A method for evaluating a restoration effect for an environmental restoration project

Abstract

The present application relates to the technical field of image processing, in particular to a kind of repair effect evaluation method for environmental remediation engineering, comprising: collecting groundwater image, and obtaining groundwater gray image by graying;Groundwater gray image is segmented to obtain pixel point to be handled, and obtain several windows of pixel point to be handled;According to the number of each window of pixel point to be handled and the number of abnormal pixel points in each window, the abnormal degree of the box chart of each window of pixel point to be handled is obtained, and the noise intensity of pixel point to be handled is obtained according to abnormal degree;According to noise intensity, noise intensity parameter sequence is obtained, and the preferred degree of each noise intensity parameter is obtained;According to preferred degree and noise intensity, groundwater remediation evaluation is completed.The present application reduces the influence that impurities and noise points in groundwater are more similar in gray performance in image, improves the accuracy of noise point identification, and further improves the robustness of system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and particularly relates to a repair effect evaluation method for environmental remediation engineering. BACKGROUND

[0002] Groundwater is an important part of water resources and an important source of agricultural irrigation, industrial and urban water use. With the rapid advancement of industrialization and urbanization, groundwater pollution is becoming more and more serious, and groundwater remediation is an important part of environmental engineering remediation.

[0003] In the prior art of groundwater remediation detection, the quality of the collected groundwater image is poor due to the environmental light of the image collection area, and the collected image often needs to be denoised to enhance the quality of the image. The box plot is a method for identifying noise points. This method identifies noise points by the size and distribution of pixel values in a local range, that is, the identification of pixel points with a gray value that deviates too much in a local range. Since the impurities and noise points in the groundwater have similar gray scale in the image, the noise determined by a single box plot may have limitations and randomness, resulting in deviations in noise point identification, and thus losing the details of the image in the denoising process, and affecting the accuracy and efficiency of subsequent groundwater remediation detection. SUMMARY

[0004] To solve the above problems, the present application provides a repair effect evaluation method for environmental remediation engineering.

[0005] The repair effect evaluation method for environmental remediation engineering of the present application adopts the following technical scheme:

[0006] One embodiment of the present application provides a repair effect evaluation method for environmental remediation engineering, which comprises the following steps:

[0007] Collecting a groundwater image and obtaining a groundwater gray scale image by gray scale conversion;

[0008] Segmenting the groundwater gray scale image to obtain a plurality of to-be-processed pixel points; taking any one to-be-processed pixel point as a target to-be-processed pixel point; constructing a window for the target to-be-processed pixel point and translating the window to obtain a plurality of windows for the target to-be-processed pixel point;

[0009] According to the windows of the target to-be-processed pixel point, obtaining the box plot of each window of the target to-be-processed pixel point and the number of abnormal pixel points in each window; according to the box plot of each window of the target to-be-processed pixel point and the number of abnormal pixel points in each window, obtaining the abnormality degree of the target to-be-processed pixel point in the box plot of each window; according to the abnormality degree and the distance between the target to-be-processed pixel point and the center pixel point of the window, obtaining the noise intensity of the target to-be-processed pixel point;

[0010] According to the noise intensity parameter sequence, a first set and a second set are obtained; according to the distance between the noise intensity corresponding to the pixel points in the first set, the regularity of the first set is obtained; the regularity of the second set is obtained; and the preferred degree of each noise intensity parameter is obtained according to the regularity of the first set and the regularity of the second set.

[0011] According to the preferred degree and the noise intensity, a third set and a fourth set are obtained; according to the third set, the fourth set and the regularity of the first set, a plurality of noise pixel points and a plurality of impurity pixel points of the groundwater grayscale image are obtained; and a denoised groundwater grayscale image is obtained according to the groundwater grayscale image and the noise pixel points.

[0012] The repaired groundwater image is collected, and the groundwater repair evaluation is completed according to the denoised groundwater grayscale image.

[0013] Further, the box plot of each window of the target pixel point to be processed and the number of abnormal pixel points in each window are obtained according to the window of the target pixel point to be processed, and the specific steps include the following:

[0014] Any window of the target pixel point to be processed is denoted as a first window; the grayscale values of all pixel points in the first window are arranged in ascending order to obtain a sequence, denoted as a first sequence, and the first sequence is input into a box plot algorithm to obtain a box plot of the first window, denoted as a first window box plot, and the number of abnormal pixel points in the first window is obtained according to the first window box plot.

[0015] Further, the abnormal degree of the target pixel point to be processed in the box plot of each window is obtained according to the box plot of each window of the target pixel point to be processed and the number of abnormal pixel points in each window, and the specific steps include the following:

[0016]

[0017] In the formula, is the number of abnormal pixel points in the first window, is the grayscale value of the target pixel point to be processed, is the median of the first window box plot, is the absolute value, is the exponential function with a natural constant as the base, is the abnormal degree of the target pixel point to be processed in the first window box plot.

[0018] Furthermore, the specific steps for obtaining the noise intensity of the target pixel based on the degree of anomaly and the distance between the target pixel and the center pixel of the window are as follows:

[0019]

[0020] In the formula, For the target pixel to be processed at the 1st The degree of anomaly in the box plot of each window. The number of windows containing the target pixels to be processed; For the target pixel to be processed The Euclidean distance between the center pixel of the window and the target pixel to be processed; It is an exponential function with the natural constant as its base; The variance of the anomaly degree of the target pixel across all windows in the box plot. For normalization function, The noise intensity of the target pixel to be processed.

[0021] Furthermore, the specific steps for obtaining the noise intensity parameter sequence based on the noise intensity are as follows:

[0022] Obtain the noise intensity of each pixel to be processed, and find the maximum and minimum noise intensity values. Record the minimum noise intensity as the minimum noise intensity and the maximum noise intensity as the maximum noise intensity. Starting from the minimum noise intensity, increase the noise intensity each time. , The minimum noise intensity is set to a third value, and the noise intensity is increased until it is greater than or equal to the maximum noise intensity for the first time. Several noise intensity parameters are obtained. The minimum noise intensity, all noise intensity parameters and the maximum noise intensity are arranged in ascending order to obtain a sequence, which is denoted as the noise intensity parameter sequence.

[0023] Furthermore, the specific steps for obtaining the first set and the second set based on the noise intensity parameter sequence are as follows:

[0024] Let any one of the noise intensity parameters in the noise intensity parameter sequence be denoted as the target noise intensity parameter; obtain the noise intensity of each pixel to be processed, and denote the set of all noise intensities less than or equal to the target noise intensity parameter as the first set, and the set of all noise intensities greater than the target noise intensity parameter as the second set.

[0025] Furthermore, the specific steps for obtaining the regularity of the first set based on the distance between the noise intensity and the pixels to be processed in the first set are as follows:

[0026]

[0027] wherein, is the regularity of the first set; is the number of noise intensities in the first set, is the number of combinations of any two noise intensities in the first set; is the first combination, and the noise intensity corresponding to one of the noise intensities in the first combination is denoted as a first pixel point; combination is denoted as a second pixel point; is the Euclidean distance between the first pixel point and the second pixel point; is the variance of the Euclidean distances between the noise intensities in the first set; is an exponential function with a natural constant as the base.

[0028] Further, the preferred degree of each noise intensity parameter is obtained according to the regularity of the first set and the regularity of the second set, including the following specific steps:

[0029]

[0030] wherein, is the preferred degree of the target noise intensity parameter; is the regularity of the first set; is the regularity of the second set; is the absolute value.

[0031] Further, the noise pixel points and the impurity pixel points of the groundwater grayscale image are obtained according to the regularity of the third set, the regularity of the fourth set, and the regularity of the first set, including the following specific steps:

[0032] The regularity of the third set is obtained, the regularity of the fourth set is obtained, and in the regularity of the third set and the regularity of the fourth set, the noise intensity corresponding to the noise pixel point in the set with the smallest regularity is taken as the noise intensity, and the noise intensity corresponding to the impurity pixel point in the set with the largest regularity is taken as the noise intensity.

[0033] Further, the groundwater repair evaluation is completed according to the denoised groundwater grayscale image, including the following specific steps:

[0034] The difference value of the number of impurity pixel points in the denoised groundwater grayscale image and the repaired groundwater grayscale image is obtained, the ratio of the difference value to the total number of pixel points is compared with the empirical threshold value, and the groundwater repair evaluation is completed.

[0035] The beneficial effects of the technical scheme of the present application are as follows: after the groundwater gray image is collected, a plurality of to-be-processed pixel points are obtained by segmenting the groundwater gray image, the to-be-processed pixel points include pixel points corresponding to groundwater impurities and noise pixel points, the noise intensity of the to-be-processed pixel points is obtained by analyzing the abnormal degree of the to-be-processed pixel points in the box plot of different windows, the noise characteristics of the to-be-processed pixel points are better reflected, the influence of the impurities and the noise points on the gray display in the image is reduced, the subsequent image denoising effect is better, and then the groundwater is effectively and intuitively repaired, the work efficiency is improved and the labor cost is reduced, and then the pixel points in the groundwater gray image are classified into two sets by the noise intensity, the noise pixel points and the impurity pixel points in the groundwater gray image are more accurately distinguished by analyzing different noise intensity parameters as classification thresholds, that is, the optimization degree of the noise intensity parameters, a plurality of noise pixel points and a plurality of impurity pixel points of the groundwater gray image are obtained, and the contingency of the noise degree determination of the traditional single box plot is effectively solved. Meanwhile, the threshold of the adaptive noise degree is utilized according to the distribution characteristics of the impurities and the noise points, the robustness of the system is improved, the noise points are identified, the detail information of the image itself is retained, the groundwater repair detection is more intuitive, and the work efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0037] Figure 1 The step flow chart of the repair effect evaluation method for the environmental remediation engineering provided by an embodiment of the present application is shown in the figure.

[0038] Figure 2 The groundwater gray image of the repair effect evaluation method for the environmental remediation engineering provided by an embodiment of the present application is shown in the figure.

[0039] Figure 3 The segmentation result graph of the repair effect evaluation method for the environmental remediation engineering provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0040] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the following describes in detail the specific implementation, structure, features and effects of a repair effect evaluation method for environmental remediation engineering according to the present application, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0042] The specific scheme of the repair effect evaluation method for environmental remediation engineering provided by the present application is described in detail below in combination with the accompanying drawings.

[0043] Please refer to Figure 1 , which shows the step flowchart of the repair effect evaluation method for environmental remediation engineering provided by one embodiment of the present application. The method includes the following steps:

[0044] Step S001, collect the underground water image and obtain the underground water gray image by gray scale.

[0045] It should be noted that the noise degree of the pixel point is obtained by combining the abnormal performance of a single pixel point in multiple box plots, and the noise points are accurately identified by the preferred degree under each noise degree threshold, so that interpolation denoising is performed to enhance the image, which is convenient for subsequent underground water remediation detection. Before starting the analysis, the image needs to be collected first.

[0046] Specifically, the image in the underground water is collected by a high-resolution camera, denoted as an underground water image; the underground water image is processed by gray scale to obtain a gray image, denoted as an underground water gray image; it should be noted that the underground water image is an RGB image, please refer to Figure 2 , Figure 2 The underground water gray image of the present embodiment, Figure 2 contains impurities and a plurality of noise points in the underground water, and the noise points are smaller than the impurities.

[0047] Thus, the underground water gray image is obtained.

[0048] Step S002, segment the underground water gray image to obtain the pixel points to be processed, and obtain a plurality of windows of the pixel points to be processed.

[0049] It should be noted that because the abnormality determined by a single box plot is accidental and the noise is smaller than the impurities, the noise degree of the pixel points can be obtained by synthesizing a plurality of box plots containing the pixel points. Meanwhile, after obtaining the noise degree of the pixel points, a specific noise degree threshold is determined by analyzing the distribution characteristics of the noise and the impurities to distinguish the noise and the impurities.

[0050] It should be noted that because the noise and the impurities can both be bright spots or dark spots in the underground water image, which have a large difference from the background gray value, the Otsu method can be used to segment the impurities and the noise, and then only the segmented points are analyzed for the noise degree, so as to reduce the calculation amount.

[0051] Specifically, the underground water gray image is segmented to obtain a plurality of pixel points to be processed, as follows:

[0052] The underground water gray image is subjected to Otsu threshold segmentation to obtain an optimal segmentation threshold, the pixel points with a gray value less than the optimal segmentation threshold in the underground water gray image are set to 0, the pixel points with a gray value greater than or equal to the optimal segmentation threshold in the underground water gray image are set to 1, a segmentation result image is obtained, and the pixel points with a gray value of 0 in the segmentation result image are taken as the pixel points to be processed; it should be noted that the pixel points to be processed include the pixel points corresponding to the noise and the impurities, please refer to Figure 3 , Figure 3 is the segmentation result image of the present embodiment.

[0053] It should be noted that the above-mentioned plurality of pixel points to be processed are obtained by constructing different windows for a single pixel point, analyzing the abnormality degree weighted average of the current pixel point in a plurality of windows to determine the noise points, the abnormality degree of the window is obtained by the deviation degree of the current pixel point and the number of abnormal pixel points, and the weight is the distance of the current pixel point from the center of the window.

[0054] Specifically, any one of the pixel points to be processed is taken as a target pixel point to be processed, a window is constructed for the target pixel point to be processed and the window is translated to obtain a plurality of windows of the target pixel point to be processed, as follows:

[0055] In the underground water gray image, a window of is constructed with the target pixel point to be processed as the center, which is taken as the target window of the target pixel point to be processed; the target window is translated times along the horizontal right direction, one pixel point each time, to obtain windows of the target pixel point to be processed; is a preset first value, is a preset second value, and , narrate; translate the target window in a horizontal left direction times, each time shifting one pixel point, to obtain target pixel points to be processed n windows; translate the target window in a vertical upward direction times, each time shifting one pixel point, to obtain target pixel points to be processed n windows; translate the target window in a vertical downward direction times, each time shifting one pixel point, to obtain target pixel points to be processed n windows.

[0056] It should be noted that the construction of the target window or the translation of the target window may exceed the boundary of the underground water grayscale image, at this time, the embodiment uses the method of quadratic linear interpolation to interpolate the data of the part of the underground water grayscale image exceeding the boundary.

[0057] At this point, a plurality of windows of the target pixel point to be processed are obtained.

[0058] Step S003, according to the box plot of each window of the pixel point to be processed and the number of abnormal pixel points in each window, the abnormal degree of the box plot of each window of the pixel point to be processed is obtained, and the noise intensity of the pixel point to be processed is obtained according to the abnormal degree.

[0059] It should be noted that because the noise point is usually a relatively isolated point, the number of abnormal points (points outside the upper and lower boundaries of the box plot) in the box plot constructed in its window range is small, and the degree of deviation from the median is large, so the abnormal degree in a window can be represented by the number of abnormal points in the box plot and the degree of deviation of the current pixel point grayscale value from the median in the box plot, and the logical relationship is that the fewer the number of abnormal points, the greater the degree of deviation, and the higher the abnormal degree.

[0060] Specifically, according to the window of the target pixel point to be processed, the box plot of each window of the target pixel point to be processed and the number of abnormal pixel points in each window are obtained, which is as follows:

[0061] Any one window of the target pixel point to be processed is denoted as a first window; the grayscale values of all pixel points in the first window are arranged in ascending order to obtain a sequence, denoted as a first sequence, and the first sequence is input into a box plot algorithm to output a box plot, denoted as a box plot of the first window, and the number of abnormal pixel points in the first window is obtained according to the box plot of the first window; it should be noted that the first sequence is input into the box plot algorithm to output a box plot, and the number of abnormal pixel points in the first window is obtained according to the box plot of the first window, which is an existing method of the box plot algorithm, and the embodiment will not be repeated.

[0062] Furthermore, based on the box plot of each window of the target pixel to be processed and the number of abnormal pixels in each window, the degree of abnormality of the box plot of the target pixel to be processed in each window is obtained, as follows:

[0063]

[0064] In the formula, This represents the number of abnormal pixels in the first window. The grayscale value of the target pixel to be processed. The median of the box plot in the first window. To take the absolute value, This embodiment uses an exponential function with the natural constant as its base. The model is used to represent the inverse proportional relationship. As input to the model, implementers can set an inverse proportional function according to the specific implementation situation. The degree of anomaly of the target pixel in the box plot of the first window.

[0065] It should be noted that, The box plot in the first window indicates the degree of deviation of the target pixel to be processed. The fewer abnormal pixels there are in the first window, the greater the degree of deviation of the target pixel to be processed, and the higher the degree of abnormality of the target pixel to be processed in the box plot in the first window.

[0066] It should be noted that since the pixels in different box plots are different, that is, the distance between different box plots and the target pixel to be processed is different, the importance of box plots at different distances is different. The logical relationship is that the closer the distance, the higher the importance, that is, the higher the weight. At the same time, for noise, it is usually an isolated point. Therefore, the more stable the anomalous points in its surrounding window are, the better. That is, the more stable the degree of anomalousness of the surrounding box plot is, the more isolated the pixel is, and the higher the probability that it is a noise pixel.

[0067] Specifically, the noise intensity of the target pixel is obtained based on the degree of anomaly and the distance between the target pixel and the center pixel of the window, as follows:

[0068]

[0069] In the formula, For the target pixel to be processed in the th The degree of anomaly in the box plot of each window. The number of windows representing the target pixels to be processed; it should be noted that in this embodiment, the number of windows representing the target pixels to be processed is 13, that is, the target window and 12 windows that are translated a total of 12 times in four directions respectively; For the target pixel to be processed the Euclidean distance between the center pixel point of the window and the target pixel point to be processed; For the exponential function with a natural constant as the base, the embodiment adopts a model to present an inverse proportional relationship, The input of the model, the implementer can set the inverse proportional function according to the specific implementation; the variance of the abnormal degree of the target pixel point to be processed in the box plot of all windows, the linear normalization function, the object of normalization is the , the noise intensity of the target pixel point to be processed.

[0070] It should be noted that, the mean value of the weighted abnormal degree of the target pixel point to be processed in the box plot of all windows, that is, the smaller the Euclidean distance between the center pixel point of the window and the target pixel point to be processed, the greater the weight, that is, the smaller the corresponding abnormal degree reference; the smaller the variance of the abnormal degree of the target pixel point to be processed in the box plot of all windows, the more stable the abnormal degree of the target pixel point to be processed in the box plot of different windows, indicating that the pixel point is more isolated, and the greater the noise intensity of the target pixel point to be processed.

[0071] At this point, the noise intensity of the target pixel point to be processed is obtained.

[0072] Step S004, obtaining a noise intensity parameter sequence according to the noise intensity, and obtaining the preferred degree of each noise intensity parameter.

[0073] It should be noted that, because the distribution of impurities in groundwater is usually in the trend of aggregation or diffusion (regularity), and noise points are usually randomly distributed (irregularity), therefore, the embodiment divides the above-mentioned several pixel points to be processed into two categories by iterating the noise degree threshold, and realizes adaptive threshold by calculating the preferred degree, so as to identify the noise points from the pixel points to be processed.

[0074] Specifically, the noise intensity parameter sequence is obtained according to the noise intensity, and the specific process is as follows:

[0075] The noise intensity of each pixel point to be processed is obtained, the maximum value and the minimum value of the noise intensity are obtained, the minimum value of the noise intensity is recorded as the minimum noise intensity, and the maximum value of the noise intensity is recorded as the maximum noise intensity; starting from the minimum noise intensity, each time , is a preset third numerical value, and the embodiment takes The process continues until the noise intensity is greater than or equal to the maximum noise intensity for the first time, resulting in several noise intensity parameters. The minimum noise intensity, all noise intensity parameters, and the maximum noise intensity are then arranged in ascending order to obtain a sequence, which is denoted as the noise intensity parameter sequence. It should be noted that the noise intensity parameter sequence contains several noise intensity parameters, and the minimum and maximum noise intensities are also noise intensity parameters.

[0076] It should be noted that each noise intensity parameter in the noise intensity parameter sequence can divide the noise intensity of all pixels to be processed into two categories, representing noise and impurities. For a good selection threshold, the more irregular the pixel position distribution in the noise category, the better, while the more regular the pixel distribution in the impurity category, the better. In this embodiment, the average distance between any two pixels in a pixel category is used to characterize the pixel clustering trend, where a smaller average value indicates greater clustering. Since pixels in the impurity category may also exhibit diffusion values, using only the average distance as a regularity criterion will have errors, while the distribution of noise is random, meaning the distance between its pixels is relatively discrete. In summary, the average distance and the dispersion of pixel distances can be used together to characterize the regularity of pixels. The smaller the average value and the smaller the dispersion, the stronger the regularity, and the higher the probability that the category belongs to the impurity category; conversely, the same applies. Furthermore, the difference in regularity between the two categories characterizes the selection degree. The larger the difference between the two categories, the better the current noise intensity parameter separates noise and impurities, and the higher its selection degree.

[0077] Specifically, the first set and the second set are obtained based on the noise intensity parameter sequence, as follows:

[0078] Let any one of the noise intensity parameters in the noise intensity parameter sequence be denoted as the target noise intensity parameter; obtain the noise intensity of each pixel to be processed, and denote the set of all noise intensities less than or equal to the target noise intensity parameter as the first set, and the set of all noise intensities greater than the target noise intensity parameter as the second set.

[0079] Furthermore, based on the distance between the noise intensity and the pixels to be processed in the first set, the regularity of the first set is obtained, as follows:

[0080]

[0081] In the formula, The regularity of the first set; The number of noise intensities in the first set. The number of combinations of any two noise intensities in the first set; For the first The pixel point to be processed corresponding to one of the noise intensities in the combination is denoted as the first pixel point; For the first The pixel point to be processed corresponding to the other noise intensity in the combination is denoted as the second pixel point; The Euclidean distance between the first pixel and the second pixel; The variance of the Euclidean distance between the pixels to be processed corresponding to all noise intensities in the first set; This embodiment uses an exponential function with the natural constant as its base. The model is used to represent the inverse proportional relationship. As input to the model, implementers can set an inverse proportional function according to the specific implementation situation.

[0082] It should be noted that, This represents the average distance between the pixels to be processed corresponding to the noise intensity in the first set, characterizing the clustering trend of the pixels. The smaller the average value and the smaller the dispersion, the stronger the regularity.

[0083] Furthermore, based on the distance between the noise intensity and the pixels to be processed in the second set, the regularity of the second set is obtained, as follows:

[0084]

[0085] In the formula, The regularity of the second set; The number of noise intensities in the second set. The number of combinations of any two noise intensities in the second set; For the first The pixel point to be processed corresponding to one of the noise intensities in the combination is denoted as the third pixel point. For the first The pixel point to be processed corresponding to the other noise intensity in the combination is denoted as the fourth pixel point; The Euclidean distance between the third and fourth pixels; The variance of the Euclidean distance between the pixels to be processed corresponding to all noise intensities in the second set; This embodiment uses an exponential function with the natural constant as its base. The model is used to represent the inverse proportional relationship. As input to the model, implementers can set an inverse proportional function according to the specific implementation situation.

[0086] Furthermore, based on the regularities of the first set and the second set, the optimality of each noise intensity parameter is obtained, as follows:

[0087]

[0088] In the formula, a preferred degree of the target noise intensity parameter; regularity of the first set; regularity of the second set; taking an absolute value.

[0089] At this point, the preferred degree of the target noise intensity parameter is obtained.

[0090] Step S005, obtaining noise pixel points and impurity pixel points of the groundwater grayscale image according to the preferred degree and the noise intensity, obtaining the groundwater grayscale image after denoising, and completing the groundwater repair evaluation according to the impurity pixel points and the groundwater grayscale image after denoising.

[0091] It should be noted that the preferred degree is represented by the difference between the regularities of the two sets, and the greater the difference between the regularities of the two sets, the better the current noise intensity parameter separates the noise points and impurities, and the higher the preferred degree as the threshold.

[0092] Further, the third set and the fourth set are obtained according to the preferred degree and the noise intensity, as follows:

[0093] The preferred degree of each noise intensity parameter is obtained, the noise intensity parameter corresponding to the maximum preferred degree is taken as the best threshold, the noise intensity of each pixel point to be processed is obtained, a set composed of all noise intensities less than or equal to the best threshold is recorded as the third set, and a set composed of all noise intensities greater than the best threshold is recorded as the fourth set.

[0094] It should be noted that the noise degree corresponding to the maximum preferred degree is the best threshold, and under this threshold, the noise points and impurities are best separated, and then the regularity of the two types of pixel points under the best threshold is calculated, and the type with the smaller regularity is the noise point type, thereby completing the identification of the noise points.

[0095] Specifically, according to the regularity of the third set, the fourth set and the first set, a plurality of noise pixel points and a plurality of impurity pixel points of the groundwater grayscale image are obtained, as follows:

[0096] The regularity of the third set is obtained, the regularity of the fourth set is obtained, and in the regularity of the third set and the regularity of the fourth set, the pixel points to be processed corresponding to the noise intensity in the set with the smallest regularity are taken as the noise pixel points, and the pixel points to be processed corresponding to the noise intensity in the set with the largest regularity are taken as the impurity pixel points; it should be noted that the method of obtaining the regularity of the third set, the regularity of the fourth set and the regularity of the first set is the same, and this embodiment will not be described again.

[0097] Further, the groundwater grayscale image after denoising is obtained according to the groundwater grayscale image and the noise pixel points, as follows:

[0098] Any one noise pixel point is recorded as a target noise pixel point, in the groundwater gray image, the gray value of the four neighborhood pixel points of the target noise pixel point is obtained by bilinear interpolation method, and the gray value of the center pixel point of the four neighborhood pixel points is recorded as the first gray value, the gray value of the target noise pixel point is replaced by the first gray value, and the gray value of each noise pixel point is replaced, and finally the denoising groundwater gray image is obtained.

[0099] Further, the groundwater repair evaluation is completed according to the denoising groundwater gray image, as follows:

[0100] By detecting the impurity content in the groundwater, adding appropriate adsorbent or activated carbon adsorbent material according to the impurity content in the groundwater, adsorbing the organic matter and heavy metal ions in the groundwater to the surface of the material, thereby removing the impurities, and then collecting the groundwater gray image after removing the impurities and carrying out the above denoising, obtaining the repaired groundwater gray image, comparing the difference between the number of impurity pixel points in the denoised groundwater gray image and the repaired groundwater gray image, thereby effectively and intuitively repairing the groundwater, improving the work efficiency and reducing the labor cost. The experience threshold value of the embodiment is 0.8, when the ratio of the number of differences of the impurity pixel points to the total number of points is less than the experience threshold value, it is considered that the repair effect is poor at this time.

[0101] It should be noted that the method of obtaining the new denoising groundwater gray image and the denoising groundwater gray image is the same, and the embodiment will not be repeated, and the impurity pixel point in the denoising groundwater gray image is the impurity pixel point of the groundwater gray image.

[0102] At this point, by denoising the groundwater gray image, the groundwater repair evaluation is completed.

[0103] The above only describes the preferred embodiments of the present application, and does not limit the present application, any modification, equivalent replacement, improvement, etc. within the principles of the present application, should be included in the protection scope of the present application.

Claims

1. A method for evaluating a remediation effect for environmental remediation engineering, characterized by, The method comprises the following steps: Collecting a groundwater image, and obtaining a groundwater grayscale image through grayscale processing; Segmenting the groundwater grayscale image to obtain a plurality of pixel points to be processed; taking any one of the pixel points to be processed as a target pixel point to be processed; constructing a window for the target pixel point to be processed and translating the window to obtain a plurality of windows for the target pixel point to be processed; According to the windows for the target pixel point to be processed, obtaining a box plot for each window of the target pixel point to be processed and the number of abnormal pixel points in each window; according to the box plot for each window of the target pixel point to be processed and the number of abnormal pixel points in each window, obtaining an abnormality degree of the target pixel point to be processed in the box plot for each window; according to the abnormality degree and the distance between the target pixel point to be processed and the center pixel point of the window, obtaining a noise intensity of the target pixel point to be processed; According to the noise intensity, obtaining a noise intensity parameter sequence, wherein the noise intensity parameter sequence comprises a plurality of noise intensity parameters; obtaining a first set and a second set according to the noise intensity parameter sequence; obtaining regularity of the first set according to the distance between the target pixel points corresponding to the noise intensity in the first set; obtaining regularity of the second set; obtaining an optimal degree of each noise intensity parameter according to the regularity of the first set and the regularity of the second set; According to the optimal degree and the noise intensity, obtaining a third set and a fourth set; according to the third set, the fourth set and the regularity of the first set, obtaining a plurality of noise pixel points and a plurality of impurity pixel points of the groundwater grayscale image; obtaining a denoised groundwater grayscale image according to the groundwater grayscale image and the noise pixel points; Collecting a repaired groundwater image, and completing groundwater repair evaluation according to the denoised groundwater grayscale image; The calculation formula of the abnormality degree is: In the formula, is the number of abnormal pixel points in the first window, is the gray value of the target pixel point to be processed, is the median of the box plot of the first window, is the absolute value, is the exponential function with a natural constant as the base, is the abnormality degree of the target pixel point to be processed in the box plot of the first window. The calculation formula of the noise intensity is: In the formula, is the abnormal degree of the target pixel point in the box plot of the first window, is the abnormal degree of the target pixel point in the box plot of the first window, is the number of windows of the target pixel point; is the center pixel point of the first window of the target pixel point, is the Euclidean distance between the center pixel point of the first window of the target pixel point and the target pixel point; is an exponential function with a natural constant as the base; is the variance of the abnormal degree of the target pixel point in the box plot of all windows, is a normalization function, is the noise intensity of the target pixel point; The calculation formula of the optimal degree is: wherein is a preferred degree of the target noise intensity parameter; is regularity of the first set; is regularity of the second set; is taking an absolute value; The calculation formula of the regularity of the first set is: In the formula, regularity of the first set; number of noise intensities in the first set, combination number of any two noise intensities in the first set; the first combination, denoted as a first pixel point, corresponding to one of the noise intensities in the first combination, denoted as a second pixel point, corresponding to the other of the noise intensities in the first combination; Euclidean distance between the first pixel point and the second pixel point; variance of Euclidean distances between pixel points corresponding to all noise intensities in the first set; exponential function with a natural constant as a base; The method for obtaining the plurality of noise pixel points and the plurality of impurity pixel points of the groundwater grayscale image is: obtaining the regularity of the third set, obtaining the regularity of the fourth set, taking the target pixel points corresponding to the noise intensity in the set with the smallest regularity as the noise pixel points, and taking the target pixel points corresponding to the noise intensity in the set with the largest regularity as the impurity pixel points; The method for obtaining the denoised groundwater grayscale image is: taking any one of the noise pixel points as a target noise pixel point, obtaining the gray value of the center pixel point of the four neighborhood pixel points of the target noise pixel point in the groundwater grayscale image through bilinear interpolation according to the gray values of the four neighborhood pixel points, taking the gray value as a first gray value, replacing the gray value of the target noise pixel point with the first gray value, and finally obtaining the denoised groundwater grayscale image through replacement of the gray values of each noise pixel point. The groundwater repair evaluation is completed according to the denoised groundwater gray image, including: obtaining the difference value of the number of impurity pixel points in the denoised groundwater gray image and the repaired groundwater gray image, comparing the ratio of the difference value to the total number of pixel points with an empirical threshold, and completing the groundwater repair evaluation. The noise intensity parameter sequence is obtained according to the noise intensity, comprising: obtaining the noise intensity of each pixel point to be processed, obtaining the maximum value and the minimum value of the noise intensity, taking the minimum value of the noise intensity as the minimum noise intensity, and taking the maximum value of the noise intensity as the maximum noise intensity; starting from the minimum noise intensity, adding a third number each time , is a preset third number, until the noise intensity is greater than or equal to the maximum noise intensity for the first time, obtaining a plurality of noise intensity parameters, arranging the minimum noise intensity, all noise intensity parameters and the maximum noise intensity in ascending order to obtain a sequence, and taking the sequence as a noise intensity parameter sequence. The first set and the second set are obtained according to the noise intensity parameter sequence, including: taking any noise intensity parameter in the noise intensity parameter sequence as a target noise intensity parameter; obtaining the noise intensity of each pixel point to be processed, and taking the set of all noise intensities less than or equal to the target noise intensity parameter as the first set, and taking the set of all noise intensities greater than the target noise intensity parameter as the second set. The third set and the fourth set are obtained by: obtaining the optimization degree of each noise intensity parameter, taking the noise intensity parameter corresponding to the maximum optimization degree as the best threshold, obtaining the noise intensity of each pixel point to be processed, taking the set of all noise intensities less than or equal to the best threshold as the third set, and taking the set of all noise intensities greater than the best threshold as the fourth set.

2. The method for evaluating a restoration effect according to claim 1, wherein The box plot of each window of the target pixel point to be processed and the number of abnormal pixel points in each window are obtained according to the window of the target pixel point to be processed, including the following specific steps: Taking any window of the target pixel point to be processed as a first window; arranging the gray values of all pixel points in the first window in ascending order to obtain a sequence, denoted as a first sequence, inputting the first sequence into a box plot algorithm to obtain a box plot, denoted as a box plot of the first window, and obtaining the number of abnormal pixel points in the first window according to the box plot of the first window.

Images