An Intelligent Identification Method for Underwater Dredging Pipelines Based on Multi-Temporal Remote Sensing Images

Abstract

The present invention discloses an intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images, including: obtaining multi-temporal remote sensing images of the same study area; extracting the blue band, green band and near-infrared band, or the blue band, green band and short-wave infrared band; calculating the optical remote sensing water body index and displaying it using a rainbow color map; determining the optimal threshold for water body segmentation of the optical remote sensing water body index; clarifying the water body and non-water body areas; using a Lee filter for image enhancement; using a custom convolution kernel for directional filtering enhancement; using a double-threshold sobel edge detection method to obtain an edge detection result map; superimposing the edge detection result map on the rainbow color map, patching the discontinuous parts at the middle position of the continuous line of the edge detection result map, and removing the debris non-segment edges to form a complete display map of the distribution of dredging pipelines; it can achieve efficient, large-scale scanning and precise positioning of the pipelines without disturbing the normal operation of the reservoir.

Description

Technical Field

[0001] The present invention relates to the technical field of dredging pipeline identification, and particularly to an intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images. Background Art

[0002] As an important water conservancy project facility, a reservoir undertakes multiple functions such as flood control, water supply, irrigation, and power generation. However, over time, the reservoir inevitably faces the problem of siltation. The deposition of sediment and organic matter will lead to a reduction in the reservoir capacity, affecting the exertion of its functions. In order to maintain the normal functions of the reservoir, regular dredging has become a necessary maintenance work. Accurately identifying the location of the dredging pipeline is crucial for understanding the dredging progress and the dredging location, which is not only related to the efficiency and safety of the dredging work but also directly affects the protection of the reservoir infrastructure.

[0003] Traditional pipeline identification methods mainly rely on manual detection or sonar technology. These methods are often time-consuming and laborious, and have limited accuracy in complex underwater environments. With the development of remote sensing technology, the water body identification method based on satellite remote sensing images provides new possibilities for reservoir management. Among them, the Normalized Difference Water Index (NDWI), as an effective water body extraction index, has shown great potential in water body boundary identification and underwater feature detection. The NDWI index is used to enhance the water body information in remote sensing images while suppressing vegetation and soil information. The calculation formula of NDWI is: NDWI = (Green - NIR) / (Green + NIR); where Green represents the reflectance of the green light band and NIR represents the reflectance of the near-infrared band. The NDWI value is usually between -1 and 1. Water bodies usually have positive values, while other features such as vegetation and soil usually have negative values. This index is mainly used for water body range identification. Different researchers have proposed various water body extraction methods based on the NDWI index, such as patents CN 201510272030, CN 201810520933.1, CN202011272866.X.

[0004] As a pipeline for sucking and transporting silt, the dredging pipeline usually extends from the bottom of the water to the water surface, spanning both the water and underwater environments, and needs to coordinate work between land and water. Considering the complexity of the environment and the special requirements for identification, innovative remote sensing technology is required to improve the accuracy and applicability for the identification of this specific object. Summary of the Invention

[0005] The present invention provides an intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images to solve the problem that the existing technology needs to be improved in terms of the accuracy and applicability of dredging pipeline identification.

[0006] The present invention provides an intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images, including:

[0007] Step 1: Obtain multi-temporal remote sensing images of the same study area;

[0008] Step 2: Extract the blue band, green band, and near-infrared band, or the blue band, green band, and short-wave infrared band of each multi-temporal remote sensing image;

[0009] Step 3: Calculate the optical remote sensing water index based on the extracted bands;

[0010] Step 4: Display the optical remote sensing water index using a rainbow color map;

[0011] Step 5: Determine the optimal threshold for water body segmentation of the optical remote sensing water index using the maximum inter-class variance method;

[0012] Step 6: Process the optical remote sensing water index according to the optimal threshold to clarify the water body and non-water body areas, and obtain the processed optical remote sensing water index image;

[0013] Step 7: Use the Lee filter to enhance the image of the processed optical remote sensing water index to remove noise and obtain the optical remote sensing water index image after noise filtering;

[0014] Step 8: Use a custom convolution kernel to perform directional filtering enhancement on the optical remote sensing water index image after noise filtering;

[0015] Step 9: Use the double-threshold sobel edge detection method for the optical remote sensing water index image after directional filtering enhancement to obtain the edge detection result map;

[0016] Step 10: Superimpose the edge detection result map on the rainbow color map, repair the discontinuous parts at the middle position of the continuous line of the edge detection result map, and remove the debris non-segment edges to form a complete display map of the distribution of dredging pipelines.

[0017] Further, in Step 1, the remote sensing images are sourced from MODIS, Landsat, Sentinel, or GF series. Download the cloud-free satellite remote sensing images for the required time period from the selected satellite data platform, and preprocess the remote sensing images in combination with the preset regional scope.

[0018] Further, in Step 1, the remote sensing images are sourced from Sentinel-2; in Step 2, extract the blue band, green band, and near-infrared band; or, in Step 1, the remote sensing images are sourced from Landsat-8; in Step 2, extract the blue band, green band, and short-wave infrared band.

[0019] Furthermore, in step three, for multi-temporal remote sensing images containing the green light band and the near-infrared band, calculate the NDWI index:

[0020] NDWI = (R(GREEN) - R(NIR)) / (R(GREEN) + R(NIR));

[0021] For multi-temporal remote sensing images containing the green light band and the short-wave infrared band, calculate the MNDWI index:

[0022] MNDWI = (R(GREEN) - R(MIR)) / (R(GREEN) + R(MIR));

[0023] In the above formulas, NDWI is the NDWI index, MNDWI is the MNDWI index, R(GREEN)- is the green light band, R(NIR) is the near-infrared band, and R(MIR) is the short-wave infrared band.

[0024] Furthermore, in step four, when displaying the NDWI index using a rainbow color map, display it according to the NDWI values from large to small using the rainbow color map.

[0025] Furthermore, in step five, the Otsu method is used to automatically determine the optimal threshold for dividing the image into two classes. For each threshold t, calculate the inter-class variance; the inter-class variance is the weighted sum of the variances of the foreground and the background, and the formula is:

[0026]

[0027] Among them, is the inter-class variance, and are the weights of the foreground and the background respectively, and are the average gray values of the foreground and the background respectively, and thus select the threshold t that maximizes the inter-class variance as the optimal threshold t1.

[0028] Furthermore, in step six, when processing the NDWI index, when NDWI ≥ t1, it is the water area, and process the area where NDWI < t1. Assign the value 0 to the area where NDWI < t1, so as to clarify the water area and the non-water area.

[0029] Furthermore, in step seven, the output pixel value of the Lee filter is calculated by the following formula:

[0030]

[0031] Among them, is the input pixel value, is the mean within the local window, W is the weight factor, defined as:

[0032]

[0033] where, is the variance within the local window, is the variance of the noise.

[0034] Furthermore, in step eight, by adjusting the parameters of the filter, the optimal kernel parameters are determined, and the optimal kernel parameters include the convolution template and the gradient direction of the convolution kernel.

[0035] Furthermore, step nine includes:

[0036] For the optically remote-sensed water body index image after directional filtering, the OTSU threshold t2 is calculated using the maximum inter-class variance method;

[0037] Use the sobel operator to perform edge detection on the optically remote-sensed water body index image after directional filtering enhancement. The convolution kernel in the horizontal direction is as follows:

[0038]

[0039] The convolution kernel in the vertical direction is as follows:

[0040]

[0041] By performing a convolution operation on the optically remote-sensed water body index image after directional filtering, the gradients of each pixel in the horizontal and vertical directions are calculated, and the final edge intensity is obtained by calculating through the following formula:

[0042] +

[0043] Based on the OTSU threshold t2, the low threshold T low [[ID=5३]]and the weight k of the high threshold T high are determined. Since t2 has effectively segmented the foreground (high-value pixels) and background (low-value pixels) of the image, the weight k of the high threshold is 1, that is, the high threshold T high = k * t2 = t2; At the same time, a low threshold lower than the high threshold is set to capture possible weak edges, which is obtained by multiplying the high threshold by the weight, T low = k * T high, in the example, the weight of the low threshold is k = 0.5; among them, the high threshold is used to determine strong edges, and any pixel above this threshold is considered an edge; the low threshold is used to connect edges. Any pixel below the high threshold but above the low threshold, if connected to a strong edge, is also considered an edge, and other pixels are suppressed as non-edges;

[0044] Output a binary image, where the pixel values are 0 or 1, representing non-edge and edge respectively.

[0045] The present invention has the following beneficial effects: An intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images of the present invention can accurately locate underwater dredging pipelines through satellite remote sensing technology, which can significantly improve the efficiency of dredging management work. Compared with the traditional manual inspection method, the satellite remote sensing identification technology can perform large-scale scanning and precise positioning efficiently and safely without interfering with the normal operation of the reservoir. By reducing the investment in manpower and equipment, the work efficiency and safety are greatly improved, the risk of personnel entering potential dangerous areas is reduced, the dredging work can be more accurate and efficient, and the interference to the reservoir ecosystem is minimized. Through accurate pipeline distribution data, it helps decision-makers more accurately determine the pipeline condition, facilitates the formulation of more precise dredging plans, makes up for the lack of management data, and provides valuable data support for the long-term management of the reservoir. Through the innovative application of satellite remote sensing technology in the field of reservoir management, the efficiency and accuracy of reservoir management are significantly improved, and more scientific and sustainable reservoir management is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is a flowchart of an intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images provided by the present invention;

[0048] Figure 2 is an RGB remote sensing image;

[0049] Figure 3 is an NDWI image after threshold processing;

[0050] Figure 4 is an NDWI image after Lee filtering;

[0051] Figure 5 is an NDWI image enhanced by directional filtering;

[0052] Figure 6 is a double-threshold sobel edge detection result image;

[0053] Figure 7 It is a superimposed image of the edge detection result and the NDWI rainbow-colored map;

[0054] Figure 8 It is the effect after superimposing the modified pipeline distribution map and the NDWI rainbow-colored map. Specific implementation manners

[0055] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. The technical solutions provided by the embodiments of the present invention will be described in detail below with reference to the drawings.

[0056] Please refer to Figures 1 to 8 , an underwater dredging pipeline intelligent recognition method based on multi-temporal remote sensing images provided by an embodiment of the present invention, includes:

[0057] Step 1: Obtain multi-temporal remote sensing images of the same study area.

[0058] Specifically, the source of the remote sensing images can be the Moderate-resolution Imaging Spectroradiometer (MODIS), Landsat or Sentinel satellites, or the GF series satellites. Download cloud-free satellite remote sensing images of the required time period from the selected satellite data platform, and perform conventional preprocessing such as cropping on the remote sensing images in combination with the preset regional scope.

[0059] Step 2: Extract the blue band, green band and near-infrared band, or the blue band, green band and short-wave infrared band of each multi-temporal remote sensing image.

[0060] Specifically, the extracted bands mainly depend on the satellite source. If the source of the remote sensing images is Sentinel-2, then extract the blue band, green band and near-infrared band; if the source of the remote sensing images is Landsat-8, then extract the blue band, green band and short-wave infrared band. The bands extracted in this step ensure that the combination of blue, green and near-infrared or short-wave infrared bands can be achieved.

[0061] Step 3: Calculate the optical remote sensing water body index according to the extracted bands.

[0062] For multi-temporal remote sensing images containing the green light band and the near-infrared band, calculate the NDWI index:

[0063] NDWI = (R(GREEN) - R(NIR)) / (R(GREEN) + R(NIR));

[0064] For multi-temporal remote sensing images containing the green light band and the short-wave infrared band, calculate the MNDWI index:

[0065] MNDWI = (R(GREEN) - R(MIR)) / (R(GREEN) + R(MIR));

[0066] In the above formulas, NDWI is the NDWI index, MNDWI is the MNDWI index, R(GREEN)- is the green light band, R(NIR) is the near-infrared band, and R(MIR) is the short-wave infrared band.

[0067] The idea of applying the NDWI index in the identification of dredging pipelines stems from the following observations. In the areas where the dredging pipelines are not covered by water, there are obvious spectral characteristics. In the areas covered by water, due to the special materials and shapes, they may still produce reflections and scatterings of incident light different from those of the surrounding water. This difference may be reflected in the NDWI index, thus providing the possibility for the development of remote sensing identification of dredging pipelines.

[0068] Step Four, display the optical remote sensing water index using a rainbow color map.

[0069] Taking the NDWI image as an example, when displaying the NDWI index using a rainbow color map, display it according to the decreasing NDWI values. The rainbow color map can better judge the local subtle differences and the changes in different value segments. Different values are reflected by gradient colors, making it possible to observe the distribution of specific values in space.

[0070] Step Five, use the Otsu method to determine the optimal threshold for water body segmentation of the optical remote sensing water index.

[0071] Specifically, the Otsu method is used to automatically determine the optimal threshold for dividing the image into two classes. For each possible threshold t, calculate the between-class variance; the between-class variance is the weighted sum of the variances of the foreground and the background, and the formula is:

[0072]

[0073] Among them, is the between-class variance, and are the weights of the foreground and the background respectively, and They are the average gray values of the foreground and background respectively, and thus the threshold t that maximizes the between-class variance is selected as the optimal threshold t1.

[0074] Step Six: Process the optical remote sensing water index according to the optimal threshold to clarify the water and non-water areas, and obtain the processed optical remote sensing water index image.

[0075] Specifically, usually according to the determined optimal threshold t1, the NDWI or MNDWI image can be segmented into water and non-water areas. Taking NDWI as an example, when NDWI ≥ t1, it is the water area, and the area where NDWI < t1 is processed. The area where NDWI < t1 is assigned a value of 0, and thus the water and non-water areas can be clarified.

[0076] Step Seven: Use the Lee filter to enhance the processed optical remote sensing water index image to remove noise and obtain the optical remote sensing water index image after noise filtering.

[0077] This method adjusts the pixel values through local statistical characteristics, smooths the uniform areas, and at the same time retains the edges and details of the image. A typical method of filtering using the local statistical characteristics of the image is based on the weighted average of the input pixel values and the local mean. Specifically, the output pixel value of the Lee filter is calculated by the following formula:

[0078]

[0079] where, is the input pixel value, is the mean within the local window, is the weight factor, defined as:

[0080]

[0081] where, is the variance within the local window, is the variance of the noise.

[0082] In the example, the size of the Lee filter is determined to be a 3×3 matrix.

[0083] Step Eight: Use a custom convolution kernel to perform directional filtering enhancement on the optical remote sensing water index image after noise filtering.

[0084] Highlight the edges and details with specific direction features in the image, which is beneficial to identifying the linear structures in the image. By adjusting the parameters of the filter, the optimal kernel parameters are determined, including the appropriate convolution template and the gradient direction of the convolution kernel. Among them, the edges changing along the specified direction in the image can be enhanced through the specified direction.

[0085] In the example, the appropriate size of the convolutional kernel is determined to be a 3×3 matrix, and the direction of the convolutional kernel is 30°. The convolutional kernel is as follows:

[0086]

[0087] Step Nine: For the optically remote-sensed water index image enhanced by directional filtering, use the double-threshold sobel edge detection method to obtain the edge detection result map.

[0088] Specifically, for the optically remote-sensed water index image after directional filtering, use the maximum inter-class variance method to calculate the OTSU threshold t2;

[0089] Use the sobel operator to perform edge detection on the optically remote-sensed water index image enhanced by directional filtering. The Sobel operator is a discrete differential operator mainly used to calculate the gradient magnitude and direction of an image, and highlights the regions with intensity changes in the image through convolution operations. The convolutional kernel in the horizontal direction is as follows:

[0090]

[0091] The convolutional kernel in the vertical direction is as follows:

[0092]

[0093] By performing convolution operations on the optically remote-sensed water index image after directional filtering, calculate the gradients of each pixel in the horizontal and vertical directions. The final edge intensity is obtained by calculating through the following formula:

[0094] +

[0095] Using the double-threshold detection technique, apply the double-threshold technique to the results of the Sobel edge detection to enhance the continuity and accuracy of the edges. Based on the OTSU threshold t2, determine the weights k of the low threshold T low and the high threshold T high Since t2 has effectively segmented the foreground (high-value pixels) and background (low-value pixels) of the image, the weight k of the high threshold is 1, that is, the high threshold T high =k*t2=t2; At the same time, set a low threshold lower than the high threshold to capture possible weak edges, which is obtained by multiplying the high threshold by the weight, T low =k*T high, in the example, the weight of the low threshold is k = 0.5; among them, the high threshold is used to determine strong edges, and any pixel above this threshold is considered an edge; the low threshold is used to connect edges. Any pixel below the high threshold but above the low threshold, if connected to a strong edge, is also considered an edge, and other pixels are suppressed as non-edges;

[0096] Output a binary image, where the pixel values are 0 or 1, representing non-edge and edge respectively.

[0097] In the example, the size of the determined Sobel convolution kernel is 3×3 pixels, and the weights of the low threshold and the high threshold are 0.6 and 1 respectively.

[0098] Step ten, superimpose the edge detection result map on the rainbow color map, repair the discontinuous parts at the middle position of the continuous line of the edge detection result map, and remove the debris non-segment edges to form a complete display map of the dredging pipeline distribution.

[0099] The present invention uses a remote sensing optical water body index, and other indices that can reflect characteristic differences obtained from optical or SAR images, etc. may also be used instead. For the image denoising method, other suitable denoising methods can be selected in combination with image features, not limited to the Lee method. For the determination of the threshold, the OTSU threshold can be used, or it can also be determined in combination with artificial image features. For the edge detection and recognition method, new image processing algorithms or machine learning models can also be introduced in combination with the image characteristics.

[0100] As can be seen from the above embodiments, the present invention is an efficient non-invasive recognition method, which uses remote sensing image processing algorithms to accurately identify dredging pipelines, without the need for on-site manual investigation, greatly improving the monitoring efficiency and coverage. This remote monitoring ability enables the monitoring of pipelines in large areas and difficult-to-reach regions. The present invention uses image processing algorithms to automatically identify dredging pipelines, reducing the subjectivity and errors of manual interpretation, and improving the accuracy and consistency of recognition. This automated processing method can quickly process a large amount of remote sensing data, significantly improving work efficiency. Compared with traditional on-site investigation methods, the present invention greatly reduces the manpower and time costs. By analyzing remote sensing images, the pipeline areas that need to be dredged can be quickly located, reducing unnecessary on-site investigations and optimizing resource allocation. By reducing the need for on-site investigations, this method minimizes the interference to the environment, especially having significant advantages in the application in ecologically sensitive areas. The method framework of the present invention has good scalability, and the recognition accuracy and efficiency can also be improved by introducing new image processing algorithms or machine learning models.

[0101] The above-described embodiments of the present invention do not constitute a limitation on the protection scope of the present invention.

Claims

1. An intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images, characterized in that, Including: Step 1: Obtain multi-temporal remote sensing images of the same study area; Step 2: Extract the blue, green, and near-infrared bands, or the blue, green, and short-wave infrared bands of each multi-temporal remote sensing image; Step 3: Calculate the optical remote sensing water index based on the extracted bands; Step 4: Display the optical remote sensing water index using a rainbow color map; Step 5: Determine the optimal threshold for water body segmentation of the optical remote sensing water index using the Otsu method; Step 6: Process the optical remote sensing water index according to the optimal threshold to clarify the water body and non-water body areas, and obtain the processed optical remote sensing water index image; Step 7: Perform image enhancement on the processed optical remote sensing water index image using the Lee filter to remove noise and obtain the optical remote sensing water index image after noise filtering; Step 8: Perform directional filtering enhancement on the optical remote sensing water index image after noise filtering using a custom convolution kernel; Step 9: Use the double-threshold sobel edge detection method on the optical remote sensing water index image after directional filtering enhancement to obtain the edge detection result map; Step 10: Superimpose the edge detection result map on the rainbow color map, repair the discontinuous parts at the middle position of the continuous line of the edge detection result map, and remove the debris non-segment edges to form a complete display map of the dredging pipeline distribution.

2. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 1, wherein, In Step 1, the source of the remote sensing image is MODIS, Landsat, Sentinel, or GF series. Download the cloud-free satellite remote sensing images of the required time period from the selected satellite data platform, and preprocess the remote sensing images in combination with the preset regional scope.

3. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 2, characterized in that, In Step 1, the source of the remote sensing image is Sentinel-2; in Step 2, extract the blue, green, and near-infrared bands; or, in Step 1, the source of the remote sensing image is Landsat-8; in Step 2, extract the blue, green, and short-wave infrared bands.

4. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 3, wherein, In Step 3, for multi-temporal remote sensing images containing the green and near-infrared bands, calculate the NDWI index: NDWI = (R(GREEN) - R(NIR)) / (R(GREEN) + R(NIR)); For multi-temporal remote sensing images containing the green and short-wave infrared bands, calculate the MNDWI index: MNDWI = (R(GREEN) - R(MIR)) / (R(GREEN) + R(MIR)); In the above formulas, NDWI is the NDWI index, MNDWI is the MNDWI index, R(GREEN) is the green band, R(NIR) is the near-infrared band, and R(MIR) is the short-wave infrared band.

5. The intelligent identification method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 4, characterized in that In Step 4, when displaying the NDWI index using a rainbow color map, display it using the rainbow color map according to the decreasing order of the NDWI value.

6. The intelligent recognition method of underwater dredging pipeline based on multi-temporal remote sensing images according to claim 5, characterized in that In step 5, the Otsu method is used to automatically determine the optimal threshold for dividing the image into two classes. For each threshold t , the variance between classes is calculated; the variance between classes is the weighted sum of the variances of the foreground and background, and the formula is: Among them, is the between-class variance, and are the weights of the foreground and background respectively, and are the average gray values of the foreground and background respectively, so as to select the threshold t that maximizes the between-class variance as the optimal threshold t1.

7. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 6, wherein, In Step 6, when processing the NDWI index, when NDWI ≥ t1, it is the water body area, process the area where NDWI < t1, and assign a value of 0 to the area where NDWI < t1 to clarify the water body and non-water body areas.

8. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 7, characterized in that, In step seven, the output pixel value of the Lee filter is calculated by the following formula: wherein, is the input pixel value, is the mean value within the local window, W is the weight factor, defined as: Among them, is the variance within the local window, is the variance of the noise.

9. The intelligent recognition method for underwater dredging pipelines based on multi-temporal remote sensing images according to claim 8, characterized in that, In Step 8, the optimal kernel parameters are determined by adjusting the parameters of the filter, and the optimal kernel parameters include the convolution template and the gradient direction of the convolution kernel.

10. The intelligent identification method of underwater dredging pipeline based on multi-temporal remote sensing images according to claim 9, characterized in that, Step 9 includes: For the optically remote-sensed water body index image after directional filtering, the OTSU threshold t2 is also calculated using the maximum inter-class variance method; Perform edge detection on the optically remote-sensed water body index image enhanced by directional filtering using the Sobel operator. The convolution kernel in the horizontal direction is as follows: Vertical convolutional kernel As follows: By performing a convolution operation on the optically remote-sensed water body index image after directional filtering, the gradients of each pixel in the horizontal and vertical directions are calculated, and the final edge intensity is obtained by calculating according to the following formula: + Based on the OTSU threshold t2, determine the low threshold T low and the high threshold T high and the weight k of the high threshold T high = k * t2 = t2; At the same time, set the low threshold lower than the high threshold. The low threshold is obtained by multiplying the high threshold by the weight, T low = k * T high ; Among them, the high threshold is used to determine strong edges, and any pixel higher than this threshold is considered an edge; the low threshold is used to connect edges. Any pixel lower than the high threshold but higher than the low threshold, if connected to a strong edge, is also considered an edge, and other pixels are suppressed as non-edges; Output a binary image, where the pixel values are 0 or 1, representing non-edge and edge respectively.

Images