contrast adjustment algorithm, a regularization term can be introduced to constrain the smoothness of the adjusted image, so as to maintain the detail information of the image. At the same time, since the regularization term takes into account the smoothness of the image, the adjusted image is more natural and smooth. Therefore, after the regularization term is introduced, the contrast adjustment algorithm can better balance the contrast enhancement and detail retention of the image.
[0067] Optionally, in step S400, the following method is used to perform illumination balance adjustment:
[0068] The local contrast information is introduced for equalization adjustment. The calculation formula is as follows:
[0069] G(x,y)=F(x,y)+λ*C(x,y) * (F(x,y) -μ(x,y));
[0070] In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, C(x,y) represents the local contrast, μ(x,y) represents the local mean of the image, and λ is the regularization parameter.
[0071] It should be noted that in the illumination equalization algorithm, more complex image statistics can be considered, such as local contrast, color information, etc., to overcome some problems that traditional illumination equalization algorithms encounter when processing complex scenes. An improved method is to introduce local contrast information and perform equalization adjustments by considering the contrast of the neighborhood around each pixel in the image.
[0072] In the above implementation, by introducing local contrast information, the illumination equalization algorithm can better handle illumination changes in complex scenes and maintain image details and color information.
[0073] Optionally, in step S500, the following method is used to perform nonlinear transformation and adaptive parameter adjustment:
[0074] The nonlinear transformation function is implemented by nonlinearly mapping the image pixels;
[0075] Adaptive adjustment parameters dynamically adjust the parameters in the algorithm according to the local characteristics and statistical information of the image. The calculation formula is as follows:
[0076] λ(x,y) =μ(x,y) + k *σ(x,y);
[0077] In the above formula, μx and μy are the mean values of images x and y, respectively, and σx and σy are the standard deviations of images x and y, respectively.
[0078] It should be noted that the nonlinear transformation function can improve the contrast adjustment and illumination equalization effects by performing nonlinear mapping on image pixels. Since the nonlinear transformation function can provide better contrast adjustment and illumination equalization effects, it is particularly suitable for processing images with a wide brightness range.
[0079] Furthermore, the adaptive adjustment parameters can dynamically adjust the parameters in the algorithm according to the local features and statistical information of the image to achieve better image enhancement effects. By dynamically calculating the adaptive adjustment parameters according to the local statistical information of the image, it is possible to better balance contrast enhancement and detail preservation and adapt to the needs of different image regions.
[0080] In the above implementation, the introduction of nonlinear transformation functions and adaptive adjustment parameters can further improve the contrast adjustment and illumination equalization algorithms, making them suitable for a wider range of image enhancement scenarios for underground water-sealed cavern face joint acquisition.
[0081] Optionally, the following formula is used for nonlinear mapping:
[0082] G(x,y) = A*log(1+F(x,y));
[0083] In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, and A is the scaling parameter of the transformation function.
[0084] In the above implementation, the nonlinear transformation function is relatively simple and can perform nonlinear mapping on image pixels, thereby improving the effects of contrast adjustment and illumination equalization.
[0085] Optionally, the structural similarity index (SSIM) is a method for measuring the similarity of two images of rock joints of underground water-sealed caverns, which takes into account the structural information of the image and is sensitive to brightness, contrast and direction. In step S600, the structural similarity index (SSIM) is calculated using the following method:
[0086] SSIM(x,y)=(2*μx*μy+σx*σy) / (μx 2 +μy 2 +σx 2 +σy 2 );
[0087] In the above formula, μx and μy are the mean values of images x and y, respectively, and σx and σy are the standard deviations of images x and y, respectively.
[0088] In the above implementation, by calculating the structural similarity index (SSIM) to evaluate the similarity between the processed image and the target object image, the quality of the acquired image of the underground water-sealed cavern face can be effectively improved, the detail information in the image can be enhanced, and the accuracy of joint recognition can be improved.
[0089] Optionally, the illumination of the first lighting condition ranges from 10 to 50 lux.
[0090] In the above implementation, it is helpful to process the first joint image, so as to compare the structural similarity index of the object to be processed and the target object, thereby making the image clearer and brighter and reducing the interference of illumination changes on joint identification.
[0091] Optionally, the illumination of the second lighting condition ranges from 100 to 500 lux.
[0092] In the above implementation, it is helpful to make the image clearer and brighter by comparing the structural similarity index of the object to be processed and the target object, and reduce the interference of illumination changes on joint identification.
[0093] It should be noted that by limiting the range of illumination values, the following purposes can be achieved:
[0094] Improved adaptability: Limiting the illumination range can help the algorithm adapt to different underground environments, as actual lighting conditions may vary from place to place and time to time.
[0095] Improve image quality: Within a limited illumination range, it can ensure that the image will not lose details due to being too dark, nor will it cause highlight overflow due to being too bright.
[0096] Improve algorithm performance: Different lighting conditions have a significant impact on the performance of image processing algorithms. Limiting the illumination range helps optimize algorithm parameters and improve algorithm stability and accuracy.
[0097] Ensure balanced lighting: Within a limited range of illumination, the lighting balance algorithm can work more effectively and reduce the impact of uneven lighting on image quality.
[0098] Ensure the accuracy of joint identification: Appropriate illumination helps to improve the recognition accuracy of joint images, because the visibility of joints is closely related to lighting conditions.
[0099] Easy to experiment and test: In practical applications, the best parameter settings can be found within these illumination ranges through experiments and testing, thereby optimizing the image enhancement effect.
[0100] Optionally, the constraint parameter search method includes random search, simulated annealing, particle swarm optimization and genetic algorithm.
[0101] In the above implementation, the parameter space involved in the improved algorithm in each step can be quickly and accurately searched through the constrained parameter search method, and the structural similarity index between the object to be processed and the target object can be quickly achieved to be less than the target threshold by adjusting the parameter space.
[0102] Optionally, the target threshold value ranges from 0.9 to 0.995.
[0103] In the above implementation, the similarity between the processed image and the target object image can be evaluated through the value range of the target threshold, which can effectively improve the quality of the image collected at the underground water-sealed cavern face, enhance the detail information in the image, and thus improve the accuracy of joint recognition.
[0104] Target thresholds are often used to evaluate the effectiveness of image processing algorithms, especially in image quality metrics such as structural similarity (SSIM). The SSIM index is a value between -1 and 1, where 1 means that the two images are exactly the same and 0 or close to 0 means that the image quality is poor.
[0105] It should be noted that the lower limit of the target threshold is 0.9, which means that the structural similarity between the image to be processed and the target image must reach at least 90% similarity. The upper limit of the target threshold is close to 1 but not 1, such as 0.995, to allow a certain degree of natural difference, because completely identical images are difficult to achieve in practical applications. By limiting the target threshold, it can not only ensure that the results produced by the image enhancement algorithm at least meet a certain quality standard, but also provide a clear optimization target for the target threshold. The algorithm can adjust the parameters according to this threshold to achieve better image enhancement effects. At the same time, using the same target threshold in different image processing tasks can ensure the consistency of the results and facilitate comparison and evaluation. Furthermore, in practical applications, complete structural similarity (SSIM=1) may not be necessary, so setting a threshold close to 1 can balance image quality and processing efficiency. In addition, by adjusting the target threshold, the performance of the image enhancement algorithm can be flexibly adjusted according to the needs of different applications.
[0106] See also Figures 2 to 6 , Figures 2 to 6 They correspond to images acquired in different steps of the present invention, and each image represents the processing effect of the algorithm at different stages.
[0107] in, Figure 2 is the image obtained in step S100, which represents the original joint image of the underground water-sealed cavern face obtained under the first lighting condition. Figure 2Reflects image quality in lower lighting conditions, which may include uneven lighting, low contrast, or blurred details.
[0108] Figure 3 is the image obtained in step S200, which represents the joint image of the tunnel face obtained under the second lighting condition, as the target object.
[0109] and Figure 2 compared to, Figure 3 Improvements in images under higher illumination conditions, such as better lighting conditions and contrast, are reflected, but further enhancements are still needed to improve the identification accuracy of joints.
[0110] Figure 4 is the image acquired in step S300, which is an image after contrast adjustment, that is, the first sub-image. Figure 4 This reflects the effect of the contrast adjustment step, including enhancing the local contrast of the image and making the joint features more obvious.
[0111] Figure 5 It is the image acquired in step S400, which is the image after illumination balance adjustment, that is, the second sub-image. Figure 5 It reflects the effect of lighting balance adjustment and shows how to reduce the problem of uneven lighting in the image, making the image look more balanced and natural as a whole.
[0112] Figure 6 It is the image obtained in step S500, which is the image after nonlinear transformation and adaptive parameter adjustment, that is, the third sub-image. Figure 6 The resulting image enhancements are shown, including improved image detail, contrast, and overall visual quality for easier joint identification and analysis.
[0113] pass Figure 2 and Figure 6 , it can be shown how the algorithm can gradually improve the image quality and ultimately achieve the goal of improving the accuracy of joint identification.
[0114] In an embodiment of the present application, the enhancement method for the joint image of the underground water-sealed cavern face improves the effects of contrast adjustment and lighting equalization, and can reduce the loss of image details, thereby achieving accurate identification of the joints of the underground water-sealed cavern face, and is suitable for practical application needs.
[0115] Therefore, the enhancement method for the joint image of the underground water-sealed tunnel face has wide application value and economic benefits in the field of joint identification of the underground water-sealed tunnel face.
[0116] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.
Claims
1. A method for enhancing joint images of underground water-sealed cavern faces, characterized in that: The steps include: Step S100, obtaining a first joint image of the tunnel face of an underground water-sealed cavern under a first lighting condition as an object to be processed; Step S200, acquiring a second joint image of the underground water-sealed cavern face as a target object under a second lighting condition; wherein the illumination of the first lighting condition is less than the illumination of the second lighting condition; Step S300, adjusting the contrast of the first joint image to obtain a first sub-image; In step S300, the contrast is adjusted by the following method: The regularization term is introduced to constrain the smoothness of the first sub-image to maintain the detail information; the regularization term adopts the total variation regularization method, the goal is to minimize the total variation of the image, and the calculation formula is as follows: min G(x, y)=|grad(G(x,y))|+λ*∑|G(x,y)-F(x,y)|; In the above formula, G(x,y) represents the adjusted image pixel value; F(x,y) represents the first joint image pixel value; grad represents the gradient operator, and λ represents the regularization parameter; Step S400, performing illumination balance adjustment on the first sub-image to obtain a second sub-image; In step S400, the following method is used to perform illumination balance adjustment: The local contrast information is introduced for equalization adjustment. The calculation formula is as follows: G(x,y)=F(x,y)+λ*C(x,y) * (F(x,y) -μ(x,y)); In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, C(x,y) represents the local contrast, μ(x,y) represents the local mean of the image, and λ is the regularization parameter; Step S500, performing nonlinear transformation and adaptive parameter adjustment on the second sub-image to obtain a third sub-image; In step S500, the following method is used to perform nonlinear transformation and adaptive parameter adjustment: The nonlinear transformation function is implemented by nonlinearly mapping the image pixels; Adaptive adjustment parameters dynamically adjust the parameters in the algorithm according to the local characteristics and statistical information of the image. The calculation formula is as follows: λ(x,y) =μ(x,y) + k *σ(x,y); In the above formula, μ(x,y) represents the local mean of the image, σ(x,y) represents the local standard deviation of the image, and k is the adjustment factor; The following formula is used for nonlinear mapping: G(x,y) = A*log(1+F(x,y)); In the above formula, G(x,y) represents the adjusted image pixel value, F(x,y) represents the first joint image pixel value, and A is the scaling parameter of the transformation function; Step S600, comparing and calculating the structural similarity index of the pixels of the third sub-image and the second joint image; Step S700, using a constrained parameter search method to search the parameter space involved in the improved algorithm in steps S300~S500, adjusting the parameter space according to the search results and repeating steps S300~S600 until the structural similarity index of the object to be processed and the target object is less than the target threshold; the constrained parameter search method includes random search, simulated annealing, particle swarm optimization and genetic algorithm.
2. The method for enhancing the joint image of the underground water-sealed cavern face according to claim 1, characterized in that: In step S600, the structural similarity index is calculated using the following method: SSIM(x,y)=(2*μx*μy+σx*σy) / (μx 2 +μy 2 +σx 2 +σy 2 ); In the above formula, μx and μy are the mean values of images x and y, respectively, and σx and σy are the standard deviations of images x and y, respectively.
3. The method for enhancing the joint image of the underground water-sealed cavern face according to claim 1, characterized in that: The illumination value of the first illumination condition ranges from 10 to 50 lux.
4. The method for enhancing the joint image of the underground water-sealed cavern face according to claim 1, characterized in that: The illumination value of the second illumination condition ranges from 100 to 500 lux.
5. The method for enhancing the joint image of the underground water-sealed cavern face according to claim 1, characterized in that: The target threshold value ranges from 0.9 to 0.995.
