Heating pipe network image edge detection method and system based on wavelet transformation

Abstract

The embodiment of the invention provides a heating pipe network image edge detection method and system based on wavelet transformation, and belongs to the technical field of heating pipe network image recognition. The edge detection method comprises the following steps: acquiring a heat supply pipe network image to be detected; processing the heat supply pipe network image by adopting a preset first edge extraction method to obtain a first edge image; processing the heat supply pipe network image by adopting a preset second edge extraction method to obtain a second edge image; processing the heat supply pipe network image by adopting a preset third edge extraction method to obtain a third edge image; and respectively performing wavelet decomposition operation on the first edge image, the second edge image and the third edge image to obtain a low-frequency approximate sub-image, a horizontal high-frequency detail sub-image, a vertical high-frequency detail sub-image and a diagonal high-frequency sub-image corresponding to each of the first edge image, the second edge image and the third edge image.

Description

Technical Field

[0001] The present invention relates to the technical field of heat supply pipe network image recognition, and in particular to a heat supply pipe network image edge detection method and system based on wavelet transform. Background Art

[0002] The accuracy of infrared image recognition is directly related to the accuracy of leak detection. Factors such as complex pipeline distribution and diverse background interference make it difficult for traditional image recognition algorithms to accurately distinguish between leaking points and normal points. Pipelines of different materials and specifications will also appear differently in infrared thermal images, further increasing the difficulty of image recognition. When performing edge detection on pipeline defect images, traditional edge detection cannot detect slight changes in the image, so the accuracy and efficiency are low. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide a method and system for edge detection of a heating pipe network image based on wavelet transform, wherein the method and system can realize accurate edge detection of a heating pipe network image.

[0004] In order to achieve the above object, an embodiment of the present invention provides a method for edge detection of a heating pipe network image based on wavelet transform, comprising:

[0005] Acquire the image of the heating pipe network to be inspected;

[0006] Processing the heating pipe network image using a preset first edge extraction method to obtain a first edge image;

[0007] Processing the heating pipe network image using a preset second edge extraction method to obtain a second edge image;

[0008] Processing the heating pipe network image using a preset third edge extraction method to obtain a third edge image;

[0009] Performing wavelet decomposition operations on the first edge image, the second edge image, and the third edge image respectively to obtain a low-frequency approximate sub-image, a horizontal high-frequency detail sub-image, a vertical high-frequency detail sub-image, and a diagonal high-frequency sub-image corresponding to each of the first edge image, the second edge image, and the third edge image;

[0010] Using a coefficient method to fuse each of the low-frequency approximate sub-image, the horizontal high-frequency detail sub-image, the vertical high-frequency detail sub-image, and the diagonal high-frequency sub-image, respectively, to obtain a fused sub-image;

[0011] The fused sub-images are fused using inverse wavelet transform to obtain an edge image.

[0012] Optionally, the heating pipe network image is processed using a preset first edge extraction method to obtain a first edge image, including:

[0013] Using a first operator to perform convolution calculation on the heating pipe network image in the x-axis direction to obtain an x-axis gradient;

[0014] Using a second operator to perform convolution calculation on the heating pipe network image in the y-axis direction to obtain a y-axis gradient;

[0015] Using a third operator to perform convolution calculation on the heating pipe network image in a 45° direction to obtain a 45° axial gradient;

[0016] Using a fourth operator to perform convolution calculation on the heating pipe network image in the 135° direction to obtain a 135° axial gradient;

[0017] Determine a gradient value of the first edge image according to the x-axis gradient, the y-axis gradient, the 45° axis gradient, and the 135° gradient;

[0018] The first edge image is determined according to the gradient value of the first edge image using a preset first gradient threshold.

[0019] Optionally, the first operator is the following formula (1):

[0020]

[0021] Wherein, T(x) is the first operator;

[0022] The second operator is the following formula (2):

[0023]

[0024] Wherein, T(y) is the second operator;

[0025] The third operator is the following formula (3):

[0026]

[0027] Wherein, T(45) is the third operator;

[0028] The fourth operator is the following formula (4):

[0029]

[0030] Among them, T(135) is the fourth operator.

[0031] Optionally, a first operator is used to perform convolution calculation on the heating pipe network image in the x-axis direction to obtain an x-axis gradient, including:

[0032] The x-axis gradient is calculated according to formula (5):

[0033] G x =(F 31 +2F 32 +F 33 )-(F 11 +2F 12 +F 13 ), (5)

[0034] Among them, G x is the x-axis gradient, F 31 、F 32 、F 33 、F 11 、F 12 and F 13 The pixel of the coordinate point calculated for the current convolution;

[0035] The second operator is used to perform convolution calculation on the heating pipe network image in the y-axis direction to obtain a y-axis gradient, including:

[0036] The y-axis gradient is calculated according to formula (6):

[0037] G y =(F 13 +2F 23 +F 33 )-(F 11 +2F 21 +F 31 ), (6)

[0038] Among them, G y is the y-axis gradient, F 13 、F 23 、F 33 、F 11 、F 21 and F 31 The pixel of the coordinate point calculated for the current convolution;

[0039] The third operator is used to perform convolution calculation on the heating network image in the 45° direction to obtain a 45° axis gradient, including:

[0040] The 45° axis gradient is calculated according to formula (7):

[0041] G 45 =(F 21 +2F 31 +F 32 )-(F 12 +2F 31 +F 23 ), (7)

[0042] Among them, G 45 For the 45° axis gradient, F 21 、F 31 、F 32 、F 12 、F 23 The pixel of the coordinate point calculated for the current convolution;

[0043] The fourth operator is used to perform convolution calculation on the heating network image in the 135° direction to obtain the 135° axial gradient, including:

[0044] G 135 =(F 23 +2F 33 +F 32 )-(F 12 +2F 11 +F 21 ), (8)

[0045] Among them, G 135 is the 135° axial gradient;

[0046] Determining the gradient value of the first edge image according to the x-axis gradient, the y-axis gradient, the 45° axis gradient, and the 135° gradient includes:

[0047] The gradient value of the first edge image is determined according to formula (9):

[0048] G=max{G x , G y , G 45 , G 135}, (9)

[0049] Wherein, G is the gradient value of the first edge image.

[0050] Optionally, the heating pipe network image is processed using a preset second edge extraction method to obtain a second edge image, including:

[0051] Performing a two-dimensional Gaussian filtering operation on the heating pipe network image;

[0052] The result of the two-dimensional Gaussian filtering operation is processed using a first-order directional time domain filter convolution;

[0053] Calculate the maximum value of the gradient of the result of the time domain filtering convolution processing;

[0054] The calculated maximum value is determined as the position of the edge pixel point to obtain the second edge image.

[0055] Optionally, performing a two-dimensional Gaussian filtering operation on the heating pipe network image includes:

[0056] The two-dimensional Gaussian filtering operation is performed according to formula (10):

[0057]

[0058] Where G(x, y) is the pixel value at position (x, y) in the result of the two-dimensional Gaussian filter operation, and δ is the standard deviation;

[0059] The result of the two-dimensional Gaussian filtering operation is processed using a first-order directional time domain filter convolution, including:

[0060] The result of the two-dimensional Gaussian filtering operation is processed according to formula (11) and formula (12):

[0061]

[0062] Among them, ex is the gradient in the x-axis direction, l(x,y) is the pixel point at position (x,y) on the heating network image, and ey is the gradient in the y-axis direction.

[0063] Optionally, the heating pipe network image is processed using a preset third edge extraction method to obtain a third edge image, including:

[0064] Determining an infrared image of the heating network image;

[0065] Using Gaussian filtering method to construct the wavelet function of the infrared image;

[0066] Performing a convolution operation on the wavelet function and the infrared image to obtain a convolved infrared image;

[0067] Performing a wavelet transform operation on the convolution infrared image;

[0068] Calculating the gradient and direction on the convolution infrared image after the wavelet transform operation;

[0069] Searching for the maximum value of the gradient in each direction on the infrared image;

[0070] An edge detection point is determined according to the maximum value using a preset threshold value to obtain the third edge image.

[0071] Optionally, a Gaussian filtering method is used to construct a wavelet function of the infrared image, including:

[0072] The wavelet function is determined according to formula (11) to formula (15):

[0073] f(x,y)>0, (11)

[0074]

[0075]

[0076] Among them, f(x, y) is a smooth binary wavelet basis function defined, H x (x, y) is the wavelet function in the x-axis direction, H y (x, y) is the wavelet function in the y-axis direction, and x and y are function variables;

[0077] The wavelet function is convolved with the infrared image to obtain a convolved infrared image, including:

[0078] The convolution operation is performed using formula (16) to formula (18):

[0079]

[0080] Among them, W S x is the scale function in the horizontal direction, and s is the scale value;

[0081]

[0082] Among them, H S y is the scale function in the height direction;

[0083]

[0084] Among them, W S f(x, y) is the result of the convolution operation;

[0085] Performing a wavelet transform operation on the convolution infrared image includes:

[0086] The wavelet transform operation is performed according to formula (19):

[0087]

[0088] Among them, W S f(x, y) is the result of the wavelet transform operation, f s In order to meet the requirements of convolution calculation between wavelet basis function and infrared image at different wavelet decomposition scales S, the Gaussian smoothing function required by wavelet is defined based on smooth binary wavelet basis function;

[0089] Determining edge detection points according to the maximum value using a preset threshold to obtain the third edge image includes:

[0090] The third edge image is determined according to formula (20):

[0091]

[0092] Among them, W S y is the scale function in the y-axis direction.

[0093] On the other hand, the present invention further provides a heating pipe network image edge detection system based on wavelet transform, the edge detection system comprising a processor, and the processor is configured to execute any of the edge detection methods described above.

[0094] Through the above technical scheme, the embodiment of the present invention provides a method and system for edge detection of heating pipe network images based on wavelet transform. The method and system extract edge features by combining multiple edge extraction methods, and then combine low-frequency, high-frequency, horizontal, and diagonal edge feature extraction for fusion judgment. This overcomes the defect that the commonly used edge detection methods in the prior art are difficult to identify the edges of heating pipe network images, and improves the accuracy of edge detection.

[0095] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0097] Figure 1 is a flow chart of a method for edge detection of a heating pipe network image based on wavelet transform according to an embodiment of the present invention;

[0098] Figure 2 is a flow chart of a first edge extraction method according to one embodiment of the present invention;

[0099] Figure 3 is a flow chart of a second edge extraction method according to one embodiment of the present invention;

[0100] Figure 4 is a flowchart of a third edge extraction method according to one embodiment of the present invention. DETAILED DESCRIPTION

[0101] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0102] like Figure 1 FIG. 1 is a flow chart of a method for edge detection of a heating network image based on wavelet transform according to an embodiment of the present invention. Figure 1In the embodiment of the present invention, the edge detection method may include the following steps:

[0103] In step S10, an image of a heating network to be detected is obtained;

[0104] In step S11, a preset first edge extraction method is used to process the heating pipe network image to obtain a first edge image;

[0105] In step S12, a preset second edge extraction method is used to process the heating pipe network image to obtain a second edge image;

[0106] In step S13, a preset third edge extraction method is used to process the heating pipe network image to obtain a third edge image;

[0107] In step S14, wavelet decomposition operations are performed on the first edge image, the second edge image, and the third edge image, respectively, to obtain a low-frequency approximate sub-image, a horizontal high-frequency detail sub-image, a vertical high-frequency detail sub-image, and a diagonal high-frequency sub-image corresponding to each of the first edge image, the second edge image, and the third edge image;

[0108] In step S15, each low-frequency approximate sub-image, horizontal high-frequency detail sub-image, vertical high-frequency detail sub-image, and diagonal high-frequency sub-image are fused respectively using a coefficient method to obtain a fused sub-image;

[0109] In step S16, the sub-images are fused by inverse wavelet transform to obtain an edge image.

[0110] In this Figure 1 In the edge detection method shown, step S11 can be used to process the heating pipe network image using a preset first edge extraction method to obtain a first edge image. The first edge extraction method can further include: Figure 2 In the steps shown in Figure 2 In the first edge extraction method, the first edge extraction method may include the following steps:

[0111] In step S20, a first operator is used to perform convolution calculation on the heating network image in the x-axis direction to obtain an x-axis gradient. In this embodiment, the first operator may be the following formula (1):

[0112]

[0113] Among them, T(x) is the first operator.

[0114] Correspondingly, when the first operator is formula (1), step S20 may also be to calculate the x-axis gradient according to formula (5):

[0115] G x=(F 31 +2F 32 +F 33 )-(F 11 +2F 12 +F 13 ), (5)

[0116] Among them, G x is the x-axis gradient, F 31 、F 32 、F 33 、F 11 、F 12 and F 13 The pixel of the coordinate point calculated for the current convolution;

[0117] In step S21, a second operator is used to perform convolution calculation on the heating network image in the y-axis direction to obtain a y-axis gradient. In this embodiment, the second operator may be the following formula (2):

[0118]

[0119] Among them, T(y) is the second operator.

[0120] Correspondingly, when the second operator is formula (2), step S20 may also be to calculate the y-axis gradient according to formula (6):

[0121] G y =(F 13 +2F 23 +F 33 )-(F 11 +2F 21 +F 31 ), (6)

[0122] Among them, G y is the y-axis gradient, F 13 、F 23 、F 33 、F 11 、F 21 and F 31 The pixel of the coordinate point calculated for the current convolution.

[0123] In step S22, a third operator is used to perform convolution calculation on the heating network image in the 45° direction to obtain a 45° axial gradient. In this embodiment, the third operator may be the following formula (3):

[0124]

[0125] Among them, T(45) is the third operator.

[0126] Correspondingly, when the second operator is formula (3), step S20 may also be to calculate the 45° axis gradient according to formula (7):

[0127] G 45 =(F 21 +2F 31 +F 32 )-(F 12 +2F 31 +F 23 ), (7)

[0128] Among them, G 45 For a 45° axis gradient, F 21 、F 31 、F 32 、F 12 、F 23 The pixel of the coordinate point calculated for the current convolution.

[0129] In step S23, a fourth operator is used to perform convolution calculation on the heating network image in the 135° direction to obtain a 135° axial gradient. In this embodiment, the fourth operator may be the following formula (4):

[0130]

[0131] Among them, T(135) is the fourth operator.

[0132] Correspondingly, when the second operator is formula (4), step S20 may also be to calculate the 135° axis gradient according to formula (8):

[0133] G 135 =(F 23 +2F 33 +F 32 )-(F 12 +2F 11 +F 21 ), (8)

[0134] Among them, G 135 The axial gradient is 135°.

[0135] In step S24, the gradient value of the first edge image is determined according to the x-axis gradient, the y-axis gradient, the 45° axis gradient, and the 135° gradient. In particular, as to the specific method for determining the gradient value of the first edge image in step S24, in one example of the present invention, since the edge of an image generally refers to an area where a large difference in color or the like occurs, the following formula (9) may be used to determine the gradient value of the first edge image:

[0136] G=max{G x , Gy , G 45 , G 135}, (9)

[0137] Wherein, G is the gradient value of the first edge image.

[0138] In step S25, a preset first gradient threshold is used to determine the first edge image according to the gradient value of the first edge image.

[0139] Step S12 may be used to process the heating pipe network image using a preset second edge extraction method to obtain a second edge image. In this embodiment, the second edge extraction method may include: Figure 3 In the steps shown in Figure 3 In the second edge extraction method, the second edge extraction method may include the following steps:

[0140] In step S30, a two-dimensional Gaussian filtering operation is performed on the heating network image. The two-dimensional Gaussian filtering operation may be the following formula (10):

[0141]

[0142] Wherein, G(x, y) is the pixel value at position (x, y) in the result of the two-dimensional Gaussian filtering operation, and δ is the standard deviation.

[0143] In step S31, the result of the two-dimensional Gaussian filtering operation is processed by first-order time domain filtering convolution. Specifically, step S31 can also be represented by the following formula (11) and formula (12):

[0144]

[0145] Among them, ex is the gradient in the x-axis direction, l(x,y) is the pixel point at position (x,y) on the heating network image, and ey is the gradient in the y-axis direction.

[0146] In step S32, the maximum value of the gradient of the result of the time domain filtering convolution process is calculated;

[0147] In step S33, the calculated maximum value is determined as the position of the edge pixel point to obtain a second edge image.

[0148] Step S13 may be used to process the heating pipe network image using a preset third edge extraction method to obtain a third edge image. In this embodiment, step S13 may include: Figure 4 In the method shown in Figure 4 In the step S13, the step S13 may further include the following steps:

[0149] In step S40, an infrared image of the heating network image is determined;

[0150] In step S41, a wavelet function of the infrared image is constructed by using a Gaussian filtering method. The wavelet function can be expressed by the following formulas (11) to (15):

[0151] f(x,y)>0, (11)

[0152]

[0153] Among them, f(x,y) is a smooth binary wavelet basis function defined, H x (x,y) is the wavelet function in the x-axis direction, H y (x,y) is the wavelet function in the y-axis direction, and x and y are function variables;

[0154] In step S42, the wavelet function is convolved with the infrared image to obtain a convolved infrared image. In this embodiment, step S42 can also be represented by the following formulas (16) to (18):

[0155]

[0156] Among them, W S x is the scale function in the horizontal direction, and s is the scale value;

[0157]

[0158] Among them, H S y is the scale function in the height direction;

[0159]

[0160] Among them, W S f(x, y) is the result of the convolution operation.

[0161] In step S43, a wavelet transform operation is performed on the convolved infrared image. Specifically, in this embodiment, the wavelet transform operation can also be expressed by the following formula (19):

[0162]

[0163] Among them, W s f(x, y) is the result of the wavelet transform operation, f s In order to meet the requirements of convolution calculation between wavelet basis function and infrared image at different wavelet decomposition scales S, the Gaussian smoothing function required by wavelet is defined based on smooth binary wavelet basis function.

[0164] In step S44, the gradient and direction on the convolution infrared image after the wavelet transform operation are calculated;

[0165] In step S45, the maximum value of the gradient is searched in each direction on the infrared image;

[0166] In step S46, a preset threshold is used to determine the edge detection point according to the maximum value to obtain a third edge image. Specifically, in this embodiment, the process of determining the edge detection point can also be expressed by the following formula (20):

[0167]

[0168] Among them, W S y is the scale function in the y-axis direction.

[0169] On the other hand, the present invention further provides a heating pipe network image edge detection system based on wavelet transform, the edge detection system comprising a processor, and the processor is configured to execute any of the edge detection methods described above.

[0170] Through the above technical scheme, the embodiment of the present invention provides a method and system for edge detection of heating pipe network images based on wavelet transform. The method and system extract edge features by combining multiple edge extraction methods, and then combine low-frequency, high-frequency, horizontal, and diagonal edge feature extraction for fusion judgment. This overcomes the defect that the commonly used edge detection methods in the prior art are difficult to identify the edges of heating pipe network images, and improves the accuracy of edge detection.

[0171] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0172] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0173] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0174] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0175] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0176] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0177] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0178] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0179] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for edge detection of heating pipe network images based on wavelet transform, characterized in that: The edge detection method comprises: Acquire the image of the heating pipe network to be inspected; Processing the heating pipe network image using a preset first edge extraction method to obtain a first edge image; Processing the heating pipe network image using a preset second edge extraction method to obtain a second edge image; Processing the heating pipe network image using a preset third edge extraction method to obtain a third edge image; Performing wavelet decomposition operations on the first edge image, the second edge image, and the third edge image respectively to obtain a low-frequency approximate sub-image, a horizontal high-frequency detail sub-image, a vertical high-frequency detail sub-image, and a diagonal high-frequency sub-image corresponding to each of the first edge image, the second edge image, and the third edge image; Using a coefficient method to fuse each of the low-frequency approximate sub-image, the horizontal high-frequency detail sub-image, the vertical high-frequency detail sub-image, and the diagonal high-frequency sub-image, respectively, to obtain a fused sub-image; The fused sub-images are fused using inverse wavelet transform to obtain an edge image.

2. The edge detection method according to claim 1, characterized in that: The heating pipe network image is processed using a preset first edge extraction method to obtain a first edge image, including: Using a first operator to perform convolution calculation on the heating pipe network image in the x-axis direction to obtain an x-axis gradient; Using a second operator to perform convolution calculation on the heating pipe network image in the y-axis direction to obtain a y-axis gradient; Using a third operator to perform convolution calculation on the heating pipe network image in a 45° direction to obtain a 45° axial gradient; Using a fourth operator to perform convolution calculation on the heating pipe network image in the 135° direction to obtain a 135° axial gradient; Determine a gradient value of the first edge image according to the x-axis gradient, the y-axis gradient, the 45° axis gradient, and the 135° gradient; The first edge image is determined according to the gradient value of the first edge image using a preset first gradient threshold.

3. The edge detection method according to claim 2, characterized in that: The first operator is the following formula (1): Wherein, T(x) is the first operator; The second operator is the following formula (2): Wherein, T(y) is the second operator; The third operator is the following formula (3): Wherein, T(45) is the third operator; The fourth operator is the following formula (4): Among them, T(135) is the fourth operator.

4. The edge detection method according to claim 3, characterized in that: The first operator is used to perform convolution calculation on the heating pipe network image in the x-axis direction to obtain an x-axis gradient, including: The x-axis gradient is calculated according to formula (5): G x =(F 31 +2F 32 +F 33 )-(F 11 +2F 12 +F 13 ), (5) Among them, G x is the x-axis gradient, F 31 、F 32 、F 33 、F 11 、F 12 and F 13 The pixel of the coordinate point calculated for the current convolution; The second operator is used to perform convolution calculation on the heating pipe network image in the y-axis direction to obtain a y-axis gradient, including: The y-axis gradient is calculated according to formula (6): G y =(F 13 +2F 23 +F 33 )-(F 11 +2F 21 +F 31 ), (6) Among them, G y is the y-axis gradient, F 13 、F 23 、F 33 、F 11 、F 21 and F 31 The pixel of the coordinate point calculated for the current convolution; The third operator is used to perform convolution calculation on the heating network image in the 45° direction to obtain a 45° axial gradient, including: The 45° axis gradient is calculated according to formula (7): G 45 =(F 21 +2F 31 +F 32 )-(F 12 +2F 31 +F 23 ), (7) Among them, G 45 For the 45° axis gradient, F 21 、F 31 、F 32 、F 12 、F 23 The pixel of the coordinate point calculated for the current convolution; The fourth operator is used to perform convolution calculation on the heating network image in the 135° direction to obtain the 135° axial gradient, including: G 135 =(F 23 +2F 33 +F 32 )-(F 12 +2F 11 +F 21 ), (8) Among them, G 135 is the 135° axial gradient; Determining the gradient value of the first edge image according to the x-axis gradient, the y-axis gradient, the 45° axis gradient, and the 135° gradient includes: The gradient value of the first edge image is determined according to formula (9): G=max{G x ,G y ,G 45 ,G 135 }, (9) Wherein, G is the gradient value of the first edge image.

5. The edge detection method according to claim 1, characterized in that: The heating pipe network image is processed using a preset second edge extraction method to obtain a second edge image, including: Performing a two-dimensional Gaussian filtering operation on the heating pipe network image; The result of the two-dimensional Gaussian filtering operation is processed using a first-order directional time domain filter convolution; Calculate the maximum value of the gradient of the result of the time domain filtering convolution processing; The calculated maximum value is determined as the position of the edge pixel point to obtain the second edge image.

6. The edge detection method according to claim 5, characterized in that: Performing a two-dimensional Gaussian filtering operation on the heating pipe network image includes: The two-dimensional Gaussian filtering operation is performed according to formula (10): Where G(x,y) is the pixel value at position (x,y) in the result of the two-dimensional Gaussian filter operation, and δ is the standard deviation; The result of the two-dimensional Gaussian filtering operation is processed using a first-order directional time domain filter convolution, including: The result of the two-dimensional Gaussian filtering operation is processed according to formula (11) and formula (12): Among them, ex is the gradient in the x-axis direction, l(x,y) is the pixel point at position (x,y) on the heating network image, and ey is the gradient in the y-axis direction.

7. The edge detection method according to claim 1, characterized in that: The heating pipe network image is processed using a preset third edge extraction method to obtain a third edge image, including: Determining an infrared image of the heating network image; Using Gaussian filtering method to construct the wavelet function of the infrared image; Performing a convolution operation on the wavelet function and the infrared image to obtain a convolved infrared image; Performing a wavelet transform operation on the convolution infrared image; Calculating the gradient and direction on the convolution infrared image after the wavelet transform operation; Searching for the maximum value of the gradient in each direction on the infrared image; An edge detection point is determined according to the maximum value using a preset threshold value to obtain the third edge image.

8. The edge detection method according to claim 7, characterized in that: The wavelet function of the infrared image is constructed by using a Gaussian filtering method, including: The wavelet function is determined according to formula (11) to formula (15): f(x,y)>0, (11) Among them, f(x,y) is a smooth binary wavelet basis function defined, H x (x,y) is the wavelet function in the x-axis direction, H y (x,y) is the wavelet function in the y-axis direction, and x and y are function variables; The wavelet function is convolved with the infrared image to obtain a convolved infrared image, including: The convolution operation is performed using formula (16) to formula (18): Among them, W S x is the scale function in the horizontal direction, and s is the scale value; Among them, H S y is the scale function in the height direction; Among them, W S f (x,y) is the result of the convolution operation; Performing a wavelet transform operation on the convolution infrared image includes: The wavelet transform operation is performed according to formula (19): Among them, W s f(x,y) is the result of the wavelet transform operation, f s In order to meet the requirements of convolution calculation between wavelet basis function and infrared image at different wavelet decomposition scales S, the Gaussian smoothing function required by wavelet is defined based on smooth binary wavelet basis function; Determining edge detection points according to the maximum value using a preset threshold to obtain the third edge image includes: The third edge image is determined according to formula (20): Among them, W S y is the scale function in the y-axis direction.

9. A heating pipe network image edge detection system based on wavelet transform, characterized in that: The edge detection system comprises a processor, wherein the processor is configured to execute the edge detection method according to any one of claims 1 to 8.

Images