Intelligent flow calculation system

Abstract

The present invention provides an intelligent flow calculation system, comprising: a type determination end, which is used to determine a set of fluid types contained in the measured fluid based on an acquired fluid video when the measured fluid flows through a measuring pipeline; a video analysis end, which is used to determine a target measured type in the fluid type set, and divide a local video of the measured fluid corresponding to the target measured type in the fluid video, and analyze flow detection data based on the local video of the measured fluid; a flow calculation end, which is used to calculate a real-time flow value of the measured fluid based on the flow detection data and attribute information of the measuring pipeline; it is used to determine the fluid type contained in the measured fluid based on the fluid video, intelligently select the target measured fluid, and then realize tracking detection of the target measured fluid based on image recognition and video tracking technology, so as to detect the flow of one fluid among multiple immiscible fluids.

Description

Technical Field

[0001] The present invention relates to the technical field of flow calculation, and in particular to an intelligent flow calculation system. Background Art

[0002] At present, flow detection and calculation may be required in both industrial manufacturing and natural monitoring. However, most of the existing flow detection and calculation devices can only detect and calculate the flow of one type of fluid. For example, when water and oil are immiscible fluids in the fluid, only the flow of the mixed fluid of water and oil can be detected and calculated, but the flow of water and the flow of oil cannot be detected and calculated separately. Most of them use a pressure detection device set on the inner wall of the detection pipeline, and then combine a series of physical principles such as dynamic volume method to realize the calculation of flow. In this way, when there are multiple immiscible fluids in the detection pipeline, only the flow of the fluid in the preset object shape can be detected and calculated according to the preset flow calculation principle, and the real target fluid to be measured cannot be intelligently selected from the multiple immiscible fluids in the detection pipeline. In addition, the existing fluid detection device is overly dependent on the accuracy of the pressure detection device. Once the pressure detection device fails or has an accuracy error, it will cause a huge error in the flow detection result, so the performance stability of flow detection also needs to be improved.

[0003] Therefore, the present invention proposes an intelligent flow calculation system. Summary of the invention

[0004] The present invention provides an intelligent flow calculation system, which is used to calculate the real-time flow value of the measured fluid based on the flow detection data and the attribute information of the measuring pipeline; to determine the type of fluid contained in the measured fluid based on the fluid video, intelligently select the target measured fluid, and then realize tracking detection of the target measured fluid based on image recognition and video tracking technology. It can detect the flow of one fluid among multiple immiscible fluids, and also greatly reduce the dependence on the pressure detection device in the flow detection process, thereby improving the performance stability of flow detection.

[0005] The present invention provides an intelligent flow calculation system, comprising:

[0006] A type determination terminal, used to determine a set of fluid types contained in the measured fluid based on the acquired fluid video when the measured fluid flows through the measurement pipeline;

[0007] The video analysis end is used to determine the target detected type in the fluid type set, divide the detected fluid local video corresponding to the target detected type in the fluid video, and analyze the flow detection data based on the detected fluid local video;

[0008] The flow calculation end is used to calculate the real-time flow value of the measured fluid based on the flow detection data and the property information of the measurement pipeline.

[0009] Preferably, the type determination terminal includes:

[0010] A video acquisition module, used for acquiring a video of the fluid in the measuring pipe when the measured fluid flows through the measuring pipe;

[0011] A position determination module, used to determine a pressure detection position based on the fluid video;

[0012] The type determination module is used to obtain the pressure value at the pressure detection position, and determine the set of fluid types contained in the measured fluid based on the pressure value and the fluid pressure value range list.

[0013] Preferably, the location determination module comprises:

[0014] a boundary judgment unit, used to judge whether there is a fluid boundary in each video frame of the fluid video, and if so, to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region, otherwise, to determine a pressure detection position based on a preset detection position list;

[0015] The position determination unit is used to determine the pressure detection position based on a partial fluid area included in the video frame where the fluid boundary exists.

[0016] Preferably, the boundary judgment unit includes:

[0017] A larger screening subunit is used to calculate a first number based on the total number of pixels contained in the video frame and a first preset proportion, sort the pixel values ​​of all pixels in the video frame from large to small to obtain a pixel value sequence of the video frame, aggregate the first several pixel values ​​in the pixel value sequence to obtain a larger pixel value set of the video frame, and determine a larger pixel representation value of the video frame based on the larger pixel value set;

[0018] a boundary judgment subunit, configured to calculate a pixel value difference between the larger pixel representation value and a minimum pixel value in the video frame, and when the pixel value difference is greater than a pixel value difference threshold, determine that a fluid boundary exists in the video frame; otherwise, determine that no fluid boundary exists in the video frame, and determine a pressure detection position based on a preset detection position list;

[0019] The region division subunit is used to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region when it is determined that there is a fluid boundary in the video frame.

[0020] Preferably, the area division subunit includes:

[0021] A Gaussian filter subunit, configured to perform Gaussian filtering on the video frame to obtain a corresponding denoised image when it is determined that a fluid boundary exists in the video frame;

[0022] a gradient determination subunit, configured to determine a horizontal pixel distribution function and a vertical pixel distribution function based on the pixel distribution data in the denoised image, and to determine a transverse gradient value and a longitudinal gradient value of each pixel in the denoised image based on a first-order derivative of the horizontal pixel distribution function and the vertical pixel distribution function and a coordinate value of each pixel, and to determine a comprehensive gradient value and a gradient angle value of the pixel based on the transverse gradient value and the longitudinal gradient value;

[0023] an adjacent determination subunit, configured to determine, in a preset gradient angle list, a preset gradient angle having a minimum difference with the gradient angle value as a standard approximate gradient value of the pixel point, and determine a first adjacent pixel point adjacent to the pixel point in a positive direction of the standard approximate gradient value and a second adjacent pixel point adjacent to the pixel point in a negative direction of the standard approximate gradient value;

[0024] a gradient comparison subunit, configured to retain the corresponding pixel when the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the first adjacent pixel and the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the second pixel; otherwise, set the corresponding pixel to zero, traverse and compare each pixel in the denoised image, and obtain a first image;

[0025] a threshold extraction subunit, configured to identify an initial fluid boundary in the first image, determine a first threshold based on pixel distribution data of the initial fluid boundary, calculate a second threshold based on the first threshold and a preset ratio, and perform threshold extraction on the denoised image based on the first threshold and the second threshold, respectively, to obtain a corresponding second image and a third image;

[0026] a position corresponding subunit, configured to select any first non-zero pixel point in the second image, and perform edge tracing in the second image with the first non-zero pixel point as a starting point until an end point of a contour line is determined, and determine a first pixel point at the same position corresponding to the end point of the contour line in the third image;

[0027] A contour compensation subunit is used to determine whether there is a second non-zero pixel point adjacent to the first co-positioned pixel point in the third image. If so, a second co-positioned pixel point corresponding to the second non-zero pixel point in the second image will be determined, and the pixel value of the second co-positioned pixel point will be set to the pixel value of the second non-zero pixel point, and edge tracking will continue in the second image with the second co-positioned pixel point as the starting point. When there is no non-zero pixel point adjacent to the latest co-positioned pixel point in the third image, contour tracking will be stopped, and the fluid boundary of the corresponding video frame will be obtained, and the corresponding video frame will be divided into regions based on the fluid boundary to obtain the corresponding partial fluid region.

[0028] Preferably, the position determination unit comprises:

[0029] A matrix construction subunit, used to take each pixel point included in a partial fluid area in the fluid video as a central pixel point, and construct a local pixel matrix of the central pixel point based on the pixel value of the central pixel point and the pixel value of the corresponding field pixel point;

[0030] a comprehensive determination subunit, configured to determine a local gradient value of each pixel in the local pixel matrix, determine a local gradient matrix corresponding to the local pixel matrix based on the local gradient value, rotate the local gradient values ​​contained in the local pixel matrix in sequence according to the preset gradients contained in the preset gradient list, obtain a rotated gradient matrix corresponding to each preset gradient, and use an average matrix of all the rotated gradient matrices as the corresponding comprehensive rotated gradient matrix;

[0031] A pixel screening subunit, used for treating corresponding pixel points as rotation invariant pixel points when the comprehensive rotation gradient matrix and the local gradient matrix are the same;

[0032] A vector determination subunit, configured to determine the coordinate value of each rotationally invariant pixel point contained in the partial fluid region, and construct a corresponding local gradient vector based on the coordinate value of the rotationally invariant pixel point and the local gradient value;

[0033] A weighted processing subunit, configured to perform Gaussian weighted processing on the local gradient vector based on the coordinate values ​​of all rotationally invariant pixels contained in the partial fluid region and the local gradient vector, so as to obtain a stable gradient vector corresponding to the local gradient vector;

[0034] A difference determination subunit, configured to determine a corresponding comprehensive stable point feature vector based on all stable gradient vectors contained in the partial fluid region, and determine an angle difference of the comprehensive stable point feature vector between every two partial fluid regions in the fluid video;

[0035] The classification and aggregation subunit is used to classify two partial fluid regions whose angle difference of the comprehensive stable point characteristic vector is less than the angle difference threshold as the same fluid type, until all partial fluid regions contained in the partial fluid region are divided, and the partial fluid regions belonging to the same fluid type are aggregated to obtain a fluid region set of each fluid type;

[0036] a center determination subunit, configured to extract all video frames with fluid boundaries from the fluid video as fluid boundary video frames, and determine the contact center position of each partial fluid area based on the contact boundary with the measuring pipe in each partial fluid area in the fluid boundary video frames;

[0037] The position determination subunit is used to determine the individual pressure detection position of the corresponding fluid type based on all contact center positions in the fluid area set, and to obtain the pressure detection position by summarizing the individual pressure detection positions of all fluid types.

[0038] Preferably, the type determination module includes:

[0039] an initial determination unit, configured to obtain a pressure value at the pressure detection position, and determine a fluid type of each fluid region set based on the pressure value and a list of fluid pressure value ranges;

[0040] The type summarizing unit is used to summarize the fluid types of all fluid area sets to obtain a set of fluid types contained in the measured fluid.

[0041] Preferably, the video analysis terminal includes:

[0042] An evaluation calculation unit, used to determine the total area of ​​all fluid regions included in the fluid region set of each fluid type, and calculate a corresponding measured evaluation value based on a preset measured weight corresponding to the fluid type and the corresponding total area;

[0043] A final determination unit, configured to take the fluid type corresponding to the maximum measured evaluation value in the fluid type set as the target measured type;

[0044] A video segmentation module, used to segment the fluid video into a local video of the target measured type of the measured fluid;

[0045] The data analysis module is used to analyze the flow detection data in the measuring pipeline based on the local video of the measured fluid.

[0046] Preferably, the data analysis module includes:

[0047] A displacement tracking unit, used for performing video tracking on each pixel point included in the local video of the measured fluid to obtain a corresponding tracking displacement record;

[0048] A flow rate determination unit is used to determine the real-time flow rate of the pixel point based on the tracking displacement record, and use the real-time flow rate of all pixel points contained in the local video of the measured fluid as flow detection data in the measurement pipeline.

[0049] Preferably, the flow calculation end includes:

[0050] A flow rate determination module, used to take the average value of the real-time flow rates of all pixels included in the flow detection data as the real-time comprehensive flow rate of the measured fluid in the measurement pipeline;

[0051] The flow calculation module is used to calculate the real-time flow value of the measured fluid based on the property information of the measuring pipeline and the real-time comprehensive flow velocity.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0053] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 A schematic diagram of an intelligent flow calculation system in an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of a type determination terminal in an embodiment of the present invention;

[0057] Figure 3 A schematic diagram of a position determination module in an embodiment of the present invention;

[0058] Figure 4 A schematic diagram of a boundary determination unit in an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of a region division subunit in an embodiment of the present invention;

[0060] Figure 6 is a schematic diagram of a position determination unit in an embodiment of the present invention;

[0061] Figure 7A schematic diagram of a type determination module in an embodiment of the present invention;

[0062] Figure 8 A schematic diagram of a video analysis terminal in an embodiment of the present invention;

[0063] Fig. 9 This is a schematic diagram of a data analysis module in an embodiment of the present invention;

[0064] Fig.10 Schematic diagram of a flow calculation terminal in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0066] Embodiment 1:

[0067] The present invention provides an intelligent flow calculation system, referring to Figure 1 ,include:

[0068] A type determination terminal, used to determine a set of fluid types contained in the measured fluid based on the acquired fluid video when the measured fluid flows through the measurement pipeline;

[0069] The video analysis end is used to determine the target detected type in the fluid type set, divide the detected fluid local video corresponding to the target detected type in the fluid video, and analyze the flow detection data based on the detected fluid local video;

[0070] The flow calculation end is used to calculate the real-time flow value of the measured fluid based on the flow detection data and the property information of the measurement pipeline.

[0071] In this embodiment, the measured fluid is the fluid (such as water or oil or a mixed fluid of water and oil) flowing through the measuring pipe and containing the flow rate to be measured.

[0072] In this embodiment, the measuring pipe is a pipe placed in the measured fluid for allowing the measured fluid to flow through and thereby detecting the flow rate of the measured fluid.

[0073] In this embodiment, the type of the measured fluid can be a single liquid or a plurality of immiscible fluids whose stratification boundaries can be captured within a certain flow rate range, such as water or oil, or a mixed fluid of water and oil.

[0074] In this embodiment, the fluid type set is a set consisting of fluid types contained in the measured fluid.

[0075] In this embodiment, the fluid video is a video monitoring the flow of the measured fluid through the measuring pipeline.

[0076] In this embodiment, the target type to be detected is the type of fluid whose flow rate needs to be detected in this embodiment.

[0077] In this embodiment, the local video of the fluid being measured is the local video corresponding to the target type of fluid being measured in the fluid video.

[0078] In this embodiment, the flow detection data is data related to the current flow condition of the measured fluid in the measuring pipeline analyzed based on the type of the measured fluid and the fluid video.

[0079] In this embodiment, the attribute information of the measuring pipeline is: the spatial dimension data of the measured fluid at the setting position of the measuring pipeline.

[0080] In this embodiment, the real-time flow value is the flow value of the measured fluid calculated in real time based on the flow detection data and the property information of the measuring pipeline.

[0081] The beneficial effects of the above technology are: based on the flow detection data and the attribute information of the measuring pipeline, the real-time flow value of the measured fluid is calculated; based on the fluid video, the type of fluid contained in the measured fluid is determined, and the target measured fluid is intelligently selected, and then the tracking detection of the target measured fluid is achieved based on image recognition and video tracking technology, and the flow of one fluid among multiple immiscible fluids can be detected.

[0082] Embodiment 2:

[0083] Based on Example 1, the type determination end refers to Figure 2 ,include:

[0084] A video acquisition module, used for acquiring a video of the fluid in the measuring pipe when the measured fluid flows through the measuring pipe;

[0085] A position determination module, used to determine a pressure detection position based on the fluid video;

[0086] The type determination module is used to obtain the pressure value at the pressure detection position, and determine the set of fluid types contained in the measured fluid based on the pressure value and the fluid pressure value range list.

[0087] In this embodiment, the pressure detection position is the position where the pressure value needs to be obtained, which is determined based on the fluid video.

[0088] In this embodiment, the fluid pressure value range list is a list containing value ranges of fluid pressure values ​​corresponding to different types of fluids.

[0089] In this embodiment, based on the pressure value and the fluid pressure value range list, the type of the measured fluid is determined, that is:

[0090] It is determined which fluid in the fluid pressure value range list corresponds to the fluid pressure value range, and then the type of the corresponding measured fluid is determined.

[0091] The beneficial effects of the above technology are: based on the video of the measured fluid flowing through the measuring pipeline, the pressure detection position is determined in a targeted manner, and the pressure value can be flexibly obtained based on the content of different fluids and the contact position with the detection pipeline under different circumstances. Based on the pressure value obtained at the pressure detection position, all types of fluids flowing through the detection pipeline are judged, thereby realizing the discrimination of the types of fluids flowing through the detection pipeline, which provides an important basis for the subsequent preparation of calculating the flow rate of the measured fluid.

[0092] Embodiment 3:

[0093] Based on Example 2, the position determination module refers to Figure 3 ,include:

[0094] a boundary judgment unit, used to judge whether there is a fluid boundary in each video frame of the fluid video, and if so, to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region, otherwise, to determine a pressure detection position based on a preset detection position list;

[0095] The position determination unit is used to determine the pressure detection position based on a partial fluid area included in the video frame where the fluid boundary exists.

[0096] In this embodiment, the fluid boundary is the boundary between fluids of different forms that may be included in the video frame of the fluid video. For example, when water and oil pass through the detection pipeline at the same time, the boundary between water and oil is the fluid boundary.

[0097] In this embodiment, the partial fluid region is an image region corresponding to fluids of different forms divided into corresponding video frames based on the fluid boundary.

[0098] In this embodiment, the preset detection position list is a list of pre-prepared pressure detection positions that should be acquired when the fluid flowing through the detection pipeline includes only one type of fluid.

[0099] The beneficial effects of the above technology are: by judging whether there is a fluid boundary between fluids of different forms in each frame of the fluid video, the image area corresponding to different types of fluids can be divided, and then the pressure detection position can be determined specifically according to the type of fluid contained in the fluid flowing through the detection pipeline, so that before measuring the fluid flow rate, the type of fluid flowing through the detection pipeline can be determined more accurately, which also lays the foundation for the subsequent accurate calculation of the flow rate of the measured fluid.

[0100] Embodiment 4:

[0101] On the basis of Example 3, the boundary judgment unit refers to Figure 4 ,include:

[0102] A larger screening subunit is used to calculate a first number based on the total number of pixels contained in the video frame and a first preset proportion, sort the pixel values ​​of all pixels in the video frame from large to small to obtain a pixel value sequence of the video frame, aggregate the first several pixel values ​​in the pixel value sequence to obtain a larger pixel value set of the video frame, and determine a larger pixel representation value of the video frame based on the larger pixel value set;

[0103] a boundary judgment subunit, configured to calculate a pixel value difference between the larger pixel representation value and a minimum pixel value in the video frame, and when the pixel value difference is greater than a pixel value difference threshold, determine that a fluid boundary exists in the video frame; otherwise, determine that no fluid boundary exists in the video frame, and determine a pressure detection position based on a preset detection position list;

[0104] The region division subunit is used to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region when it is determined that there is a fluid boundary in the video frame.

[0105] In this embodiment, the first number is the product of the total number of pixels included in the video frame and the first preset proportion.

[0106] In this embodiment, the first ratio is set to 0.1.

[0107] In this embodiment, the pixel value sequence is a sequence obtained by sorting the pixel values ​​of all pixels in the video frame from large to small.

[0108] In this embodiment, the larger pixel value set is a set of partial pixel values ​​of the video frame obtained by aggregating the first several pixel values ​​in the pixel value sequence.

[0109] In this embodiment, the larger pixel representation value of the video frame is determined based on the larger pixel value set, that is, the average value of all larger pixel values ​​included in the larger pixel value set is used as the larger pixel representation value of the video frame.

[0110] In this embodiment, the pixel value difference threshold is the maximum pixel value difference corresponding to determining that there is no fluid boundary in the video frame.

[0111] In this embodiment, the partial fluid region is an image region obtained by dividing the corresponding video frame into regions based on the fluid boundary.

[0112] The beneficial effects of the above technology are: based on the difference between the representation value representing the larger pixel value contained in the video frame and the minimum pixel value contained in the video frame, it is possible to judge whether there is a fluid boundary in the video frame based on the principle that the boundary pixel points are composed of a large number of pixel points with a large difference in pixel value with other pixel points, and thus the division of image areas corresponding to different types of fluids is also realized.

[0113] Embodiment 5:

[0114] On the basis of Example 4, the area division subunit, referring to Figure 5 ,include:

[0115] A Gaussian filter subunit, configured to perform Gaussian filtering on the video frame to obtain a corresponding denoised image when it is determined that a fluid boundary exists in the video frame;

[0116] a gradient determination subunit, configured to determine a horizontal pixel distribution function and a vertical pixel distribution function based on the pixel distribution data in the denoised image, and to determine a transverse gradient value and a longitudinal gradient value of each pixel in the denoised image based on a first-order derivative of the horizontal pixel distribution function and the vertical pixel distribution function and a coordinate value of each pixel, and to determine a comprehensive gradient value and a gradient angle value of the pixel based on the transverse gradient value and the longitudinal gradient value;

[0117] an adjacent determination subunit, configured to determine, in a preset gradient angle list, a preset gradient angle having a minimum difference with the gradient angle value as a standard approximate gradient value of the pixel point, and determine a first adjacent pixel point adjacent to the pixel point in a positive direction of the standard approximate gradient value and a second adjacent pixel point adjacent to the pixel point in a negative direction of the standard approximate gradient value;

[0118] a gradient comparison subunit, configured to retain the corresponding pixel when the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the first adjacent pixel and the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the second pixel; otherwise, set the corresponding pixel to zero, traverse and compare each pixel in the denoised image, and obtain a first image;

[0119] a threshold extraction subunit, configured to identify an initial fluid boundary in the first image, determine a first threshold based on pixel distribution data of the initial fluid boundary, calculate a second threshold based on the first threshold and a preset ratio, and perform threshold extraction on the denoised image based on the first threshold and the second threshold, respectively, to obtain a corresponding second image and a third image;

[0120] a position corresponding subunit, configured to select any first non-zero pixel point in the second image, and perform edge tracing in the second image with the first non-zero pixel point as a starting point until an end point of a contour line is determined, and determine a first pixel point at the same position corresponding to the end point of the contour line in the third image;

[0121] A contour compensation subunit is used to determine whether there is a second non-zero pixel point adjacent to the first co-positioned pixel point in the third image. If so, a second co-positioned pixel point corresponding to the second non-zero pixel point in the second image will be determined, and the pixel value of the second co-positioned pixel point will be set to the pixel value of the second non-zero pixel point, and edge tracking will continue in the second image with the second co-positioned pixel point as the starting point. When there is no non-zero pixel point adjacent to the latest co-positioned pixel point in the third image, contour tracking will be stopped, and the fluid boundary of the corresponding video frame will be obtained, and the corresponding video frame will be divided into regions based on the fluid boundary to obtain the corresponding partial fluid region.

[0122] In this embodiment, the denoised image is an image obtained by performing Gaussian filtering on the video frame when it is determined that a fluid boundary exists in the video frame.

[0123] In this embodiment, the horizontal pixel distribution function is a function of the pixel distribution data in each row of pixels determined based on the pixel distribution data in the denoised image.

[0124] In this embodiment, the pixel distribution data is data representing the pixel value of each pixel point included in the denoised image.

[0125] In this embodiment, the vertical pixel distribution function is a function of the pixel distribution data in each column of pixels determined based on the pixel distribution data in the denoised image.

[0126] In this embodiment, the transverse gradient value is a value determined by substituting the coordinate value of the pixel point into the first-order derivative of the horizontal pixel distribution function of the row where the pixel point is located.

[0127] In this embodiment, the longitudinal gradient value is a value determined by substituting the coordinate value of the pixel point into the first-order derivative of the vertical pixel distribution function of the column where the pixel point is located.

[0128] In this embodiment, the comprehensive gradient value is a value obtained by taking the square root of the sum of the square of the transverse gradient value and the square of the longitudinal gradient value.

[0129] In this embodiment, the gradient angle value is the arc tangent value of the ratio of the longitudinal gradient value to the transverse gradient value.

[0130] In this embodiment, the preset gradient angle list is: 0°, 45°, 90°, 135°, 180°, 225°, 270°, 315°, and 360°.

[0131] In this embodiment, the standard approximate gradient value is the preset gradient determined in the preset gradient angle list and having the smallest difference with the gradient angle value.

[0132] In this embodiment, the first adjacent pixel point is a pixel point adjacent to the pixel point in the positive direction of the standard approximate gradient value.

[0133] In this embodiment, the second adjacent pixel point is a pixel point adjacent to the pixel point in the negative direction of the standard approximate gradient value.

[0134] In this embodiment, the first image is obtained by the following steps: when the comprehensive gradient value of the pixel point is greater than the comprehensive gradient value of the first adjacent pixel point and the comprehensive gradient value of the pixel point is greater than the comprehensive gradient value of the second pixel point, the corresponding pixel point is retained; otherwise, the corresponding pixel point is set to zero, and each pixel point in the denoised image is traversed and compared to obtain an image.

[0135] In this embodiment, the initial fluid boundary is the edge line identified in the first image.

[0136] In this embodiment, the first threshold is determined based on the pixel distribution data of the initial fluid boundary, that is, the average value of the pixel values ​​of all the pixel points included in the initial fluid boundary is used as the first threshold.

[0137] In this embodiment, the first threshold is a pixel threshold determined based on the pixel distribution data of the initial fluid boundary and used for performing threshold extraction on the denoised image.

[0138] In this embodiment, the preset ratio is 0.3.

[0139] In this embodiment, the second threshold is the product of the first threshold and the preset proportion.

[0140] In this embodiment, the second image is an image obtained by setting all the pixels in the denoised image whose pixel values ​​do not exceed the first threshold to zero.

[0141] In this embodiment, the third image is an image obtained by setting all the pixels in the denoised image whose pixel values ​​do not exceed the second threshold to zero.

[0142] In this embodiment, the first non-zero pixel point is a non-zero pixel point randomly selected in the second image.

[0143] In this embodiment, the end point of the contour line is the point where the edge is traced in the second image starting from the first non-zero pixel point until the determined contour ends.

[0144] In this embodiment, the first co-located pixel point is the pixel point where the end point of the contour line is located at the same position in the third image.

[0145] In this embodiment, the second non-zero pixel point is a non-zero pixel point in the third image that is adjacent to the first pixel point at the same position.

[0146] In this embodiment, the second co-position pixel point is a pixel point at the same position of the second non-zero pixel point in the second image.

[0147] In this embodiment, the fluid boundary is the edge line obtained after the entire contour tracing process in the second image is stopped.

[0148] The beneficial effects of the above technology are as follows: based on the gradient values ​​and gradient angles determined by the horizontal and vertical derivatives in the denoised image corresponding to the video frame, and then based on the non-maximum suppression operation of each pixel point and the adjacent pixel points in the denoised image, the first image is determined, and the rough fluid boundary is initially determined in the denoised image, which provides a basis for the subsequent determination of the accurate fluid boundary. Different pixel extraction thresholds are determined based on the pixel distribution data in the initially determined fluid boundary, which provides an extraction standard for the subsequent fine extraction of the fluid boundary in the denoised image based on the dual threshold algorithm. The two fluid boundaries with different completeness extracted based on the dual threshold algorithm complement each other, and the accurate extraction of the fluid boundary in the video frame is achieved.

[0149] Embodiment 6:

[0150] Based on Example 5, the position determination unit refers to Figure 6 ,include:

[0151] A matrix construction subunit, used to take each pixel point included in a partial fluid area in the fluid video as a central pixel point, and construct a local pixel matrix of the central pixel point based on the pixel value of the central pixel point and the pixel value of the corresponding field pixel point;

[0152] a comprehensive determination subunit, configured to determine a local gradient value of each pixel in the local pixel matrix, determine a local gradient matrix corresponding to the local pixel matrix based on the local gradient value, rotate the local gradient values ​​contained in the local pixel matrix in sequence according to the preset gradients contained in the preset gradient list, obtain a rotated gradient matrix corresponding to each preset gradient, and use an average matrix of all the rotated gradient matrices as the corresponding comprehensive rotated gradient matrix;

[0153] A pixel screening subunit, used for treating corresponding pixel points as rotation invariant pixel points when the comprehensive rotation gradient matrix and the local gradient matrix are the same;

[0154] A vector determination subunit, configured to determine the coordinate value of each rotationally invariant pixel point contained in the partial fluid region, and construct a corresponding local gradient vector based on the coordinate value of the rotationally invariant pixel point and the local gradient value;

[0155] A weighted processing subunit, configured to perform Gaussian weighted processing on the local gradient vector based on the coordinate values ​​of all rotationally invariant pixels contained in the partial fluid region and the local gradient vector, so as to obtain a stable gradient vector corresponding to the local gradient vector;

[0156] A difference determination subunit, configured to determine a corresponding comprehensive stable point feature vector based on all stable gradient vectors contained in the partial fluid region, and determine an angle difference of the comprehensive stable point feature vector between every two partial fluid regions in the fluid video;

[0157] The classification and aggregation subunit is used to classify two partial fluid regions whose angle difference of the comprehensive stable point characteristic vector is less than the angle difference threshold as the same fluid type, until all partial fluid regions contained in the partial fluid region are divided, and the partial fluid regions belonging to the same fluid type are aggregated to obtain a fluid region set of each fluid type;

[0158] a center determination subunit, configured to extract all video frames with fluid boundaries from the fluid video as fluid boundary video frames, and determine the contact center position of each partial fluid area based on the contact boundary with the measuring pipe in each partial fluid area in the fluid boundary video frames;

[0159] The position determination subunit is used to determine the individual pressure detection position of the corresponding fluid type based on all contact center positions in the fluid area set, and to obtain the pressure detection position by summarizing the individual pressure detection positions of all fluid types.

[0160] In this embodiment, the central pixel point is a pixel point contained in a partial fluid area in the fluid video.

[0161] In this embodiment, the local pixel matrix is ​​a third-order matrix corresponding to the central pixel point constructed by taking the pixel value of the central pixel point and the pixel value of the corresponding area pixel point as the value at the corresponding position in the matrix.

[0162] In this embodiment, the comprehensive rotation gradient matrix is ​​the average matrix of all rotation gradient matrices.

[0163] In this embodiment, determining the local gradient value of each pixel in the local pixel matrix includes:

[0164] Based on the local pixel matrix, the horizontal pixel distribution function and the vertical pixel distribution function of the pixel point are determined, the first derivative of the horizontal pixel distribution function is used as the horizontal gradient value, the first derivative of the vertical pixel distribution function is used as the vertical gradient value, and the square root of the sum of the square of the horizontal gradient value and the square of the vertical gradient value is obtained as the local gradient value of the pixel point.

[0165] In this embodiment, the local gradient matrix is ​​a matrix obtained by replacing the pixel values ​​of corresponding pixel points in the local pixel matrix with corresponding gradient values.

[0166] In this embodiment, the rotated gradient matrix is ​​a new matrix obtained by rotating the local gradient values ​​in the local gradient matrix in sequence according to the preset gradients included in the preset gradient list.

[0167] In this embodiment, the rotation invariant pixel point is a pixel point whose corresponding comprehensive rotation gradient matrix and corresponding local gradient matrix are the same.

[0168] In this embodiment, the coordinate value of each rotationally invariant pixel point included in the partial fluid region is determined, and the corresponding local gradient vector is constructed based on the coordinate value of the rotationally invariant pixel point and the local gradient value, including: x is the horizontal coordinate value of the rotation-invariant pixel point, y is the vertical coordinate value of the rotation-invariant pixel point, and t is the local gradient value of the rotation-invariant pixel point.

[0169] In this embodiment, Gaussian weighted processing is performed on the local gradient vector based on the coordinate values ​​of all rotationally invariant pixels contained in the partial fluid region and the local gradient vector to obtain a stable gradient vector corresponding to the local gradient vector, including:

[0170] The local gradient vector currently being Gaussian weighted is regarded as a target vector, and based on the coordinate values ​​of all rotationally invariant pixels contained in the partial fluid region, the interval distance between the target vector and the remaining local gradient vectors in the partial fluid region except the target vector is calculated:

[0171]

[0172] Where l is the distance between the target vector and the remaining currently calculated local gradient vectors in the partial fluid region except the target vector, and x m1 is the starting point abscissa of the target vector, x j1 is the starting point abscissa of the currently calculated local gradient vector remaining in the partial fluid region except the target vector, y m1 is the horizontal coordinate of the vertical point of the target vector, y j1 is the starting point ordinate of the remaining currently calculated local gradient vector in the partial fluid region except the target vector, t m1 is the vertical coordinate of the starting point of the target vector, t j1 is the vertical coordinate of the starting point of the currently calculated local gradient vector remaining in the partial fluid region except the target vector, x m2 is the abscissa of the end point of the target vector, x j2 is the end point abscissa of the currently calculated local gradient vector remaining in the partial fluid region except the target vector, y m2 is the ordinate of the end point of the target vector, y j2 is the end point ordinate of the currently calculated local gradient vector remaining in the partial fluid region except the target vector, t m2 is the vertical coordinate of the end point of the target vector, t j2 The vertical coordinate of the end point of the currently calculated local gradient vector remaining in the partial fluid region except the target vector;

[0173] For example, the starting coordinate value of the target vector is (3,4,0), and the ending ordinate is (5,5,0); the starting coordinate value of the remaining currently calculated local gradient vector in the partial fluid area except the target vector is (4,4,0), and the ending ordinate is (5,5,0), then l is 0.5.

[0174] Based on the interval between the target vector and the remaining local gradient vectors in the partial fluid region except the target vector, Gaussian weighted processing is performed on the target vector to obtain a stable gradient vector corresponding to the target vector:

[0175]

[0176] In the formula, is the stable gradient vector corresponding to the target vector, is the target vector, i is the currently calculated local gradient vector remaining in the partial fluid region except the target vector, n is the total number of local gradient vectors remaining in the partial fluid region except the target vector, is the i-th local gradient vector remaining in the partial fluid region except the target vector, l i is the interval distance between the target vector and the i-th local gradient vector remaining in the partial fluid region except the target vector;

[0177] For example, the starting coordinate value of the target vector is (3,4,0), and the ending ordinate is (5,5,0); in addition to the target vector, there is a local gradient vector in the partial fluid region, and the starting coordinate value of the currently calculated local gradient vector is (4,4,0), and the ending ordinate is (5,5,0), and the interval distance between the target vector and the remaining local gradient vector in the partial fluid region except the target vector is 0.5, then The starting point coordinate value is (5,6,0) and the end point vertical coordinate is (7.5,7.5,0).

[0178] In this embodiment, the stable gradient vector is a vector obtained by performing Gaussian weighting processing on the local gradient vector based on the coordinate values ​​of all rotationally invariant pixel points contained in the partial fluid region and the local gradient vector.

[0179] In this embodiment, the angle difference of the comprehensive stabilization point feature vector between every two partial fluid regions in the fluid video is determined, that is, the angle difference of the comprehensive stabilization point feature vector between the two partial fluid regions is used as the angle difference of the comprehensive stabilization point feature vector.

[0180] In this embodiment, the fluid region set is a region set corresponding to each fluid type obtained by aggregating some fluid regions belonging to the same fluid type.

[0181] In this embodiment, the angle difference threshold is the maximum angle difference threshold corresponding to dividing two partial fluid regions into the same fluid type.

[0182] In this embodiment, the fluid boundary video frame is a video frame having a fluid boundary.

[0183] In this embodiment, the contact boundary line is an edge line in the partial fluid region that contacts the measuring pipe.

[0184] In this embodiment, the contact center position is the midpoint position of the contact boundary line.

[0185] In this embodiment, the individual pressure detection positions are all contact center positions in the fluid region set corresponding to the fluid type.

[0186] The beneficial effects of the above technology are: based on the comparison of the comprehensive rotation gradient matrix determined by rotating and averaging the local gradient matrix corresponding to each pixel point contained in the partial fluid area according to the preset gradient angle list and the local gradient matrix, the points whose gradient properties are not affected by the rotation in the partial fluid area can be screened out, and then the local gradient vectors corresponding to the screened points are Gaussian-weighted averaged to determine the vector that can characterize the comprehensive gradient characteristics of the partial fluid area, thereby reducing the influence of the vector characterizing the comprehensive gradient characteristics caused by the change of fluid morphology during the fluid movement, and then realizing the classification and division of partial fluid areas with similar comprehensive gradient characteristics based on the comprehensive stable point feature vector, realizing the visual classification of partial fluid areas, and also realizing the targeted determination of the pressure detection position according to the fluid type contained in the fluid flowing through the detection pipeline, and also laying the foundation for the subsequent accurate calculation of the flow rate of the measured fluid.

[0187] Embodiment 7:

[0188] On the basis of Example 6, the type determination module refers to Figure 7 ,include:

[0189] an initial determination unit, configured to obtain a pressure value at the pressure detection position, and determine a fluid type of each fluid region set based on the pressure value and a list of fluid pressure value ranges;

[0190] The type summarizing unit is used to summarize the fluid types of all fluid area sets to obtain a set of fluid types contained in the measured fluid.

[0191] In this embodiment, the fluid region set is a set consisting of partial fluid regions corresponding to the corresponding fluid type.

[0192] The beneficial effects of the above technology are: determining the fluid type of each fluid area set based on the pressure detection results, and then determining all fluid types flowing through the detection pipeline, providing a basis for subsequent determination of the target type to be tested.

[0193] Embodiment 8:

[0194] Based on Example 1, the video analysis end refers to Figure 8 ,include:

[0195] An evaluation calculation unit, used to determine the total area of ​​all fluid regions included in the fluid region set of each fluid type, and calculate a corresponding measured evaluation value based on a preset measured weight corresponding to the fluid type and the corresponding total area;

[0196] A final determination unit, configured to take the fluid type corresponding to the maximum measured evaluation value in the fluid type set as the target measured type;

[0197] A video segmentation module, used to segment the fluid video into a local video of the target measured type of the measured fluid;

[0198] The data analysis module is used to analyze the flow detection data in the measuring pipeline based on the local video of the measured fluid.

[0199] In this embodiment, the preset measured weights are pre-prepared weight values ​​corresponding to different fluid types, which represent the possibility that the corresponding fluid type may be the target fluid that the user wants to measure (the larger the preset measured weight, the greater the possibility).

[0200] In this embodiment, the measured evaluation value is the product of the preset measured weight and the corresponding total area.

[0201] The beneficial effects of the above technology are: considering the proportion of different types of fluids flowing through the detection pipeline and the preset measured weights corresponding to the fluid types, the measured evaluation value of the corresponding fluid type is determined, thereby realizing the consideration of the two influencing factors of the proportion of different types of fluids flowing through the detection pipeline and the possibility that the fluid type may be the target fluid that the user wants to measure, and then when different types of fluids flow through the detection pipeline at the same time, the type of fluid that the user really wants to detect can be accurately determined, and the flow detection data in the measuring pipeline can be analyzed based on the local video of the measured fluid divided in the fluid video based on the target measured type, providing a data basis for the subsequent accurate calculation of the flow rate of the measured fluid.

[0202] Embodiment 9:

[0203] On the basis of Example 8, the data analysis module refers to Fig. 9 ,include:

[0204] A displacement tracking unit, used for performing video tracking on each pixel point included in the local video of the measured fluid to obtain a corresponding tracking displacement record;

[0205] A flow rate determination unit is used to determine the real-time flow rate of the pixel point based on the tracking displacement record, and use the real-time flow rate of all pixel points contained in the local video of the measured fluid as flow detection data in the measurement pipeline.

[0206] In this embodiment, the tracking displacement record is a displacement record of each pixel point obtained by performing video tracking on each pixel point included in the local video of the measured fluid.

[0207] In this embodiment, the real-time flow velocity is the real-time flow velocity of the pixel point determined based on the tracking displacement record.

[0208] In this embodiment, the real-time flow velocity of the pixel point is determined based on the tracking displacement record, that is:

[0209] Based on the tracking displacement record, the displacement length value of the corresponding pixel point per unit time is determined as the real-time flow velocity of the pixel point.

[0210] The beneficial effect of the above technology is: by performing video tracking on each pixel point contained in the local video of the measured fluid, the real-time flow velocity of each pixel point can be determined, thereby providing a data basis for the subsequent accurate calculation of the flow velocity of the measured fluid.

[0211] Embodiment 10:

[0212] Based on Example 1, the flow calculation end refers to Fig.10 ,include:

[0213] A flow rate determination module, used to take the average value of the real-time flow rates of all pixels included in the flow detection data as the real-time comprehensive flow rate of the measured fluid in the measurement pipeline;

[0214] The flow calculation module is used to calculate the real-time flow value of the measured fluid based on the property information of the measuring pipeline and the real-time comprehensive flow velocity.

[0215] In this embodiment, the real-time integrated flow velocity is the average value of the real-time flow velocities of all pixels included in the flow detection data.

[0216] In this embodiment, the real-time flow value of the measured fluid is calculated based on the property information of the measuring pipeline and the real-time integrated flow velocity, including:

[0217] Determine, based on the attribute information of the measuring pipeline, a cross section perpendicular to the flow direction of the fluid in the space carrying the measured fluid at the setting position of the measuring pipeline;

[0218] The product of the cross section and the real-time integrated flow velocity is regarded as the real-time flow value of the measured fluid.

[0219] The beneficial effects of the above technology are: it realizes the real-time flow velocity of the pixel points determined based on video tracking, and combines the attribute information of the measuring pipeline to accurately calculate the real-time flow value of the measured fluid, which greatly reduces the dependence of the traditional flow detection calculation method on the pressure detection device, making the calculated flow value more accurate.

[0220] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. Intelligent flow calculation system, characterized in that: include: A type determination terminal, used to determine a set of fluid types contained in the measured fluid based on the acquired fluid video when the measured fluid flows through the measurement pipeline; The video analysis end is used to determine the target detected type in the fluid type set, and divide the detected fluid local video corresponding to the target detected type in the fluid video, and analyze the flow detection data based on the detected fluid local video, including: An evaluation calculation unit, used to determine the total area of ​​all fluid regions included in the fluid region set of each fluid type, and calculate a corresponding measured evaluation value based on a preset measured weight corresponding to the fluid type and the corresponding total area; A final determination unit, configured to take the fluid type corresponding to the maximum measured evaluation value in the fluid type set as the target measured type; A video segmentation module, used to segment the fluid video into a local video of the target measured type of the measured fluid; A data analysis module, used for analyzing the flow detection data in the measuring pipeline based on the local video of the measured fluid; The flow calculation end is used to calculate the real-time flow value of the measured fluid based on the flow detection data and the property information of the measurement pipeline.

2. The intelligent flow calculation system according to claim 1, characterized in that: The type determination terminal includes: A video acquisition module, used for acquiring a video of the fluid in the measuring pipe when the measured fluid flows through the measuring pipe; A position determination module, used to determine a pressure detection position based on the fluid video; The type determination module is used to obtain the pressure value at the pressure detection position, and determine the set of fluid types contained in the measured fluid based on the pressure value and the fluid pressure value range list.

3. The intelligent flow calculation system according to claim 2, characterized in that: The position determination module comprises: a boundary judgment unit, used to judge whether there is a fluid boundary in each video frame of the fluid video, and if so, to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region, otherwise, to determine a pressure detection position based on a preset detection position list; The position determination unit is used to determine the pressure detection position based on a partial fluid area included in the video frame where the fluid boundary exists.

4. The intelligent flow calculation system according to claim 3, characterized in that: The boundary judgment unit includes: A larger screening subunit is used to calculate a first number based on the total number of pixels contained in the video frame and a first preset proportion, sort the pixel values ​​of all pixels in the video frame from large to small to obtain a pixel value sequence of the video frame, aggregate the first several pixel values ​​in the pixel value sequence to obtain a larger pixel value set of the video frame, and determine a larger pixel representation value of the video frame based on the larger pixel value set; a boundary judgment subunit, configured to calculate a pixel value difference between the larger pixel representation value and a minimum pixel value in the video frame, and when the pixel value difference is greater than a pixel value difference threshold, determine that a fluid boundary exists in the video frame; otherwise, determine that no fluid boundary exists in the video frame, and determine a pressure detection position based on a preset detection position list; The region division subunit is used to divide the corresponding video frame into regions based on the fluid boundary to obtain a corresponding partial fluid region when it is determined that there is a fluid boundary in the video frame.

5. The intelligent flow calculation system according to claim 4, characterized in that: The area division subunit includes: A Gaussian filter subunit, configured to perform Gaussian filtering on the video frame to obtain a corresponding denoised image when it is determined that a fluid boundary exists in the video frame; a gradient determination subunit, configured to determine a horizontal pixel distribution function and a vertical pixel distribution function based on the pixel distribution data in the denoised image, and to determine a transverse gradient value and a longitudinal gradient value of each pixel in the denoised image based on a first-order derivative of the horizontal pixel distribution function and the vertical pixel distribution function and a coordinate value of each pixel, and to determine a comprehensive gradient value and a gradient angle value of the pixel based on the transverse gradient value and the longitudinal gradient value; an adjacent determination subunit, configured to determine, in a preset gradient angle list, a preset gradient angle having a minimum difference with the gradient angle value as a standard approximate gradient value of the pixel point, and determine a first adjacent pixel point adjacent to the pixel point in a positive direction of the standard approximate gradient value and a second adjacent pixel point adjacent to the pixel point in a negative direction of the standard approximate gradient value; a gradient comparison subunit, configured to retain the corresponding pixel when the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the first adjacent pixel and the comprehensive gradient value of the pixel is greater than the comprehensive gradient value of the second adjacent pixel; otherwise, set the corresponding pixel to zero, traverse and compare each pixel in the denoised image, and obtain a first image; a threshold extraction subunit, configured to identify an initial fluid boundary in the first image, determine a first threshold based on pixel distribution data of the initial fluid boundary, calculate a second threshold based on the first threshold and a preset ratio, and perform threshold extraction on the denoised image based on the first threshold and the second threshold, respectively, to obtain a corresponding second image and a third image; a position corresponding subunit, configured to select any first non-zero pixel point in the second image, and perform edge tracing in the second image with the first non-zero pixel point as a starting point until an end point of a contour line is determined, and determine a first pixel point at the same position corresponding to the end point of the contour line in the third image; The contour compensation subunit is used to determine whether there is a second non-zero pixel point adjacent to the first co-positioned pixel point in the third image. If so, the second co-positioned pixel point corresponding to the second non-zero pixel point in the second image will be determined, and the pixel value of the second co-positioned pixel point will be set to the pixel value of the second non-zero pixel point, and edge tracking will continue in the second image with the second co-positioned pixel point as the starting point. When there is no non-zero pixel point adjacent to the latest co-positioned pixel point in the third image, contour tracking will be stopped, and the fluid boundary of the corresponding video frame will be obtained, and the corresponding video frame will be divided into regions based on the fluid boundary to obtain the corresponding partial fluid region.

6. The intelligent flow calculation system according to claim 5, characterized in that: The position determination unit comprises: A matrix construction subunit, used to take each pixel point included in the partial fluid area in the fluid video as a central pixel point, and construct a local pixel matrix of the central pixel point based on the pixel value of the central pixel point and the pixel values ​​of the corresponding neighboring pixels; a comprehensive determination subunit, configured to determine a local gradient value of each pixel in the local pixel matrix, determine a local gradient matrix corresponding to the local pixel matrix based on the local gradient value, rotate the local gradient values ​​contained in the local pixel matrix in sequence according to the preset gradients contained in the preset gradient list, obtain a rotated gradient matrix corresponding to each preset gradient, and use an average matrix of all the rotated gradient matrices as the corresponding comprehensive rotated gradient matrix; A pixel screening subunit, used for treating corresponding pixel points as rotation invariant pixel points when the comprehensive rotation gradient matrix and the local gradient matrix are the same; A vector determination subunit, configured to determine the coordinate value of each rotationally invariant pixel point contained in the partial fluid region, and construct a corresponding local gradient vector based on the coordinate value of the rotationally invariant pixel point and the local gradient value; A weighted processing subunit, configured to perform Gaussian weighted processing on the local gradient vector based on the coordinate values ​​of all rotationally invariant pixels contained in the partial fluid region and the local gradient vector, so as to obtain a stable gradient vector corresponding to the local gradient vector; A difference determination subunit, configured to determine a corresponding comprehensive stable point feature vector based on all stable gradient vectors contained in the partial fluid region, and determine an angle difference of the comprehensive stable point feature vector between every two partial fluid regions in the fluid video; The classification and aggregation subunit is used to classify two partial fluid regions whose angle difference of the comprehensive stable point characteristic vector is less than the angle difference threshold as the same fluid type, until all partial fluid regions contained in the partial fluid region are divided, and the partial fluid regions belonging to the same fluid type are aggregated to obtain a fluid region set of each fluid type; a center determination subunit, configured to extract all video frames with fluid boundaries from the fluid video as fluid boundary video frames, and determine the contact center position of each partial fluid area based on the contact boundary with the measuring pipe in each partial fluid area in the fluid boundary video frames; The position determination subunit is used to determine the individual pressure detection position of the corresponding fluid type based on all contact center positions in the fluid area set, and to obtain the pressure detection position by summarizing the individual pressure detection positions of all fluid types.

7. The intelligent flow calculation system according to claim 6, characterized in that: The type determination module includes: an initial determination unit, configured to obtain a pressure value at the pressure detection position, and determine a fluid type of each fluid region set based on the pressure value and a list of fluid pressure value ranges; The type summarizing unit is used to summarize the fluid types of all fluid area sets to obtain a set of fluid types contained in the measured fluid.

8. The intelligent flow calculation system according to claim 1, characterized in that: The data analysis module comprises: A displacement tracking unit, used for performing video tracking on each pixel point included in the local video of the measured fluid to obtain a corresponding tracking displacement record; A flow rate determination unit is used to determine the real-time flow rate of the pixel point based on the tracking displacement record, and use the real-time flow rate of all pixel points contained in the local video of the measured fluid as flow detection data in the measurement pipeline.

9. The intelligent flow calculation system according to claim 1, characterized in that: The flow calculation end includes: A flow rate determination module, used to take the average value of the real-time flow rates of all pixels included in the flow detection data as the real-time comprehensive flow rate of the measured fluid in the measurement pipeline; The flow calculation module is used to calculate the real-time flow value of the measured fluid based on the property information of the measuring pipeline and the real-time comprehensive flow velocity.

Images