Dam water seepage monitoring method based on remote sensing image

Abstract

The invention discloses a dam water seepage monitoring method based on a remote sensing image, and belongs to the technical field of image processing. The method comprises the following steps: collecting a plurality of remote sensing images of the same dam wall area in different time periods under a sunlight condition, and distinguishing a background area and a contrast area through clustering processing; acquiring a pixel water seepage characteristic value according to the color difference and generating a characteristic graph; finding out an overlapped comparison area of the same water seepage target, and calculating a water seepage diffusion goodness of fit, a mean value and a water seepage change degree; performing enhancement processing on the pixel water seepage feature map by using the indexes; and finally, processing the enhanced graph through a dam water seepage evaluation neural network to obtain a water seepage score, thereby realizing accurate monitoring and evaluation of the dam water seepage condition.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a dam seepage monitoring method based on remote sensing images. BACKGROUND

[0002] As the core facility of water conservancy projects, dams play an irreplaceable role in flood control, water supply, power generation and other fields. The safe and stable operation of dams is directly related to people's life and property safety and social and economic development. However, with the continuous development of water conservancy construction, many dams have increased in design height to meet multiple demands such as flood control and power generation, and some high dams even reach hundreds of meters. Such height makes manual visual observation extremely difficult, and the seepage condition of the upper part of the dam body is difficult to be discovered in time, greatly increasing the safety hazard.

[0003] Due to the influence of factors such as long-term water flow erosion, geological condition changes, and material aging, dam walls are prone to seepage problems. If not discovered and handled in time, seepage may gradually intensify, causing damage to the dam structure, reducing the strength, and even leading to dam collapse and other major safety accidents. Therefore, it is crucial to accurately and efficiently monitor the seepage condition of the dam.

[0004] Traditional dam seepage monitoring methods mostly rely on manual inspection or sensor deployment. Manual inspection not only consumes a lot of manpower and resources, but also is limited by the experience and field of view of the inspectors. Especially when facing high dams, it is difficult to cover all areas of the dam, and the seepage points at high altitudes of the dam body often become blind spots of monitoring, with a serious risk of missed detection. Although the sensor deployment method can achieve a certain degree of automatic monitoring, the installation and maintenance of sensors are costly and easily disturbed by environmental factors, and the monitoring range is also limited, making it difficult to meet the overall monitoring needs of large high dams.

[0005] The existing method can obtain macro information of the dam wall from remote sensing images, but how to highlight the seepage features from the remote sensing images and accurately identify the seepage condition is still a key problem currently faced. SUMMARY

[0006] In view of the above shortcomings in the prior art, the dam seepage monitoring method based on remote sensing images provided by the present application solves the problem that the seepage features in the remote sensing images are not prominent, resulting in inaccurate identification of the seepage condition.

[0007] In order to achieve the above-mentioned application purposes, the technical scheme adopted by the present application is as follows: a dam seepage monitoring method based on remote sensing images, comprising the following steps: Under daylight conditions, multiple remote sensing images of the same dam wall area at different time periods are collected, each remote sensing image is subjected to clustering processing to obtain multiple connected regions, and background regions and contrast regions are distinguished; According to the color difference between the contrast region and the background region, a pixel water seepage feature value of each pixel point in the contrast region is obtained, and a pixel water seepage feature map is obtained; On the plurality of pixel water seepage feature maps, the contrast regions belonging to the same water seepage target with area overlap are found out; The water seepage diffusion coincidence degree of each contrast region in the pixel water seepage feature map is obtained, and the water seepage diffusion coincidence degrees belonging to the same water seepage target are averaged; According to the contrast regions belonging to the same water seepage target, the water seepage change degree of each water seepage target is obtained; The water seepage diffusion coincidence degree average and the water seepage change degree are respectively used for enhancement processing on any one pixel water seepage feature map, and a first pixel water seepage feature enhancement map and a second pixel water seepage feature enhancement map are obtained; The dam water seepage evaluation neural network is used to process the first pixel water seepage feature enhancement map and the second pixel water seepage feature enhancement map, and a water seepage score is obtained.

[0008] In some preferred embodiments of the present application, the process of obtaining the pixel water seepage feature map comprises: The average of each brightness value in the background region is obtained, and the background brightness is obtained; The ratio of the background brightness to the brightness value of each pixel point in the contrast region is taken as the pixel brightness ratio; The channel difference is obtained by subtracting the channel value of each pixel point in the contrast region from the background channel average; The channel difference value is obtained by adding the weighted 3 channel differences; The pixel water seepage feature value is obtained by multiplying the channel difference value of the same pixel point in the contrast region with the pixel brightness ratio, and the multiplication result is normalized, the pixel water seepage feature value of the pixel point in the background region is set to 0, and the pixel water seepage feature map is obtained.

[0009] In some preferred embodiments of the present application, the formula for obtaining the channel difference value is: , Wherein, C i is the channel difference value of the i th pixel point in the contrast region, △R i is the R channel difference of the i th pixel point, △G i is the G channel difference of the i th pixel point, and △B i is the B channel difference of the i th pixel point.

[0010] In some preferred embodiments of the present application, the process of obtaining the water seepage diffusion coincidence degree comprises: A contrast region of the pixel water seepage feature map is divided into a plurality of N×M size image blocks, wherein N and M are positive integers; The average of the pixel water seepage feature value of each image block is taken. selecting the image block corresponding to the mean value of the pixel water seepage feature of the maximum pixel water seepage feature in the contrast region as the center image block; in the horizontal direction, taking the center image block as the starting point, calculating the difference in the mean value of the pixel water seepage feature of the adjacent image blocks on the left and right sides; in the vertical direction, taking the center image block as the starting point, calculating the difference in the mean value of the pixel water seepage feature of the adjacent image blocks on the top and bottom sides; taking the ratio of the number of difference values greater than 0 to the total number of difference values as the water seepage diffusion coincidence degree.

[0011] In some preferred embodiments of the present application, in the left direction of the center image block, the mean value of the pixel water seepage feature of the image block on the right is subtracted from the mean value of the pixel water seepage feature of the image block on the left to obtain the difference value; in the right direction of the center image block, the mean value of the pixel water seepage feature of the image block on the left is subtracted from the mean value of the pixel water seepage feature of the image block on the right to obtain the difference value; in the top direction of the center image block, the mean value of the pixel water seepage feature of the image block on the bottom is subtracted from the mean value of the pixel water seepage feature of the image block on the top to obtain the difference value; in the bottom direction of the center image block, the mean value of the pixel water seepage feature of the image block on the top is subtracted from the mean value of the pixel water seepage feature of the image block on the bottom to obtain the difference value.

[0012] In some preferred embodiments of the present application, the process of obtaining the water seepage change degree of each water seepage target includes: selecting the contrast region with the largest area and the contrast region with the smallest area in the contrast regions belonging to the same water seepage target; subtracting the contrast region with the smallest area from the contrast region with the largest area to obtain the maximum area difference; normalizing the maximum area difference to obtain the water seepage change degree of the water seepage target.

[0013] In some preferred embodiments of the present application, the enhancement process includes: using the mean value of the water seepage diffusion coincidence degree of the same water seepage target to enhance the contrast region of the same water seepage target in the pixel water seepage feature map to obtain a first pixel water seepage feature enhancement map; using the water seepage change degree of the same water seepage target to enhance the contrast region of the same water seepage target in the pixel water seepage feature map to obtain a second pixel water seepage feature enhancement map.

[0014] In some preferred embodiments of the present application, the formula for enhancement processing is: , , wherein, Vz1，i V is the i-th enhanced value of the first pixel water seepage feature enhancement map on the contrast region of the same water seepage target i V is the i-th pixel water seepage feature value on the contrast region of the same water seepage target on the pixel water seepage feature map, c is the water seepage diffusion goodness of fit mean value of the same water seepage target, V z2，i V is the i-th enhanced value of the second pixel water seepage feature enhancement map on the contrast region of the same water seepage target, and ε is the water seepage change degree of the same water seepage target.

[0015] In some preferred embodiments of the present application, the dam water seepage evaluation neural network comprises: a first parallel feature extraction unit, a second parallel feature extraction unit, a cross-branch gated interactive feature enhancement unit, a CNN unit and a fully connected layer; The input end of the first parallel feature extraction unit is used for inputting the first pixel water seepage feature enhancement map; and the input end of the second parallel feature extraction unit is used for inputting the second pixel water seepage feature enhancement map. The first input end of the cross-branch gated interactive feature enhancement unit is connected with the output end of the first parallel feature extraction unit, the second input end thereof is connected with the output end of the second parallel feature extraction unit, and the output end thereof is connected with the input end of the CNN unit; and the input end of the fully connected layer is connected with the output end of the CNN unit, and the output end thereof serves as the output end of the dam water seepage evaluation neural network.

[0016] In some preferred embodiments of the present application, the cross-branch gated interactive feature enhancement unit comprises: a first gating module, a second gating module, a multiplier M1, a multiplier M2 and an adder A1. The input end of the first gating module is connected with the first input end of the multiplier M1 and serves as the first input end of the cross-branch gated interactive feature enhancement unit; the input end of the second gating module is connected with the first input end of the multiplier M2 and serves as the second input end of the cross-branch gated interactive feature enhancement unit; the second input end of the multiplier M1 is connected with the output end of the first gating module, and the output end thereof is connected with the first input end of the adder A1; the second input end of the multiplier M2 is connected with the output end of the second gating module, and the output end thereof is connected with the second input end of the adder A1; and the output end of the adder A1 serves as the output end of the cross-branch gated interactive feature enhancement unit.

[0017] The present application has the following beneficial effects: 1. The present application can accurately distinguish the background region and the contrast region by clustering the remote sensing images of the same dam wall surface region in different time periods, and then obtain the pixel water seepage feature value by combining the color difference between the contrast region and the background region, highlight the water seepage feature of the potential water seepage region, make the water seepage feature prominent, and improve the accuracy of identifying the water seepage condition.

[0018] 2、The application can highlight the characteristics of the water seepage target, further enhance the difference between the water seepage area and other areas, reduce the influence of interference factors, and then use the dam water seepage evaluation neural network to process the first pixel water seepage feature enhancement map and the second pixel water seepage feature enhancement map, combine the enhanced water seepage features in the two maps, and improve the accuracy of identifying the water seepage condition. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0020] Figure 1 A flowchart of a dam water seepage monitoring method based on remote sensing images is provided for the embodiments of the present application. Figure 2 A schematic diagram of the division of the contrast area into multiple N*M size image blocks is provided for the embodiments of the present application. Figure 3 A structural schematic diagram of a dam water seepage evaluation neural network is provided for the embodiments of the present application. Figure 4 A structural schematic diagram of a cross-branch gate interaction feature enhancement unit is provided for the embodiments of the present application. Figure 5 A structural schematic diagram of a first gate module and a second gate module is provided for the embodiments of the present application. Figure 6 A structural schematic diagram of a first parallel feature extraction unit and a second parallel feature extraction unit is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0023] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0024] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. In addition, the terms "first," "second," "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0025] Furthermore, terms such as "horizontal," "vertical," and "sag" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0026] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0027] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0028] Example 1 like Figure 1 As shown, a method for monitoring dam seepage based on remote sensing images includes the following steps: Under sunlight conditions, multiple remote sensing images of the same dam wall surface area at different time periods are collected, each remote sensing image is clustered to obtain multiple connected regions, and the background region and the contrast region are distinguished; According to the color difference between the contrast region and the background region, the pixel water seepage feature value of each pixel point in the contrast region is obtained, and a pixel water seepage feature map is obtained; On multiple pixel water seepage feature maps, find the contrast regions belonging to the same water seepage target that exist area overlap; Obtain the water seepage diffusion coincidence degree of each contrast region in the pixel water seepage feature map, and take the mean value of the water seepage diffusion coincidence degree of each water seepage target belonging to the same water seepage target; According to the contrast region belonging to the same water seepage target, the water seepage change degree of each water seepage target is obtained; The water seepage diffusion coincidence degree mean value and the water seepage change degree are respectively used for enhancement processing on any one pixel water seepage feature map to obtain a first pixel water seepage feature enhancement map and a second pixel water seepage feature enhancement map. The dam water seepage evaluation neural network processes the first pixel water seepage feature enhancement map and the second pixel water seepage feature enhancement map to obtain a water seepage score.

[0029] In this embodiment, the clustering process includes: taking any unclassified pixel point as an initial clustering point, including the pixel points similar to it in the clustering point neighborhood into the same region, taking the newly included pixel point as a new clustering point, repeating the expansion, until the region cannot add new pixel points, forming a connected region, and continuing to classify the remaining unclassified pixel points until the dam wall surface remote sensing image is classified, obtaining multiple connected regions.

[0030] The clustering detailed process includes the following steps: A1, taking any pixel point without a connected region as a clustering point, obtaining the similarity of each pixel point in the neighborhood range of the clustering point and the clustering point in the pixel value, and when the similarity is greater than the similarity threshold, the corresponding pixel point in the neighborhood range and the clustering point are classified into a connected region; A2, taking any pixel point in the connected region as a new clustering point, obtaining the similarity of each pixel point in the neighborhood range of the new clustering point and the new clustering point in the pixel value, and when the similarity is greater than the similarity threshold, the corresponding pixel point in the neighborhood range is included in the connected region; A3, loop execution A2, until the connected region cannot be expanded, the connected region is divided; A4, loop execution A1, divide and extract the next connected region, until the dam wall surface remote sensing image is divided.

[0031] In the embodiment, the largest area connected region is taken as the background region, and other connected regions are taken as the contrast region.

[0032] In the embodiment, the process of obtaining the pixel water seepage feature map comprises: The average of each brightness value in the background region is obtained as the background brightness; The ratio of the background brightness and the brightness value of each pixel point in the contrast region is taken as the pixel brightness ratio; The channel difference is obtained by subtracting the channel value of each pixel point in the contrast region from the background channel average; The channel difference values are obtained by adding the weights of the three channel differences; The pixel water seepage feature value of each pixel point in the contrast region is obtained by multiplying the channel difference value of the same pixel point in the contrast region and the pixel brightness ratio, and the multiplication result is normalized, the pixel water seepage feature value of the pixel point in the background region is set to 0, and the pixel water seepage feature map is obtained.

[0033] In the embodiment, the formula of the normalization processing of the multiplication result is: sz=z / z max , wherein, sz is the pixel water seepage feature value, z is the multiplication result of any pixel point, and z max is the maximum multiplication result.

[0034] The surface of the water seepage region is covered with water, the reflectivity is reduced, and the overall brightness is lower than the background, so the ratio of “background brightness / pixel brightness” will be greater than 1 (the greater the pixel brightness ratio, the more significant the darkening feature).

[0035] Water has stronger absorption to visible light (especially blue light), which causes the RGB channel value (especially the blue channel) of the water seepage region to be significantly lower than the dry background region.

[0036] The application uses the multiplication operation of “channel difference value x pixel brightness ratio” to cooperatively amplify the two features of “color shift” and “brightness reduction” caused by water, so that the pixel water seepage feature value of the water seepage region is significantly higher than the non-water seepage region (such as dry stains and textures), and the feature value is positively correlated with the water seepage degree by using the action rule of water on light.

[0037] In the embodiment, the formula for obtaining the channel difference value is: , wherein, C i is the channel difference value of the i-th pixel point in the contrast region, △R i is the R channel difference of the i-th pixel point, △G i is the G channel difference of the i-th pixel point, and △B i is the B channel difference of the i-th pixel point.

[0038] In the present application, the weight of R channel difference is 0.2, the weight of G channel difference is 0.3, and the weight of B channel difference is 0.5. The water has different absorption degrees for different wavelengths of visible light, and the absorption of blue light (corresponding to the B channel) is relatively stronger. When calculating the channel difference value, the B channel difference is given the highest weight of 0.5, the G channel difference is given a weight of 0.3, and the R channel difference is given a weight of 0.2. The difference between the water seepage area and the background in the B channel can be highlighted, so that the optical characteristic change of the water seepage area caused by the presence of water can be accurately captured, and the identification accuracy of the water seepage area can be improved.

[0039] In the present embodiment, the process of finding the contrast regions of the same water seepage target includes the following steps: Any contrast region in any pixel water seepage feature map R is denoted as contrast region A; The intersection of the contrast region A and each contrast region in the remaining pixel water seepage feature map K is obtained. When the intersection is not empty, and there is only one contrast region in each pixel water seepage feature map K whose intersection with the contrast region A is not empty, the contrast region in the pixel water seepage feature map K is denoted as contrast region B. If there are multiple contrast regions, the contrast region corresponding to the largest intersection in the pixel water seepage feature map K is denoted as contrast region B; The contrast region A and each contrast region B are contrast regions belonging to the same water seepage target. The process is repeated to obtain the contrast regions belonging to each water seepage target.

[0040] The contrast regions belonging to the same water seepage target satisfy the following conditions: , wherein G1 is the pixel point coordinate set of the contrast region belonging to the water seepage target on the pixel water seepage feature map of the first time period, G2 is the pixel point coordinate set of the contrast region belonging to the water seepage target on the pixel water seepage feature map of the second time period, G t is the pixel point coordinate set of the contrast region belonging to the water seepage target on the pixel water seepage feature map of the tth time period, G T is the pixel point coordinate set of the contrast region belonging to the water seepage target on the pixel water seepage feature map of the Tth time period, and is the intersection, is an empty set.

[0041] In the present embodiment, the process of finding the contrast regions belonging to the same water seepage target includes the following steps: The contrast regions that do not satisfy the condition

[0042] In the present embodiment, the process of obtaining the water seepage diffusion coincidence degree includes: A contrast region of the pixel water seepage feature map is divided into a plurality of image blocks of size N x M, wherein N and M are positive integers; The pixel water seepage feature value of each image block is averaged; Select the maximum pixel water seepage feature mean corresponding image block inside the contrast region as the center image block; In the horizontal direction, taking the center image block as the starting point, the difference in pixel water seepage feature mean of the left and right adjacent image blocks is calculated; In the vertical direction, taking the center image block as the starting point, the difference in pixel water seepage feature mean of the upper and lower adjacent image blocks is calculated; The ratio of the number of difference values greater than 0 to the total number of difference values is taken as the water seepage diffusion coincidence degree.

[0043] In this embodiment, the size of the image block is set to 5x5, that is, vertical cutting is performed by 5 pixel width, and horizontal cutting is performed by 5 pixel width, and the size of the edge image block is less than 5x5, only the pixel water seepage feature value existing in the edge image block is considered.

[0044] In this embodiment, in the left direction of the center image block, the pixel water seepage feature mean of the right image block in the adjacent image block is subtracted from the pixel water seepage feature mean of the left image block to obtain the difference value; In the right direction of the center image block, the pixel water seepage feature mean of the left image block in the adjacent image block is subtracted from the pixel water seepage feature mean of the right image block to obtain the difference value; In the upper direction of the center image block, the pixel water seepage feature mean of the lower image block in the adjacent image block is subtracted from the pixel water seepage feature mean of the upper image block to obtain the difference value; In the lower direction of the center image block, the pixel water seepage feature mean of the upper image block in the adjacent image block is subtracted from the pixel water seepage feature mean of the lower image block to obtain the difference value.

[0045] Dam water seepage is gradually diffused from the seepage point (center region) to the surrounding, and the water seepage feature (such as humidity, optical performance) of the center region is the strongest, and gradually weakens to the periphery.

[0046] The present application selects the "maximum pixel water seepage feature mean corresponding image block as the center image block", which can accurately locate the potential leakage core region, and when calculating the feature mean difference, the left side uses "right minus left", the right side uses "left minus right", the upper side uses "down minus up", and the lower side uses "up minus down". This difference calculation method is used to judge whether the "peripheral image block feature is weaker than the adjacent block in the center direction": if the difference is greater than 0, it means that from the center to that direction, the feature shows a decreasing trend of "strong center and weak periphery", which can effectively capture the diffusion characteristics of real water seepage.

[0047] For example, Figure 2As shown, the area of the center image block O with the red arrow pointing upwards is the upper side direction, the area of the center image block O with the red arrow pointing downwards is the lower side direction, the area of the center image block O with the red arrow pointing left is the left side direction, and the area of the center image block O with the red arrow pointing right is the right side direction, in the lower side direction, H1 and H2 are a pair of adjacent image blocks, H1 is the image block above, and H2 is the image block below.

[0048] In the embodiment, the process of obtaining the water seepage change degree of each water seepage target comprises: selecting the largest-area contrast region and the smallest-area contrast region in the contrast regions belonging to the same water seepage target; subtracting the largest-area contrast region from the smallest-area contrast region to obtain the maximum area difference; normalizing the maximum area difference to obtain the water seepage change degree of the water seepage target.

[0049] The present application collects multiple remote sensing images of the same dam wall surface region at different time periods under sunlight conditions, and therefore, the dam water seepage is a dynamic process changing with time, and the water seepage region will change in area due to continuous water seepage, evaporation, etc. at different time periods (such as early, middle, and late) under sunlight. By selecting the maximum and minimum contrast region areas of the same water seepage target from multiple time-series remote sensing images, the spatial range extreme difference of the water seepage region in the time dimension can be captured. This difference directly corresponds to the activity degree of water seepage, and the greater the area difference, the more significant the development or recession of water seepage in the time series, further embodying the characteristics that water seepage is a dynamic process and the area changes with time.

[0050] In the embodiment, the formula for the normalization processing of the maximum area difference is: rE=E / (E+1), rE is the water seepage change degree of the water seepage target, and E is the maximum area difference.

[0051] In the embodiment, the enhancement processing process comprises: using the water seepage diffusion coincidence degree mean value of the same water seepage target to perform enhancement processing on the contrast region of the same water seepage target in the pixel water seepage feature map to obtain a first pixel water seepage feature enhancement map; using the water seepage change degree of the same water seepage target to perform enhancement processing on the contrast region of the same water seepage target in the pixel water seepage feature map to obtain a second pixel water seepage feature enhancement map.

[0052] The contrast regions on the first pixel water seepage feature enhancement map are all enhanced, and the contrast regions on the second pixel water seepage feature enhancement map are all enhanced.

[0053] In the embodiment, the formula for the enhancement processing is: , wherein, V z1，iV is the i-th enhanced value of the first pixel water seepage feature enhancement map on the contrast region of the same water seepage target, and V is the i-th enhanced value of the second pixel water seepage feature enhancement map on the contrast region of the same water seepage target i V is the i-th enhanced value of the first pixel water seepage feature enhancement map on the contrast region of the same water seepage target, and c is the water seepage diffusion coincidence degree mean value of the same water seepage target z2，i V is the i-th enhanced value of the first pixel water seepage feature enhancement map on the contrast region of the same water seepage target, and c is the water seepage diffusion coincidence degree mean value of the same water seepage target

[0054] The present application utilizes the water seepage diffusion coincidence degree mean value and the water seepage change degree of the same water seepage target to enhance the corresponding contrast region in the pixel water seepage feature map respectively. The water seepage diffusion coincidence degree reflects the diffusion feature fitting degree of the water seepage region, and the water seepage change degree embodies the change amplitude of the water seepage region. Through the enhancement processing, the features of the water seepage region are significantly distinguished from other regions, so that the water seepage feature is more prominent, and the water seepage monitoring precision is improved.

[0055] As shown in Figure 3 The dam water seepage evaluation neural network comprises a first parallel feature extraction unit, a second parallel feature extraction unit, a cross-branch gated interactive feature enhancement unit, a CNN unit and a full connection layer. The input end of the first parallel feature extraction unit is used for inputting the first pixel water seepage feature enhancement map; and the input end of the second parallel feature extraction unit is used for inputting the second pixel water seepage feature enhancement map. The first input end of the cross-branch gated interactive feature enhancement unit is connected with the output end of the first parallel feature extraction unit, the second input end thereof is connected with the output end of the second parallel feature extraction unit, and the output end thereof is connected with the input end of the CNN unit; and the input end of the full connection layer is connected with the output end of the CNN unit, and the output end thereof serves as the output end of the dam water seepage evaluation neural network.

[0056] The present application extracts features from the first pixel water seepage feature enhancement map and the second pixel water seepage feature enhancement map through the first parallel feature extraction unit and the second parallel feature extraction unit respectively, then performs interactive processing on the two parallel extracted features through the cross-branch gated interactive feature enhancement unit, the gated mechanism adaptively adjusts the weights of different branch features, so that the fused features can more accurately reflect the essential features of the water seepage target, improve the expression ability of the features, finally extract the features through the CNN unit, and output the water seepage score through the full connection layer.

[0057] As shown in Figure 4 The cross-branch gated interactive feature enhancement unit comprises a first gating module, a second gating module, a multiplier M1, a multiplier M2 and an adder A1. The input end of the first gating module is connected with the first input end of the multiplier M1 and serves as the first input end of the cross-branch gating interactive feature enhancement unit, and the input end of the second gating module is connected with the first input end of the multiplier M2 and serves as the second input end of the cross-branch gating interactive feature enhancement unit; the second input end of the multiplier M1 is connected with the output end of the first gating module, and the output end thereof is connected with the first input end of the adder A1; the second input end of the multiplier M2 is connected with the output end of the second gating module, and the output end thereof is connected with the second input end of the adder A1; and the output end of the adder A1 serves as the output end of the cross-branch gating interactive feature enhancement unit.

[0058] According to the characteristics of the input features, the first gating module and the second gating module can adaptively generate weights, and then the weights are multiplied with the input feature map bit by bit, and the two weighted feature maps are added bit by bit through the adder A1, so that the proportion of different source features in the fusion result can be dynamically adjusted based on the numerical value of the weight at each corresponding position of the feature map.

[0059] As shown in Figure 5 , the first gating module and the second gating module each include a first convolutional layer, a BN layer, a ReLU layer, a second convolutional layer and a Sigmoid layer. The gating module generates a weight map with the same size as the input feature map through "1x1 convolution -> BN -> ReLU -> 1x1 convolution -> Sigmoid".

[0060] As shown in Figure 6 , the first parallel feature extraction unit and the second parallel feature extraction unit each include a third convolutional layer, a fourth convolutional layer, a fifth convolutional layer and a Concat layer, the convolution kernel size of the third convolutional layer is 3x3, the convolution kernel size of the fourth convolutional layer is 5x5, and the convolution kernel size of the fifth convolutional layer is 7x7.

[0061] The water seepage score ranges from 0 to 100 points, and the corresponding conditions are as follows: 0-30 points: There is no obvious water seepage in the dam, and the dam body is in a normal stable state.

[0062] 31-60 points: There is slight water seepage in the dam, which will not pose a serious threat to the safety of the dam body, but monitoring needs to be strengthened.

[0063] 61-85 points: The dam seepage is relatively obvious, and if not handled in time, it may gradually worsen the dam structure.

[0064] 86-100 points: The dam seepage is serious, which has posed a great threat to the safety of the dam, and immediate measures need to be taken to handle it.

[0065] The application can accurately distinguish the background area and the contrast area by clustering the remote sensing images of the same dam wall surface area in different periods, and then obtain the pixel water seepage characteristic value by combining the color difference between the contrast area and the background area, highlight the water seepage characteristics of the potential water seepage area, make the water seepage characteristics obvious, and improve the accuracy of identifying the water seepage condition.

[0066] The application can highlight the characteristics of the water seepage target by finding the contrast areas belonging to the same water seepage target in multiple pixel water seepage feature maps, and enhancing the pixel water seepage feature map by combining the water seepage diffusion coincidence degree mean value and the water seepage change degree, further enhance the difference between the water seepage area and other areas, reduce the influence of interference factors, and then process the first pixel water seepage feature enhancement map and the second pixel water seepage feature enhancement map by using the dam water seepage evaluation neural network, combine the enhanced water seepage characteristics in the two maps, and improve the accuracy of identifying the water seepage condition.

[0067] In the description of the embodiments of the application, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connecting" should be understood in a broad sense, for example, can be fixed connection, can also be detachable connection, or integrally connected, can be mechanical connection, can also be electrical connection, can be directly connected, can also be indirectly connected through an intermediate medium, and can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0068] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the application or the part of the application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of software products, and the computer software product is stored in a storage medium, including a plurality of instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in the embodiments of the application. The foregoing storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and various program code storage media.

[0069] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions recorded in the above embodiments can be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1.A dam water seepage monitoring method based on remote sensing images, characterized in that, The method comprises the following steps: Under daylight conditions, a plurality of remote sensing images of the same dam wall surface region at different time periods are collected, each remote sensing image is subjected to clustering processing, a plurality of connected regions are obtained, and background regions and contrast regions are distinguished; According to the color difference between the contrast regions and the background regions, a pixel water seepage feature value of each pixel point in the contrast regions is obtained, and a pixel water seepage feature map is obtained; On the plurality of pixel water seepage feature maps, the contrast regions belonging to the same water seepage target and having area overlap are found out; The water seepage diffusion coincidence degree of each contrast region in the pixel water seepage feature map is obtained, and the water seepage diffusion coincidence degrees of the contrast regions belonging to the same water seepage target are averaged; According to the contrast regions belonging to the same water seepage target, the water seepage change degree of each water seepage target is obtained; The water seepage diffusion coincidence degree average and the water seepage change degree are respectively used for enhancing processing on any one pixel water seepage feature map, and a first pixel water seepage feature enhanced map and a second pixel water seepage feature enhanced map are obtained; The dam water seepage evaluation neural network is used for processing the first pixel water seepage feature enhanced map and the second pixel water seepage feature enhanced map, and a water seepage score is obtained. 2.The dam water seepage monitoring method based on remote sensing images of claim 1, wherein, The process of obtaining the pixel water seepage feature map comprises: The average value of the brightness values of the background regions is obtained, and the background brightness is obtained; The ratio of the background brightness to the brightness value of each pixel point in the contrast regions is taken as the pixel brightness ratio; The channel difference is obtained by subtracting the channel value of each pixel point in the contrast regions from the background channel average; The channel difference values are obtained by adding the channel differences of the three channels after weighting; The pixel water seepage feature value of the same pixel point in the contrast regions is multiplied by the pixel brightness ratio, and the multiplication result is normalized, the pixel water seepage feature value of the pixel point in the background region is set to 0, and the pixel water seepage feature map is obtained. 3.The method of claim 2, wherein, The formula for obtaining the channel difference value is: ， Wherein, C i is the channel difference value of the i-th pixel point in the contrast area, △R i is the R channel difference of the i-th pixel point, △G i is the G channel difference of the i-th pixel point, and △B i is the B channel difference of the i-th pixel point. 4.The method of claim 1, wherein, The process of obtaining the water seepage diffusion coincidence degree comprises: A contrast region of the pixel water seepage feature map is divided into a plurality of N*M size image blocks, wherein N and M are positive integers; The average value of the pixel water seepage feature values of each image block is taken; The maximum pixel water seepage feature average value corresponding to the image block located in the interior of the contrast region is selected as the center image block; In the horizontal direction, the center image block is taken as the starting point, and the difference values of the pixel water seepage feature average values of the adjacent image blocks on the left and right sides are calculated; In the vertical direction, the center image block is taken as the starting point, and the difference values of the pixel water seepage feature average values of the adjacent image blocks on the upper side and the lower side are calculated; The ratio of the number of difference values greater than 0 to the total number of difference values is taken as the water seepage diffusion coincidence degree. 5.The dam water seepage monitoring method based on remote sensing images of claim 4, wherein, In the left direction of the center image block, the pixel water seepage feature average value of the image block on the right in the adjacent image blocks is subtracted from the pixel water seepage feature average value of the image block on the left, and the difference value is obtained; In the right direction of the center image block, the pixel water seepage feature average value of the image block on the left in the adjacent image blocks is subtracted from the pixel water seepage feature average value of the image block on the right, and the difference value is obtained; In the upper direction of the center image block, the pixel water seepage feature average value of the image block on the lower side in the adjacent image blocks is subtracted from the pixel water seepage feature average value of the image block on the upper side, and the difference value is obtained; In the lower side direction of the center image block, the pixel seepage feature mean of the upper image block in the adjacent image block is subtracted from the pixel seepage feature mean of the lower image block to obtain a difference value. 6.The method for monitoring dam seepage based on remote sensing image according to claim 1, characterized in that, The process of obtaining the seepage change degree of each seepage target includes: selecting the largest area contrast region and the smallest area contrast region in the contrast regions belonging to the same seepage target; subtracting the largest area contrast region from the smallest area contrast region to obtain a maximum area difference; normalizing the maximum area difference to obtain the seepage change degree of the seepage target. 7.The method of claim 1, wherein, The enhancement process includes: using the seepage diffusion coincidence degree mean of the same seepage target to enhance the contrast region of the same seepage target in the pixel seepage feature map to obtain a first pixel seepage feature enhancement map; using the seepage change degree of the same seepage target to enhance the contrast region of the same seepage target in the pixel seepage feature map to obtain a second pixel seepage feature enhancement map. 8.The method of claim 7, wherein, The formula of the enhancement process is: ， ， wherein V z1，i is the i-th enhanced value of the contrast region of the same water seepage target on the first pixel water seepage feature enhanced map, V i is the i-th pixel water seepage feature value of the contrast region of the same water seepage target on the pixel water seepage feature map, c is the water seepage diffusion coincidence mean value of the same water seepage target, V z2，i is the i-th enhanced value of the contrast region of the same water seepage target on the second pixel water seepage feature enhanced map, and ε is the water seepage change degree of the same water seepage target. 9.The dam seepage monitoring method based on remote sensing images of claim 1, wherein, The dam seepage evaluation neural network includes: a first parallel feature extraction unit, a second parallel feature extraction unit, a cross-branch gated interactive feature enhancement unit, a CNN unit and a fully connected layer; The input end of the first parallel feature extraction unit is used to input the first pixel seepage feature enhancement map; the input end of the second parallel feature extraction unit is used to input the second pixel seepage feature enhancement map; The first input end of the cross-branch gated interactive feature enhancement unit is connected with the output end of the first parallel feature extraction unit, the second input end thereof is connected with the output end of the second parallel feature extraction unit, and the output end thereof is connected with the input end of the CNN unit; the input end of the fully connected layer is connected with the output end of the CNN unit, and the output end thereof serves as the output end of the dam seepage evaluation neural network. 10.The dam water seepage monitoring method based on remote sensing images of claim 9, wherein, The cross-branch gated interactive feature enhancement unit includes: a first gating module, a second gating module, a multiplier M1, a multiplier M2 and an adder A1; The input end of the first gating module is connected with the first input end of the multiplier M1 and serves as the first input end of the cross-branch gated interactive feature enhancement unit; the input end of the second gating module is connected with the first input end of the multiplier M2 and serves as the second input end of the cross-branch gated interactive feature enhancement unit; the second input end of the multiplier M1 is connected with the output end of the first gating module, and the output end thereof is connected with the first input end of the adder A1; the second input end of the multiplier M2 is connected with the output end of the second gating module, and the output end thereof is connected with the second input end of the adder A1; the output end of the adder A1 serves as the output end of the cross-branch gated interactive feature enhancement unit.

Images