Sediment detection method, equipment and device based on machine vision

Abstract

The present invention discloses a sediment detection method, equipment and device based on machine vision, the method comprising: when it is determined that the sediment detection operation conditions are currently met, the pre-processing device collects the water to be detected mixed with the sediment to be detected according to the preset water extraction volume, and performs sediment sedimentation, drainage and parameter calculation on the water to be detected to obtain the pre-processing result corresponding to the water to be detected; the pre-processing result includes the filtered water corresponding to the water to be detected and the sediment concentration measurement result thereof; the detection device extracts the filtered water, and performs visual recognition on the filtered water according to the preset visual recognition algorithm to obtain the sediment concentration mixed with the filtered water; the pre-processing result and the visual recognition result are determined as the target detection result for the water to be detected. It can be seen that the implementation of the present invention can realize the comprehensive quantitative and qualitative measurement of sediment concentration, improve the intelligent degree of sediment detection, detection efficiency and detection accuracy.

Description

Technical Field

[0001] The present invention relates to the field of visual recognition technology, and in particular to a sediment detection method, equipment and device based on machine vision. Background Art

[0002] In the field of water quality monitoring and sediment control, accurate detection of sediment concentration is crucial for evaluating water quality, predicting water flow changes, and formulating effective sediment control strategies. However, traditional sediment detection methods often have problems such as long detection cycles, complex operations, and insufficient accuracy, which can hardly meet the urgent needs of modern water quality monitoring and sediment control.

[0003] Traditional sediment detection methods usually include multiple steps such as sampling and laboratory analysis. These steps are not only time-consuming and laborious, but also easily affected by human factors and environmental conditions, resulting in inaccurate test results. In addition, traditional methods can often only provide a single indicator of sediment concentration and cannot fully reflect the physical and chemical properties of sediment, thus limiting its application in sediment control and water quality monitoring. It can be seen that it is particularly important to provide a corresponding solution to the technical problems of low measurement efficiency and low measurement accuracy in existing sediment detection technologies. Summary of the invention

[0004] The present invention provides a sediment detection method, equipment and device based on machine vision, which can realize comprehensive quantitative and qualitative measurement of sediment concentration, and improve the intelligent degree, detection efficiency and detection accuracy of sediment.

[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a sediment detection method based on machine vision, the method is applied to a sediment detection device based on machine vision, the device includes a pre-processing device and a detection device, the method includes: When it is determined that the sediment detection operation conditions are currently met, the pre-processing device collects water to be detected according to a preset water extraction volume, and the water to be detected is mixed with sediment to be detected; The pre-processing device performs a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected; the pre-processing operation includes sediment sedimentation, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be detected, and the sediment concentration corresponding to the filtered water calculated based on the parameters, and the filtered water is mixed with the sediment to be detected; The detection device extracts the filtered water, and performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition includes multi-dimensional parameter recognition for the sediment to be detected, and the multi-dimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; The pre-processing result and the visual recognition result are determined as target detection results for the water to be detected.

[0006] As an optional implementation, in the first aspect of the present invention, the pre-processing device performs a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected, including: The pre-processing device performs a vibration sedimentation operation on the water to be tested, and collects vibration sedimentation data of the water to be tested in real time; The pre-processing device performs data comparison on the vibration sedimentation data according to a plurality of preset settlement indicators, and determines that the vibration sedimentation operation is completed when it is determined that the corresponding data comparison information indicates that the vibration sedimentation data meets the settlement requirement of each settlement indicator; The pre-processing device performs drainage processing on the water to be detected to obtain a drainage processing result corresponding to the water to be detected, wherein the drainage processing result includes filtered water after the preset impurities are filtered out, and the filtered water is a water-sediment mixture; The pre-processing device performs parameter calculation on the drainage processing result according to a preset density calculation formula to obtain a sediment density calculation result corresponding to the drainage processing result as a pre-processing result corresponding to the water to be detected.

[0007] As an optional embodiment, in the first aspect of the present invention, the plurality of preset sedimentation indicators at least include sedimentation rate of sediment particles, water clarity, sediment layer stability, and water quality indicators; the pre-processing device performs data comparison on the vibration sedimentation sediment data according to the plurality of preset sedimentation indicators, including: The pre-processing device analyzes the vibration sedimentation data to obtain real-time index data corresponding to each of the sedimentation indicators; For each of the settlement indicators, the pre-processing device compares the standard value corresponding to the settlement indicator with the corresponding real-time indicator data to obtain a numerical comparison result corresponding to the settlement indicator; and at the same time, the numerical comparison results corresponding to all the settlement indicators are determined as data comparison information corresponding to the vibration settlement sediment data; And, the data comparison information indicates that the settlement requirements for the vibration sedimentation data to meet each of the settlement indicators are specifically: The real-time indicator data corresponding to each of the settlement indicators is within the standard value corresponding to the settlement indicator.

[0008] As an optional implementation, in the first aspect of the present invention, the detection device performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water, including: The detection device detects a first preset parameter in the filtered water according to a preset visual recognition algorithm to obtain first parameter data corresponding to the first preset parameter; the first preset parameter includes sediment weight and sediment density; The detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to second preset parameters; the second preset parameters at least include sediment concentration; the second preset parameters also include turbidity and / or particle size and distribution; The detection device generates image recognition data corresponding to the filtered water according to the second parameter data; and determines the first parameter data, the second parameter data and the image recognition data as a visual recognition result corresponding to the filtered water.

[0009] As an optional implementation, in the first aspect of the present invention, the visual recognition algorithm includes an OpenCV algorithm; the detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, including: The detection device acquires a visual recognition image corresponding to the filtered water; The detection device extracts a first recognition feature of the visual recognition image according to the visual recognition algorithm, wherein the first recognition feature includes a color feature and / or a texture feature corresponding to the silt particles in the filtered water; and according to the visual recognition algorithm, performs sediment concentration prediction on the first recognition feature to obtain a predicted sediment concentration corresponding to the filtered water, and adds the predicted sediment concentration to the second parameter data corresponding to the second preset parameter; When the second preset parameter includes the turbidity, the detection device calculates the absolute value of the Laplace operator corresponding to the visual recognition image according to the visual recognition algorithm, and calculates the standard deviation of the absolute value to obtain the turbidity value corresponding to the visual recognition image, and adds the turbidity value to the second parameter data corresponding to the second preset parameter.

[0010] As an optional implementation, in the first aspect of the present invention, the detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, further comprising: When the second preset parameter includes the particle size and distribution, the detection device performs preprocessing on the visual recognition image to obtain a preprocessed image corresponding to the visual recognition image; the preprocessing includes any one of grayscale conversion, blurring and binarization; The detection device determines the particle edge of each target particle in the preprocessed image according to a set edge detection algorithm; and determines the particle contour of each target particle according to a set contour extraction algorithm; The detection device calculates the contour area of ​​each target particle according to a set contour calculation formula and in combination with the particle contour of each target particle; The detection device determines the particle size and distribution corresponding to the filtered water according to the particle edge and the contour area corresponding to each target particle, and adds the particle size and distribution to the second parameter data corresponding to the second preset parameter.

[0011] As an optional embodiment, in the first aspect of the present invention, the second preset parameter further includes sediment type and / or particle movement characteristics; The detecting device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, further comprising: When the second preset parameter includes the particle motion characteristics, the detection device determines the motion data of the target particles in the visual recognition image according to the set optical flow method or tracking algorithm, the motion data including at least one of the motion trajectory, motion speed, and motion acceleration of each target particle; and analyzes the motion data of all the target particles to obtain the motion characteristics of all the target particles, and adds the motion characteristics to the second parameter data corresponding to the second preset parameter; When the sediment type is determined, the detection device extracts a second identification feature corresponding to the visual identification image, wherein the second identification feature includes the color feature and / or the texture feature and / or the shape feature; and performs classification identification on the second identification feature according to the set classification model to obtain a sediment type identification result corresponding to the visual image, and adds the sediment type identification result to the second parameter data corresponding to the second preset parameter.

[0012] The second aspect of the present invention discloses a sediment detection device based on machine vision, the device comprising a pre-processing device and a detection device, wherein: The pre-processing device is used to collect water to be tested according to a preset water extraction volume when it is determined that the sediment detection operation conditions are currently met, and the water to be tested is mixed with sediment to be tested; The pre-processing device is further used to perform a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected; the pre-processing operation includes sediment sedimentation, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be detected, and the sediment concentration corresponding to the filtered water calculated based on the parameters, and the filtered water is mixed with the sediment to be detected; The detection device is used to extract the filtered water and perform visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition includes multi-dimensional parameter recognition for the sediment to be detected, and the multi-dimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; The detection device is further used to determine the pre-processing result and the visual recognition result as the target detection result for the water to be detected.

[0013] As an optional implementation, in the second aspect of the present invention, the pre-processing device performs a pre-processing operation on the water to be detected, and a method of obtaining a pre-processing result corresponding to the water to be detected specifically includes: Performing a vibration sedimentation operation on the water to be tested, and collecting vibration sedimentation data of the water to be tested in real time; Performing data comparison on the vibration sedimentation data according to a plurality of preset sedimentation indices, and determining that the vibration sedimentation operation is completed when it is determined that the corresponding data comparison information indicates that the vibration sedimentation data meets the sedimentation requirement of each of the sedimentation indices; Performing drainage processing on the water to be tested to obtain a drainage processing result corresponding to the water to be tested, wherein the drainage processing result includes filtered water after removing preset impurities, and the filtered water is a water-sediment mixture; According to a preset density calculation formula, parameter calculation is performed on the drainage treatment result to obtain a sediment density calculation result corresponding to the drainage treatment result as a pre-processing result corresponding to the water to be detected.

[0014] As an optional embodiment, in the second aspect of the present invention, the plurality of preset sedimentation indicators at least include sedimentation rate of sediment particles, water clarity, sediment layer stability, and water quality indicators; the pre-processing device performs data comparison on the vibration sedimentation sediment data according to the plurality of preset sedimentation indicators specifically including: Analyze the vibration sedimentation data to obtain real-time index data corresponding to each of the sedimentation indicators; For each of the settlement indicators, the standard value corresponding to the settlement indicator is compared with the corresponding real-time indicator data to obtain the numerical comparison result corresponding to the settlement indicator; at the same time, the numerical comparison results corresponding to all the settlement indicators are determined as the data comparison information corresponding to the vibration settlement sediment data; And, the data comparison information indicates that the settlement requirements for the vibration sedimentation data to meet each of the settlement indicators are specifically: The real-time indicator data corresponding to each of the settlement indicators is within the standard value corresponding to the settlement indicator.

[0015] As an optional embodiment, in the second aspect of the present invention, the detection device performs visual recognition on the filtered water according to a preset visual recognition algorithm, and the method of obtaining the visual recognition result corresponding to the filtered water specifically includes: According to a preset visual recognition algorithm, a first preset parameter in the filtered water is detected to obtain first parameter data corresponding to the first preset parameter; the first preset parameter includes sediment weight and sediment density; According to the visual recognition algorithm, visual recognition is performed on the filtered water to obtain second parameter data corresponding to second preset parameters; the second preset parameters at least include sediment concentration; the second preset parameters also include turbidity and / or particle size and distribution; According to the second parameter data, image recognition data corresponding to the filtered water is generated; and the first parameter data, the second parameter data and the image recognition data are determined as a visual recognition result corresponding to the filtered water.

[0016] As an optional implementation, in the second aspect of the present invention, the visual recognition algorithm includes an OpenCV algorithm; the detection device performs visual recognition on the filtered water according to the visual recognition algorithm, and the method of obtaining the second parameter data corresponding to the second preset parameter specifically includes: Acquire a visual recognition image corresponding to the filtered water; Extracting a first recognition feature of the visual recognition image according to the visual recognition algorithm, the first recognition feature including a color feature and / or a texture feature corresponding to the silt particles in the filtered water; and performing a silt concentration prediction on the first recognition feature according to the visual recognition algorithm to obtain a predicted silt concentration corresponding to the filtered water, and adding the predicted silt concentration to the second parameter data corresponding to the second preset parameter; When the second preset parameter includes the turbidity, according to the visual recognition algorithm, the absolute value of the Laplace operator corresponding to the visual recognition image is calculated, and the standard deviation of the absolute value is calculated to obtain the turbidity value corresponding to the visual recognition image, and the turbidity value is added to the second parameter data corresponding to the second preset parameter.

[0017] As an optional implementation, in the second aspect of the present invention, the detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: When the second preset parameter includes the particle size and distribution, preprocessing is performed on the visual recognition image to obtain a preprocessed image corresponding to the visual recognition image; the preprocessing includes any one of grayscale conversion, blurring and binarization; Determine the particle edge of each target particle in the preprocessed image according to a set edge detection algorithm; and determine the particle contour of each target particle according to a set contour extraction algorithm; According to the set contour calculation formula, combined with the particle contour of each target particle, the contour area of ​​each target particle is calculated; According to the particle edge and the contour area corresponding to each of the target particles, the particle size and distribution corresponding to the filtered water are determined, and the particle size and distribution are added to the second parameter data corresponding to the second preset parameter.

[0018] As an optional embodiment, in the second aspect of the present invention, the second preset parameter further includes sediment type and / or particle movement characteristics; The detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: When the second preset parameter includes the particle motion characteristics, the motion data of the target particles in the visual recognition image are determined according to the set optical flow method or tracking algorithm, the motion data including at least one of the motion trajectory, motion speed, and motion acceleration of each target particle; and the motion data of all the target particles are analyzed to obtain the motion characteristics of all the target particles, and the motion characteristics are added to the second parameter data corresponding to the second preset parameter; When the sediment type is determined, a second identification feature corresponding to the visual identification image is extracted, and the second identification feature includes the color feature and / or the texture feature and / or the shape feature; and classification identification is performed on the second identification feature according to the set classification model to obtain a sediment type identification result corresponding to the visual image, and the sediment type identification result is added to the second parameter data corresponding to the second preset parameter.

[0019] The third aspect of the present invention discloses a sediment detection device based on machine vision, the device comprising: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the machine vision-based sediment detection method disclosed in the first aspect of the present invention.

[0020] A fourth aspect of the present invention discloses a computer storage medium, wherein the computer storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the machine vision-based sediment detection method disclosed in the first aspect of the present invention.

[0021] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: In an embodiment of the present invention, a sediment detection method based on machine vision is provided. The method is applied to sediment detection equipment based on machine vision. The equipment includes a pre-processing device and a detection device. The method includes: when it is determined that the sediment detection operating conditions are currently met, the pre-processing device collects water to be detected according to a preset water extraction volume, and the water to be detected is mixed with sediment to be detected; the pre-processing device performs a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected; the pre-processing operation includes sediment sedimentation, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be detected, and the filtered water is mixed with sediment to be detected; the detection device extracts the filtered water, and performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition result at least includes the sediment concentration of the mixed sediment in the filtered water; the pre-processing result and the visual recognition result are determined as the target detection result for the water to be detected. It can be seen that the implementation of the present invention realizes efficient and accurate detection of sediment content in water bodies by integrating the pre-treatment device and the detection device. Specifically: by setting the sediment detection operating conditions and automatically triggering the sediment detection when the conditions are met, it is beneficial to improve the sediment detection efficiency; and the sediment sedimentation, drainage and parameter calculation steps performed by the pre-treatment device effectively separate the water body and the sediment, providing a clearer and less-interfered sample for subsequent visual recognition, and by accurately controlling the experimental conditions, it is beneficial to improve the reliability of the visual recognition results; at the same time, the parameter calculation realizes the quantitative measurement of the sediment concentration based on the pre-treatment device; further, after the filtered water is obtained by the pre-treatment device, the detection device can accurately analyze the sediment concentration in the filtered water by using the advanced visual recognition algorithm, which is different from the traditional detection method and can provide higher resolution and more precise sediment content data. At the same time, the qualitative measurement of the sediment concentration is realized based on the detection device; the target detection result for the water to be detected is finally obtained by combining the pre-treatment result and the visual recognition result, realizing the quantitative and qualitative measurement of the water to be detected, while improving the precision of the target detection result. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a flow chart of a sediment detection method based on machine vision disclosed in an embodiment of the present invention; Figure 2 It is a flow chart of another sediment detection method based on machine vision disclosed in an embodiment of the present invention; Figure 3 It is a structural schematic diagram of a sediment detection device based on machine vision disclosed in an embodiment of the present invention; Figure 4 is a schematic structural diagram of another machine vision-based sediment detection device disclosed in an embodiment of the present invention; Figure 5 It is a structural schematic diagram of a sediment detection device based on machine vision disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0026] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0027] The invention discloses a sediment detection method, equipment and device based on machine vision. By integrating a pre-processing device and a detection device, efficient and accurate detection of sediment content in a water body is achieved. Specifically, by setting sediment detection operation conditions and automatically triggering sediment detection when the conditions are met, the sediment detection efficiency is improved. In addition, the sediment settling, drainage and parameter calculation steps performed by the pre-processing device effectively separate the water body and the sediment, providing a clearer sample with less interference for subsequent visual recognition. By accurately controlling the experimental conditions, the reliability of the visual recognition result is improved. At the same time, the parameter calculation realizes quantitative measurement of sediment concentration based on the pre-processing device. Further, after the filtered water is obtained by the pre-processing device, the detection device can accurately analyze the sediment concentration in the filtered water by using an advanced visual recognition algorithm. Different from the traditional detection method, it can provide higher resolution and more precise sediment content data. At the same time, qualitative measurement of sediment concentration is achieved based on the detection device. The target detection result for the water to be detected is finally obtained by integrating the pre-processing result and the visual recognition result, realizing quantitative and qualitative measurement of the water to be detected, while improving the precision of the target detection result. The following are detailed descriptions of each.

[0028] Embodiment 1 See also Figure 1 , Figure 1 : is a flow chart of a sediment detection method based on machine vision disclosed in an embodiment of the present invention. Figure 1 The described machine vision-based sediment detection method can be applied to a machine vision-based sediment detection device / apparatus, which at least includes a pre-processing device and a detection device. Figure 1 As shown, the sediment detection method based on machine vision may include the following operations: 101. When it is determined that the current operating conditions for sediment detection are met, the pre-processing device collects water to be detected according to a preset water extraction volume, and the water to be detected is mixed with sediment to be detected.

[0029] In the embodiment of the present invention, the current sediment detection operation conditions are specifically: The pumping component of the sediment detection equipment, such as a pumping pump, is started, and the pumping component operates normally; at the same time, the pumping pipe corresponding to the pumping component is connected to the water source to be pumped, such as any river channel that allows pumping; Specifically, when the pumping pipe is connected to the river channel and the pump is started, when the pumping pipe can stably extract river water from the river channel, it is determined that the current operating conditions for sediment detection are met.

[0030] In the embodiment of the present invention, the preset pumping water volume generally controls the pumping component to perform pumping control at a pumping speed of 1 L / min.

[0031] In an embodiment of the present invention, if it is detected that the sediment detection operating conditions are not met at present, it is necessary to perform status detection and adjustment on the pumping components. For example, if the pump is broken, the components need to be repaired or replaced; if the pumping pipe is damaged, the pumping pipe needs to be replaced; if the pumping volume does not meet the standard, it is necessary to adjust the connection position of the pumping pipe and the river channel, generally controlling the pumping pipe to be connected to a position half the height of the overall river channel; or adjust the pumping power of the pump.

[0032] 102. The pre-processing device performs a pre-processing operation on the water to be tested to obtain a pre-processing result corresponding to the water to be tested.

[0033] In the embodiment of the present invention, the pre-processing operation includes sediment settling, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be tested, the sediment concentration corresponding to the filtered water calculated based on the parameters, and the filtered water is mixed with the sediment to be tested.

[0034] 103. The detection device extracts filtered water and performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water.

[0035] In an embodiment of the present invention, the visual recognition includes multidimensional parameter recognition for the sediment to be detected, and the multidimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; correspondingly, the visual recognition result corresponding to the filtered water includes at least visual recognition results for sediment density and sediment concentration of mixed sediment in the filtered water.

[0036] 104. Determine the pre-processing result and the visual recognition result as the target detection result for the water to be detected.

[0037] It can be seen that implementation Figure 1The described machine vision-based sediment detection method realizes efficient and accurate detection of sediment content in water bodies by integrating a pre-processing device and a detection device. Specifically, by setting sediment detection operating conditions and automatically triggering sediment detection when the conditions are met, it is beneficial to improve the sediment detection efficiency; and the sediment settling, drainage and parameter calculation steps performed by the pre-processing device effectively separate the water body and sediment, providing a clearer and less-interfered sample for subsequent visual recognition, and by accurately controlling the experimental conditions, it is beneficial to improve the reliability of the visual recognition results; at the same time, the parameter calculation realizes quantitative measurement of sediment concentration based on the pre-processing device; further, after the filtered water is processed by the pre-processing device, the detection device can accurately analyze the sediment concentration in the filtered water by using an advanced visual recognition algorithm, which is different from the traditional detection method and can provide higher resolution and more precise sediment content data. At the same time, the qualitative measurement of sediment concentration is realized based on the detection device; the target detection result for the water to be detected finally obtained combines the pre-processing result and the visual recognition result, realizes quantitative and qualitative measurement of the water to be detected, and improves the precision of the target detection result.

[0038] In an optional embodiment, the above step 102 is performed by the pre-processing device on the water to be tested. The method of obtaining the pre-processing result corresponding to the water to be tested specifically includes: The pre-processing device performs a vibration sedimentation operation on the water to be tested, and collects vibration sedimentation data of the water to be tested in real time; The pre-processing device performs data comparison on the vibration sedimentation data according to a plurality of preset settlement indicators, and determines that the vibration sedimentation operation is completed when it is determined that the corresponding data comparison information indicates that the vibration sedimentation data meets the settlement requirement of each settlement indicator; The pre-treatment device performs drainage treatment on the water to be tested to obtain drainage treatment results corresponding to the water to be tested, the drainage treatment results include filtered water after the preset impurities are filtered out, and the filtered water is a water-sediment mixture; The pre-processing device performs parameter calculation on the drainage processing result according to a preset density calculation formula to obtain a sediment density calculation result corresponding to the drainage processing result as a pre-processing result corresponding to the water to be detected.

[0039] In this optional embodiment, the vibration sedimentation operation performed by the pre-processing device can accelerate the sedimentation speed of sediment particles in the water body, significantly shorten the sedimentation time compared to natural sedimentation, and help improve the detection efficiency. At the same time, real-time collection of vibration sedimentation data ensures the controllability and monitorability of the sedimentation process. In addition, in specific application scenarios, the vibration sedimentation operation can also be used to separate the required water-sediment mixture from impurities to achieve the effect of impurity filtration.

[0040] In this optional embodiment, when the data comparison information shows that the vibration sedimentation data meets the sedimentation requirements of each sedimentation index, it can be determined that the vibration sedimentation operation is completed. This step ensures the sufficiency and uniformity of sediment settlement.

[0041] In this optional embodiment, through the drainage treatment of the pre-treatment device, preset impurities can be accurately filtered out to obtain filtered water, which not only removes interference factors that affect visual recognition, but also retains key sediment information, providing a clear sample for subsequent visual recognition.

[0042] In this optional embodiment, the density calculation formula is specifically:

[0043] in, is the overall density of the water to be tested, is the mass of the sediment, is the total volume of sediment and water in the water to be tested; If the densities of sediment and water are considered to be respectively, the mixed density formula can be used:

[0044] Among them, it can be calculated by the volume and density of water; If the volume fraction of sediment in water is known ( ), the density can be calculated using the following formula:

[0045] in, is the density of the sediment; is the density of water.

[0046] It can be seen that in this optional embodiment, by performing a vibration sedimentation operation on the water to be tested, it is beneficial to accelerate the settling rate of sediment particles in the water body and improve the detection efficiency; the real-time collection of vibration sedimentation data ensures the controllability and monitorability of the sedimentation process; the vibration sedimentation data is compared according to a plurality of preset sedimentation indicators to achieve precise control of the sedimentation process; and the pre-processing device performs parameter calculation on the drainage treatment result based on a preset density calculation formula, which not only provides a direct quantitative indicator of sediment concentration, but also enhances the accuracy and reliability of the parameter calculation results through precise parameter calculation.

[0047] In another optional embodiment, the plurality of preset sedimentation indicators at least include sediment particle settling rate, water clarity, sediment layer stability, and water quality indicators; The above-mentioned method of performing data comparison on the vibration settlement sediment data by the pre-processing device according to a plurality of preset settlement indicators specifically includes: The pre-processing device analyzes the vibration sedimentation data to obtain real-time index data corresponding to each sedimentation index; For each settlement index, the pre-processing device compares the standard value corresponding to the settlement index with its corresponding real-time index data to obtain the numerical comparison result corresponding to the settlement index; at the same time, the numerical comparison results corresponding to all settlement indexes are determined as data comparison information corresponding to the vibration settlement sediment data; In addition, the data comparison information indicates that the settlement requirements for the vibration sedimentation data to achieve each settlement index are as follows: The real-time indicator data corresponding to each settlement indicator is within the standard value corresponding to the settlement indicator.

[0048] In this optional embodiment, for the sedimentation rate of sediment particles in the sedimentation index: in the initial stage of the vibration sedimentation operation, the sedimentation rate is fast because the sediment particles are loosely distributed in the water; as the sedimentation proceeds, the particles gradually accumulate and become compact, and the sedimentation rate will slow down significantly; when the sedimentation rate drops to almost zero or remains stable and no longer changes, it can be considered that the sedimentation process is complete.

[0049] In this optional embodiment, for the water clarity in the sedimentation index: during the sedimentation process, as the sediment particles gradually sink, the suspended matter in the water body decreases and the clarity gradually increases; when the water body becomes relatively clear and there is no obvious increase in suspended matter for a long time, it can be considered that the sedimentation effect is good and the sedimentation process is completed.

[0050] In this optional embodiment, for the stability of the sediment layer in the sedimentation index: during the vibration sedimentation process, the silt particles will gradually accumulate to form a sediment layer. As the sedimentation is completed, the thickness and shape of the sediment layer will tend to be stable and no longer change significantly. By observing and recording the stability and thickness changes of the sediment layer, it can be determined whether the sedimentation is complete.

[0051] In this optional embodiment, for the water quality index in the sedimentation index: during the sedimentation process, the suspension and sinking of sediment particles may have a certain impact on the water quality. By monitoring the changes in water quality indicators (such as turbidity, suspended matter concentration, etc.), the impact of sedimentation on water quality can be determined. When the water quality index returns to normal levels or reaches the expected standard, it can be considered that the sedimentation process has been completed.

[0052] In this optional embodiment, the real-time indicator data corresponding to each of the above-mentioned settlement indicators being within the standard value corresponding to the settlement indicator specifically includes: The sedimentation rate of the sediment particles is the set minimum sedimentation rate value, the water clarity reaches the set standard clarity value, the stability of the sediment layer reaches the standard stability, and the water quality indicators reach the standard water quality values.

[0053] In this optional embodiment, the minimum sedimentation rate value can be set according to actual needs, such as 0.1mm / s; the water clarity can be set to low clarity, medium clarity and high clarity; before the initial vibration sedimentation sedimentation operation is performed, the water clarity of the water to be tested is defaulted to low clarity. If after the vibration sedimentation sedimentation operation is performed, the water clarity can rise to high clarity relative to the initial water to be tested, the water clarity meets the standard; similar to the water clarity, the sediment layer stability can be divided into low stability, medium stability and high stability; the corresponding low stability corresponds to the low thickness of the sediment layer, the medium stability corresponds to the medium thickness of the sediment layer, and the high stability corresponds to the high thickness of the sediment layer; the water quality index can be set to four indicators, namely, poor, qualified, medium and excellent. Correspondingly, when the water quality index of the water to be tested can reach a rating of medium or above after the vibration sedimentation sedimentation operation is performed, the water quality index can be considered to meet the standard.

[0054] It can be seen that in this optional embodiment, by setting multiple sedimentation indicators including the sedimentation rate of sediment particles, water clarity, sediment layer stability and water quality indicators, the entire process of vibration sedimentation can be fully monitored; these indicators not only cover the physical sedimentation characteristics of sediment particles, but also take into account the overall state of water bodies and sediments, thereby ensuring the comprehensiveness and accuracy of monitoring the sedimentation process; in addition, through real-time data analysis and comparison of vibration sedimentation sediment data, intuitive presentation of data comparison information and accurate judgment of sedimentation requirements, it is beneficial to improve the detection efficiency and accuracy of sedimentation.

[0055] In yet another optional embodiment, before the above step 104 determines the pre-processing result and the visual recognition result as the target detection result for the water to be detected, the method further includes: The sediment concentration corresponding to the filtered water is selected from the pre-processing result and recorded as the first sediment concentration; the sediment concentration corresponding to the filtered water is extracted from the visual recognition result and recorded as the second sediment concentration; Calculate the deviation result corresponding to the first sediment concentration and the second sediment concentration according to the set deviation calculation formula; Compare the deviation result with the pre-calculated deviation threshold to obtain a corresponding comparison result; when the deviation result indicates that it is greater than the deviation threshold, mark the first sediment concentration and the second sediment concentration as abnormal data, and the abnormal data is used to indicate that data needs to be removed and / or to indicate that the above steps 101 to 103 are re-executed for the water to be tested; When the deviation result indicates that the deviation is less than or equal to the deviation threshold, step 104 is triggered.

[0056] The specific deviation calculation formula is: Deviation result = |first sediment concentration - second sediment concentration|, or, Deviation result = (|first sediment concentration - second sediment concentration| / first sediment concentration) * 100% It should be noted that the pre-calculated deviation threshold may be fixed or dynamically calculated based on historical data, which is not limited in this optional embodiment.

[0057] It should be noted that the pre-processing result and the visual recognition result correspond to the data of the current detection process, that is, the pre-processing result and the visual recognition result are single calculation data; the above-mentioned operation process of steps 101 to 103 can be repeated as many times as needed, and the calculation process of the data deviation can be executed each time. After determining that the data is correct (less than or equal to the deviation threshold), step 104 is triggered to end the current concentration calculation process.

[0058] It can be seen that in this optional embodiment, a data verification process is set up for the pre-processing results and the visual recognition results, the deviation value of the two results is calculated, and the data greater than the deviation threshold is eliminated, thereby achieving accurate screening of abnormal / erroneous data, which is beneficial to improving the accuracy and reliability of the target detection results obtained.

[0059] Embodiment 2 See also Figure 2 , Figure 2 is a flow chart of another method for detecting sediment based on machine vision disclosed in an embodiment of the present invention. Figure 2 The described machine vision-based sediment detection method can be applied to a machine vision-based sediment detection device, which is not limited in the embodiments of the present invention. Figure 2 As shown, the sediment detection method based on machine vision may include the following operations: 201. When it is determined that the current operating conditions for sediment detection are met, the pre-processing device collects water to be detected according to a preset water extraction volume, and the water to be detected is mixed with sediment to be detected.

[0060] 202. The pre-processing device performs a pre-processing operation on the water to be tested to obtain a pre-processing result corresponding to the water to be tested.

[0061] 203. The detection device extracts filtered water and detects a first preset parameter in the filtered water according to a preset visual recognition algorithm to obtain first parameter data corresponding to the first preset parameter.

[0062] In the embodiment of the present invention, the first preset parameter includes sediment weight and sediment density.

[0063] 204. The detection device performs visual recognition on the filtered water according to a visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter.

[0064] In the embodiment of the present invention, the second preset parameter at least includes sediment concentration; the second preset parameter also includes turbidity and / or particle size and distribution.

[0065] 205. The detection device generates image recognition data corresponding to the filtered water according to the second parameter data; and determines the first parameter data, the second parameter data and the image recognition data as a visual recognition result corresponding to the filtered water.

[0066] 206. Determine the pre-processing result and the visual recognition result as the target detection result for the water to be detected.

[0067] In the embodiment of the present invention, for other descriptions of step 201-step 202 and step 206, please refer to other specific descriptions of step 101-step 102 and step 104 in embodiment 1, which will not be repeated in the embodiment of the present invention.

[0068] It can be seen that the implementation Figure 2 The described machine vision-based sediment detection method can simultaneously detect multiple preset parameters in filtered water through a visual recognition algorithm preset in a detection device. This multi-dimensional parameter detection method provides comprehensive information on sediment characteristics, which is beneficial to improving the accuracy of water quality and sediment content assessment. Among them, the sediment feature recognition based on the visual recognition algorithm improves the accuracy and reliability of sediment detection for filtered water. In addition, the detection device can not only output specific parameter data, but also generate image recognition data corresponding to the filtered water. The image data provides intuitive sediment distribution and morphology information, which helps operators to understand water quality and sediment characteristics more intuitively, and improves the readability and practicality of the detection results.

[0069] In an optional embodiment, the visual recognition algorithm includes an OpenCV algorithm; In step 204, the detection device performs visual recognition on the filtered water according to the visual recognition algorithm, and the method of obtaining the second parameter data corresponding to the second preset parameter specifically includes: The detection device obtains a visual recognition image corresponding to the filtered water; The detection device extracts a first recognition feature of the visual recognition image according to a visual recognition algorithm, wherein the first recognition feature includes a color feature and / or a texture feature corresponding to the silt particles in the filtered water; and performs a silt concentration prediction on the first recognition feature according to the visual recognition algorithm to obtain a predicted silt concentration corresponding to the filtered water, and adds the predicted silt concentration to the second parameter data corresponding to the second preset parameter; When the second preset parameter includes turbidity, the detection device calculates the absolute value of the Laplace operator corresponding to the visual recognition image according to the visual recognition algorithm, and calculates the standard deviation of the absolute value to obtain the turbidity value corresponding to the visual recognition image, and adds the turbidity value to the second parameter data corresponding to the second preset parameter.

[0070] In this optional embodiment, the detection device achieves accurate identification of sediment particles by extracting the first identification features in the visual recognition image, including color features and / or texture features corresponding to the sediment particles. This not only improves the accuracy of sediment detection, but also facilitates the subsequent prediction of sediment concentration and calculation of turbidity.

[0071] In this optional embodiment, when performing sediment concentration prediction, the detection device can combine the powerful computing power of the visual recognition algorithm while considering the color and texture characteristics of the sediment particles, thereby ensuring the accuracy of the sediment concentration prediction.

[0072] In this optional embodiment, the detection device obtains the turbidity value corresponding to the visual recognition image by calculating the absolute value and standard deviation of the Laplace operator corresponding to the visual recognition image. This calculation method avoids the influence of human factors on turbidity assessment and ensures the objectivity and accuracy of turbidity calculation.

[0073] It can be seen that in this optional embodiment, the efficiency and intelligence level of sediment detection are significantly improved by efficiently utilizing the OpenCV algorithm, accurately extracting sediment characteristics, accurately predicting sediment concentration, and objectively calculating turbidity.

[0074] In another optional embodiment, the above step 204 is performed by the detection device according to the visual recognition algorithm to perform visual recognition on the filtered water, and the method of obtaining the second parameter data corresponding to the second preset parameter specifically includes: When the second preset parameter includes particle size and distribution, the detection device performs preprocessing on the visual recognition image to obtain a preprocessed image corresponding to the visual recognition image; the preprocessing includes any one of grayscale conversion, blurring and binarization; The detection device determines the particle edge of each target particle in the preprocessed image according to a set edge detection algorithm; and determines the particle contour of each target particle according to a set contour extraction algorithm; The detection device calculates the contour area of ​​each target particle according to the set contour calculation formula and the particle contour of each target particle; The detection device determines the particle size and distribution corresponding to the filtered water according to the particle edge and contour area corresponding to each target particle, and adds the particle size and distribution to the second parameter data corresponding to the second preset parameter.

[0075] In this optional embodiment, the preprocessing step for the visual recognition image helps to reduce image noise and enhance the contrast between the target particles (i.e., sediment particles) and the background, thereby improving the accuracy of subsequent edge detection and contour extraction.

[0076] In this optional embodiment, by setting the edge detection algorithm, the detection device can accurately determine the particle edge of each target particle in the preprocessed image. Further, by using the contour extraction algorithm, the detection device can accurately depict the particle contour of each target particle. This step provides a reliable basis for the subsequent particle size and distribution calculation.

[0077] In this optional embodiment, by integrating information such as the particle edge and contour area of ​​each target particle, the detection device can comprehensively analyze the size and distribution of particles in the filtered water; this analysis method not only provides quantitative indicators of the particles, but also reveals the distribution of particles in space, providing more detailed information for water quality assessment and sediment management.

[0078] It can be seen that in this optional embodiment, the efficiency and intelligence level of sediment detection are significantly improved through image preprocessing, precise edge detection and contour extraction, precise particle size calculation and comprehensive particle distribution analysis.

[0079] In yet another optional embodiment, the second preset parameter further includes sediment type and / or particle movement characteristics; In step 204, the detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: When the second preset parameter includes particle motion characteristics, the detection device determines motion data of target particles in the visual recognition image according to a set optical flow method or tracking algorithm, the motion data including at least one of the motion trajectory, motion speed, and motion acceleration of each target particle; and analyzes the motion data of all target particles to obtain the motion characteristics of all target particles, and adds the motion characteristics to the second parameter data corresponding to the second preset parameter; When the sediment type is detected, the detection device extracts a second identification feature corresponding to the visual identification image, the second identification feature including a color feature and / or a texture feature and / or a shape feature; and performs classification identification on the second identification feature according to the set classification model to obtain a sediment type identification result corresponding to the visual image, and adds the sediment type identification result to the second parameter data corresponding to the second preset parameter.

[0080] In this optional embodiment, the detection device can accurately capture the motion data of target particles in the visual recognition image by applying the optical flow method or tracking algorithm. This ability of precise capture and analysis helps to deeply understand the dynamic behavior of sediment particles in water.

[0081] In this optional embodiment, the detection device can accurately determine the type of sediment by extracting the second recognition feature (including color, texture, shape, etc.) in the visual recognition image and using the classification model for classification and recognition; this recognition method is not only fast and efficient, but also has high accuracy, which helps to improve the recognition speed and processing efficiency for different types of sediments.

[0082] It can be seen that in this optional embodiment, by introducing the detection of particle movement characteristics and sediment types, the comprehensiveness and practicality of sediment detection technology are significantly improved, and it is also beneficial to improve the comprehensiveness and accuracy of monitoring and analysis of sediment characteristics.

[0083] Embodiment 3 See also Figure 3 , Figure 3 : is a schematic diagram of the structure of a machine vision-based sediment detection device disclosed in an embodiment of the present invention. The machine vision-based sediment detection device can be a machine vision-based sediment detection terminal, device, system or server. The server can be a local server, a remote server, or a cloud server (also known as a cloud server). When the server is a non-cloud server, the non-cloud server can communicate with the cloud server, which is not limited in the embodiment of the present invention. Figure 3 As shown, the machine vision-based sediment detection device may include a pre-processing device 301 and a detection device 302, wherein: The pre-processing device 301 is used to collect water to be detected according to a preset water extraction volume when it is determined that the sediment detection operation conditions are currently met, and the water to be detected is mixed with the sediment to be detected.

[0084] The pre-processing device 301 is also used to perform pre-processing operations on the water to be tested to obtain pre-processing results corresponding to the water to be tested; the pre-processing operations include sediment sedimentation, drainage and parameter calculation; the pre-processing results include filtered water corresponding to the water to be tested, sediment concentration corresponding to the filtered water calculated based on parameters, and the filtered water is mixed with sediment to be tested.

[0085] The detection device 302 is used to extract filtered water and perform visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition includes multi-dimensional parameter recognition for the sediment to be detected, and the multi-dimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; correspondingly, the visual recognition result corresponding to the filtered water includes at least visual recognition results of sediment density and sediment concentration for the mixed sediment in the filtered water.

[0086] The detection device 302 is also used to determine the pre-processing result and the visual recognition result as the target detection result for the water to be detected.

[0087] See also Figure 4 , Figure 4 FIG. 1 is a schematic diagram of another structure of a sediment detection device based on machine vision disclosed in an embodiment of the present invention. Figure 4 As shown, the sediment detection equipment includes a pre-treatment device and a detection device, wherein the pre-treatment device is used to extract the sampling liquid to be detected (the above-mentioned water to be detected), and control the vibration component in the pre-treatment device to start working, so that the sediment in the water to be detected including the sediment mixture is settled, and then the pressure detection component arranged in the pre-treatment device is used to perform weighing calculation, combined with the volume parameters of the sampling liquid, to finally realize the sediment density detection function.

[0088] More specifically, the silt-sand mixture can also be filtered out after impurities other than silt are settled through a vibrating component, and the silt-sand mixture after impurities are filtered out is then transported to the detection device through a set transmission pipeline, so that the detection device can perform visual inspection and identification on the sampled liquid.

[0089] It can be seen that implementation Figure 3The described machine vision-based sediment detection equipment realizes efficient and accurate detection of sediment content in water bodies by integrating a pre-processing device and a detection device. Specifically, by setting sediment detection operating conditions and automatically triggering sediment detection when the conditions are met, it is beneficial to improve the sediment detection efficiency; and the sediment sedimentation, drainage and parameter calculation steps performed by the pre-processing device effectively separate the water body and sediment, providing a clearer and less-interfered sample for subsequent visual recognition, and by accurately controlling the experimental conditions, it is beneficial to improve the reliability of the visual recognition results; at the same time, the parameter calculation realizes quantitative measurement of sediment concentration based on the pre-processing device; further, after the filtered water is processed by the pre-processing device, the detection device can accurately analyze the sediment concentration in the filtered water by using an advanced visual recognition algorithm, which is different from the traditional detection method and can provide higher resolution and more precise sediment content data. At the same time, the qualitative measurement of sediment concentration is realized based on the detection device; the target detection result for the water to be detected finally obtained combines the pre-processing result and the visual recognition result, realizes quantitative and qualitative measurement of the water to be detected, and improves the precision of the target detection result.

[0090] In an optional embodiment, the pre-processing device 301 performs a pre-processing operation on the water to be detected, and a method of obtaining a pre-processing result corresponding to the water to be detected specifically includes: Perform vibration sedimentation operation on the water to be tested, and collect vibration sedimentation data of the water to be tested in real time; Performing data comparison on the vibration sedimentation data according to a plurality of preset settlement indicators, and determining that the vibration sedimentation operation is completed when it is determined that the corresponding data comparison information indicates that the vibration sedimentation data meets the settlement requirement of each settlement indicator; Performing drainage treatment on the water to be tested to obtain drainage treatment results corresponding to the water to be tested, the drainage treatment results including filtered water after the preset impurities are filtered out, the filtered water being a water-sediment mixture; According to the preset density calculation formula, parameter calculation is performed on the drainage treatment result to obtain the sediment density calculation result corresponding to the drainage treatment result as the pre-processing result corresponding to the water to be tested.

[0091] It can be seen that in this optional embodiment, by performing a vibration sedimentation operation on the water to be tested, it is beneficial to accelerate the settling rate of sediment particles in the water body and improve the detection efficiency; the real-time collection of vibration sedimentation data ensures the controllability and monitorability of the sedimentation process; the vibration sedimentation data is compared according to a plurality of preset sedimentation indicators to achieve precise control of the sedimentation process; and the pre-processing device performs parameter calculation on the drainage treatment result based on a preset density calculation formula, which not only provides a direct quantitative indicator of sediment concentration, but also enhances the accuracy and reliability of the parameter calculation results through precise parameter calculation.

[0092] In another optional embodiment, the plurality of preset sedimentation indicators at least include sediment particle settling rate, water clarity, sediment layer stability, and water quality indicators; The above-mentioned pre-processing device 301 performs data comparison on the vibration sedimentation data according to a plurality of preset settlement indicators, specifically including: Analyze the vibration sedimentation data to obtain real-time index data corresponding to each sedimentation index; For each settlement index, the standard value corresponding to the settlement index is compared with the corresponding real-time index data to obtain the numerical comparison result corresponding to the settlement index; at the same time, the numerical comparison results corresponding to all settlement indexes are determined as data comparison information corresponding to the vibration settlement sediment data; In addition, the data comparison information indicates that the settlement requirements for the vibration sedimentation data to achieve each settlement index are as follows: The real-time indicator data corresponding to each settlement indicator is within the standard value corresponding to the settlement indicator.

[0093] It can be seen that in this optional embodiment, by setting multiple sedimentation indicators including the sedimentation rate of sediment particles, water clarity, sediment layer stability and water quality indicators, the entire process of vibration sedimentation can be fully monitored; these indicators not only cover the physical sedimentation characteristics of sediment particles, but also take into account the overall state of water bodies and sediments, thereby ensuring the comprehensiveness and accuracy of monitoring the sedimentation process; in addition, through real-time data analysis and comparison of vibration sedimentation sediment data, intuitive presentation of data comparison information and accurate judgment of sedimentation requirements, it is beneficial to improve the detection efficiency and accuracy of sedimentation.

[0094] In another optional embodiment, the detection device 302 performs visual recognition on the filtered water according to a preset visual recognition algorithm, and the method of obtaining the visual recognition result corresponding to the filtered water specifically includes: According to a preset visual recognition algorithm, a first preset parameter in the filtered water is detected to obtain first parameter data corresponding to the first preset parameter; the first preset parameter includes sediment weight and sediment density; Perform visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter; the second preset parameter at least includes sediment concentration; the second preset parameter also includes turbidity and / or particle size and distribution; According to the second parameter data, image recognition data corresponding to the filtered water is generated; and the first parameter data, the second parameter data and the image recognition data are determined as visual recognition results corresponding to the filtered water.

[0095] It can be seen that in this optional embodiment, through the visual recognition algorithm preset by the detection device, multiple preset parameters in the filtered water can be detected simultaneously. This multi-dimensional parameter detection method provides comprehensive information on sediment characteristics, which is beneficial to improving the accuracy of the assessment of water quality conditions and sediment content; wherein, the sediment feature recognition based on the visual recognition algorithm improves the accuracy and reliability of sediment detection for filtered water; and, the detection device can not only output specific parameter data, but also generate image recognition data corresponding to the filtered water. The image data provides intuitive sediment distribution and morphology information, which helps operators to understand water quality conditions and sediment characteristics more intuitively, and improves the readability and practicality of the detection results.

[0096] In another optional embodiment, the visual recognition algorithm includes an OpenCV algorithm; The detection device 302 performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: Obtaining a visual recognition image corresponding to filtered water; Extracting a first recognition feature of the visual recognition image according to the visual recognition algorithm, the first recognition feature including a color feature and / or a texture feature corresponding to the sediment particles in the filtered water; and performing sediment concentration prediction on the first recognition feature according to the visual recognition algorithm to obtain a predicted sediment concentration corresponding to the filtered water, and adding the predicted sediment concentration to the second parameter data corresponding to the second preset parameter; When the second preset parameter includes turbidity, according to the visual recognition algorithm, the absolute value of the Laplace operator corresponding to the visual recognition image is calculated, and the standard deviation of the absolute value is calculated to obtain the turbidity value corresponding to the visual recognition image, and the turbidity value is added to the second parameter data corresponding to the second preset parameter.

[0097] It can be seen that in this optional embodiment, the efficiency and intelligence level of sediment detection are significantly improved by efficiently utilizing the OpenCV algorithm, accurately extracting sediment characteristics, accurately predicting sediment concentration, and objectively calculating turbidity.

[0098] In yet another optional embodiment, the detection device 302 performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: When the second preset parameter includes particle size and distribution, preprocessing is performed on the visual recognition image to obtain a preprocessed image corresponding to the visual recognition image; the preprocessing includes any one of grayscale conversion, blurring and binarization; According to the set edge detection algorithm, the particle edge of each target particle in the preprocessed image is determined; and according to the set contour extraction algorithm, the particle contour of each target particle is determined; According to the set contour calculation formula, combined with the particle contour of each target particle, the contour area of ​​each target particle is calculated; According to the particle edge and contour area corresponding to each target particle, the particle size and distribution corresponding to the filtered water are determined, and the particle size and distribution are added to the second parameter data corresponding to the second preset parameter.

[0099] It can be seen that in this optional embodiment, the efficiency and intelligence level of sediment detection are significantly improved through image preprocessing, precise edge detection and contour extraction, precise particle size calculation and comprehensive particle distribution analysis.

[0100] In another optional embodiment, the second preset parameter further includes sediment type and / or particle movement characteristics; The detection device 302 performs visual recognition on the filtered water according to the visual recognition algorithm to obtain the second parameter data corresponding to the second preset parameter, which specifically includes: When the second preset parameter includes particle motion characteristics, the motion data of the target particles in the visual recognition image are determined according to the set optical flow method or tracking algorithm, and the motion data includes at least one of the motion trajectory, motion speed, and motion acceleration of each target particle; and the motion data of all target particles are analyzed to obtain the motion characteristics of all target particles, and the motion characteristics are added to the second parameter data corresponding to the second preset parameter; When the sediment type is detected, a second identification feature corresponding to the visual identification image is extracted, and the second identification feature includes a color feature and / or a texture feature and / or a shape feature; and classification identification is performed on the second identification feature according to the set classification model to obtain a sediment type identification result corresponding to the visual image, and the sediment type identification result is added to the second parameter data corresponding to the second preset parameter.

[0101] It can be seen that in this optional embodiment, by introducing the detection of particle movement characteristics and sediment types, the comprehensiveness and practicality of sediment detection technology are significantly improved, and it is also beneficial to improve the comprehensiveness and accuracy of monitoring and analysis of sediment characteristics.

[0102] Embodiment 4 See also Figure 5 , Figure 5 Schematic diagram of the structure of a sediment detection device based on machine vision disclosed in an embodiment of the present invention. Figure 5 As shown, the machine vision-based sediment detection device may include: A memory 401 storing executable program codes; a processor 402 coupled to the memory 401; The processor 402 calls the executable program code stored in the memory 401 to execute the steps of the sediment detection method based on machine vision described in the first embodiment of the present invention or the second embodiment of the present invention.

[0103] Embodiment 5 An embodiment of the present invention discloses a computer storage medium storing computer instructions. When the computer instructions are called, they are used to execute the steps of the machine vision-based sediment detection method described in Embodiment 1 or Embodiment 2 of the present invention.

[0104] Embodiment 6 An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the machine vision-based sediment detection method described in Embodiment 1 or Embodiment 2.

[0105] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0106] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0107] Finally, it should be noted that the machine vision-based sediment detection method, equipment and device disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A sediment detection method based on machine vision, characterized in that: The method is applied to a sediment detection device based on machine vision, the device includes a pre-processing device and a detection device, and the method includes: When it is determined that the sediment detection operation conditions are currently met, the pre-processing device collects water to be detected according to a preset water extraction volume, and the water to be detected is mixed with sediment to be detected; The pre-processing device performs a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected; the pre-processing operation includes sediment sedimentation, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be detected, and the sediment concentration corresponding to the filtered water calculated based on the parameters, and the filtered water is mixed with the sediment to be detected; The detection device extracts the filtered water, and performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition includes multi-dimensional parameter recognition for the sediment to be detected, and the multi-dimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; The pre-processing result and the visual recognition result are determined as target detection results for the water to be detected.

2. The machine vision-based sediment detection method according to claim 1, characterized in that: The pre-processing device performs a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected, including: The pre-processing device performs a vibration sedimentation operation on the water to be tested, and collects vibration sedimentation data of the water to be tested in real time; The pre-processing device performs data comparison on the vibration sedimentation data according to a plurality of preset settlement indicators, and determines that the vibration sedimentation operation is completed when it is determined that the corresponding data comparison information indicates that the vibration sedimentation data meets the settlement requirement of each settlement indicator; The pre-processing device performs drainage processing on the water to be detected to obtain a drainage processing result corresponding to the water to be detected, wherein the drainage processing result includes filtered water after the preset impurities are filtered out, and the filtered water is a water-sediment mixture; The pre-processing device performs parameter calculation on the drainage processing result according to a preset density calculation formula to obtain a sediment density calculation result corresponding to the drainage processing result as a pre-processing result corresponding to the water to be detected.

3. The machine vision-based sediment detection method according to claim 2, characterized in that: The plurality of preset sedimentation indicators at least include sedimentation rate of sediment particles, water clarity, sediment layer stability, and water quality indicators; the pre-processing device performs data comparison on the vibration sedimentation sediment data according to the plurality of preset sedimentation indicators, including: The pre-processing device analyzes the vibration sedimentation data to obtain real-time index data corresponding to each of the sedimentation indicators; For each of the settlement indicators, the pre-processing device compares the standard value corresponding to the settlement indicator with the corresponding real-time indicator data to obtain a numerical comparison result corresponding to the settlement indicator; and at the same time, the numerical comparison results corresponding to all the settlement indicators are determined as data comparison information corresponding to the vibration settlement sediment data; And, the data comparison information indicates that the settlement requirements for the vibration sedimentation data to meet each of the settlement indicators are specifically: The real-time indicator data corresponding to each of the settlement indicators is within the standard value corresponding to the settlement indicator.

4. The machine vision-based sediment detection method according to any one of claims 1 to 3, characterized in that: The detection device performs visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water, including: The detection device detects a first preset parameter in the filtered water according to a preset visual recognition algorithm to obtain first parameter data corresponding to the first preset parameter; the first preset parameter includes sediment weight and sediment density; The detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to second preset parameters; the second preset parameters at least include sediment concentration; the second preset parameters also include turbidity and / or particle size and distribution; The detection device generates image recognition data corresponding to the filtered water according to the second parameter data; and determines the first parameter data, the second parameter data and the image recognition data as a visual recognition result corresponding to the filtered water.

5. The machine vision-based sediment detection method according to claim 4, characterized in that: The visual recognition algorithm includes an OpenCV algorithm; the detection device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, including: The detection device acquires a visual recognition image corresponding to the filtered water; The detection device extracts a first recognition feature of the visual recognition image according to the visual recognition algorithm, wherein the first recognition feature includes a color feature and / or a texture feature corresponding to the silt particles in the filtered water; and according to the visual recognition algorithm, performs sediment concentration prediction on the first recognition feature to obtain a predicted sediment concentration corresponding to the filtered water, and adds the predicted sediment concentration to the second parameter data corresponding to the second preset parameter; When the second preset parameter includes the turbidity, the detection device calculates the absolute value of the Laplace operator corresponding to the visual recognition image according to the visual recognition algorithm, and calculates the standard deviation of the absolute value to obtain the turbidity value corresponding to the visual recognition image, and adds the turbidity value to the second parameter data corresponding to the second preset parameter.

6. The machine vision-based sediment detection method according to claim 5, characterized in that: The detecting device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, further comprising: When the second preset parameter includes the particle size and distribution, the detection device performs preprocessing on the visual recognition image to obtain a preprocessed image corresponding to the visual recognition image; the preprocessing includes any one of grayscale conversion, blurring and binarization; The detection device determines the particle edge of each target particle in the preprocessed image according to a set edge detection algorithm; and determines the particle contour of each target particle according to a set contour extraction algorithm; The detection device calculates the contour area of ​​each target particle according to a set contour calculation formula and in combination with the particle contour of each target particle; The detection device determines the particle size and distribution corresponding to the filtered water according to the particle edge and the contour area corresponding to each target particle, and adds the particle size and distribution to the second parameter data corresponding to the second preset parameter.

7. The machine vision-based sediment detection method according to claim 5, characterized in that: The second preset parameters also include sediment type and / or particle movement characteristics; The detecting device performs visual recognition on the filtered water according to the visual recognition algorithm to obtain second parameter data corresponding to the second preset parameter, further comprising: When the second preset parameter includes the particle motion characteristics, the detection device determines the motion data of the target particles in the visual recognition image according to the set optical flow method or tracking algorithm, the motion data including at least one of the motion trajectory, motion speed, and motion acceleration of each target particle; and analyzes the motion data of all the target particles to obtain the motion characteristics of all the target particles, and adds the motion characteristics to the second parameter data corresponding to the second preset parameter; When the sediment type is determined, the detection device extracts a second identification feature corresponding to the visual identification image, wherein the second identification feature includes the color feature and / or the texture feature and / or the shape feature; and performs classification identification on the second identification feature according to the set classification model to obtain a sediment type identification result corresponding to the visual image, and adds the sediment type identification result to the second parameter data corresponding to the second preset parameter.

8. A sediment detection device based on machine vision, characterized in that: The device comprises a pre-processing device and a detection device, wherein: The pre-processing device is used to collect water to be tested according to a preset water extraction volume when it is determined that the sediment detection operation conditions are currently met, and the water to be tested is mixed with sediment to be tested; The pre-processing device is further used to perform a pre-processing operation on the water to be detected to obtain a pre-processing result corresponding to the water to be detected; the pre-processing operation includes sediment sedimentation, drainage and parameter calculation; the pre-processing result includes filtered water corresponding to the water to be detected, and the sediment concentration corresponding to the filtered water calculated based on the parameters, and the filtered water is mixed with the sediment to be detected; The detection device is used to extract the filtered water and perform visual recognition on the filtered water according to a preset visual recognition algorithm to obtain a visual recognition result corresponding to the filtered water; the visual recognition includes multi-dimensional parameter recognition for the sediment to be detected, and the multi-dimensional parameter recognition includes at least visual recognition of sediment density and sediment concentration; The detection device is further used to determine the pre-processing result and the visual recognition result as the target detection result for the water to be detected.

9. A sediment detection device based on machine vision, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the machine vision-based sediment detection method according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the machine vision-based sediment detection method according to any one of claims 1 to 7.

Images