Underwater pipeline inspection robot fish control method and system based on machine vision

Abstract

The invention discloses a control method and system for an underwater pipeline inspection robot fish based on machine vision, and belongs to the technical field of underwater robot control, and the method comprises the steps: obtaining an inspection image of the underwater pipeline inspection robot fish, and generating a shooting offset index; judging whether the shooting quality of the inspection image is normal or not according to the shooting offset index; obtaining a camera rotation angle of the underwater pipeline inspection robot fish, and generating a camera rotation angle evaluation index; judging whether the rotation angle of the camera has a blocking risk or not according to the camera rotation angle evaluation index; according to the shooting offset index and the camera rotation angle evaluation index, a propeller adjustment analysis model is established, and a propeller propulsion power adjustment value is generated; according to the propulsive power adjustment value of the propeller, the propulsive power of the underwater pipeline inspection robotic fish propeller is adjusted; according to the invention, the power adjustment value of the propeller is dynamically adjusted to optimize the attitude adjustment of the robotic fish, so that the stability of image acquisition and the image quality are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater robot control, and in particular relates to a control method and system for an underwater pipeline inspection robot fish based on machine vision. Background Art

[0002] With the development of industry and infrastructure, pipeline inspection and maintenance have become increasingly important, especially in underwater environments. Traditional manual inspection methods are inefficient and pose significant safety risks. To improve the efficiency and safety of underwater pipeline inspections, underwater robotic fish, as autonomous inspection equipment, are becoming a key tool in the pipeline inspection field. Equipped with sensors such as cameras, robotic fish can capture images and monitor pipelines in real time.

[0003] When inspecting pipelines, if the camera shooting angle of the robotic fish is abnormal, it will often lead to a decline in image acquisition quality, which in turn affects subsequent inspection and analysis work. In particular, when the posture of the robotic fish is abnormal, it may cause the camera shooting angle of the robotic fish to be abnormal, resulting in unstable image acquisition when the robotic fish collects images of the inspection target during the inspection process, which reduces the accuracy of the robotic fish inspection. Summary of the Invention

[0004] In response to the deficiencies in the prior art, the present invention provides a control method and system for an underwater pipeline inspection robot fish based on machine vision, which solves the above-mentioned problems.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A control method for an underwater pipeline inspection robot fish based on machine vision comprises the following steps:

[0006] Acquire inspection images of underwater pipeline inspection robot fish and generate shooting offset index;

[0007] According to the shooting offset index, judge whether the shooting quality of the inspection image is normal;

[0008] If the inspection image is offset, obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index; the camera rotation angle includes the camera horizontal rotation angle and the camera vertical rotation angle;

[0009] Based on the camera rotation angle evaluation index, determine whether there is a risk of obstruction in the camera's rotation angle;

[0010] If the camera's rotation angle is blocked, a thruster adjustment analysis model is established based on the shooting offset index and the camera rotation angle evaluation index to generate the thruster propulsion power adjustment value;

[0011] According to the propulsion power adjustment value of the propeller, the propulsion power of the underwater pipeline inspection robot fish is adjusted.

[0012] On the basis of the above technical solutions, the present invention also provides the following optional technical solutions:

[0013] Further technical solution: The method for generating the shooting offset index specifically includes the following steps:

[0014] Obtain inspection images of the underwater pipeline inspection robot fish and mark them as inspection images;

[0015] Obtaining the features of the inspection target in the inspection image and the positions of the features of the inspection target in the inspection image, and marking them as inspection target features and feature coordinates respectively;

[0016] According to the characteristic coordinates, an offset analysis model is established to generate the shooting offset index.

[0017] Further technical solution: The expression of the offset analysis model is specifically:

[0018]

[0019] In the expression, Q deviation It represents the shooting offset index, Δx i It represents the difference between the X-axis coordinate value of the feature coordinate and the X-axis coordinate value of the center point of the inspection image, Δy i It represents the difference between the Y-axis coordinate value of the feature coordinate and the Y-axis coordinate value of the inspection image center point coordinate. max It represents the maximum offset value of the feature coordinates.

[0020] Further technical solution: The method for generating the camera rotation angle evaluation index specifically includes:

[0021] By formula:

[0022] K= i i

[0023] maxθ;

[0024] Generate camera rotation angle evaluation index K;

[0025] In the formula, maxθ represents the maximum rotation angle of the camera, θ i Indicates the current rotation angle.

[0026] Further technical solution: The method for generating the thruster propulsion power adjustment value specifically includes the following steps:

[0027] If the camera's rotation angle is blocked, the posture adjustment value of the underwater pipeline inspection robot fish is generated according to the shooting offset index and the camera rotation angle evaluation index;

[0028] Acquire underwater environmental data and generate an underwater environmental impact coefficient; the underwater environmental data includes water flow impact pressure and water flow impact angle; water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish; water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish;

[0029] A thruster adjustment analysis model is established, and the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient are substituted into the thruster adjustment analysis model to generate the thruster propulsion power adjustment value.

[0030] Further technical solution: The method for generating the posture adjustment value of the underwater pipeline inspection robot fish is specifically as follows:

[0031] generating a shooting offset index adjustment value according to the shooting offset index and the shooting offset index threshold; wherein the shooting offset index adjustment value refers to the difference between the shooting offset index and the shooting offset index threshold, and the shooting offset index is greater than the shooting offset index threshold;

[0032] By formula:

[0033] Δθ=ΔQ*L max *ω;

[0034] Generate the camera shooting angle adjustment requirement value Δθ;

[0035] In the formula, ΔQ represents the shooting offset index adjustment value, L max It represents the maximum offset value of the feature coordinates, and ω represents the influence coefficient of the camera rotation angle on the change of the feature coordinates in the inspection image;

[0036] By formula:

[0037] θ lack =Δθ-(maxθ-θ i );

[0038] Generate the posture adjustment value θ of the underwater pipeline inspection robot fish lack ;

[0039] In the formula, Δθ represents the required value of the camera shooting angle adjustment, maxθ represents the maximum rotation angle of the camera, and θ i Indicates the current rotation angle.

[0040] Further technical solution: The underwater environmental impact coefficient is generated in the following manner:

[0041] Obtaining the water flow impact pressure of the underwater environment and generating a water flow velocity influence coefficient; wherein the water flow velocity influence coefficient refers to the ratio of the difference between the water flow impact pressure and the water flow impact pressure threshold to the water flow impact pressure threshold;

[0042] By formula:

[0043]

[0044] Generate underwater environmental impact coefficient σ;

[0045] In the formula, F asse It represents the water velocity influence coefficient, impactθ refers to the water impact angle, impactθ max Refers to the maximum impact angle of water flow.

[0046] Further technical solution: The expression of the propeller adjustment analysis model is specifically:

[0047]

[0048] In the expression, ΔJ represents the propulsion power adjustment value of the thruster, J η It represents the propulsion power increase value of the propeller with a known angle adjustment efficiency, σ represents the underwater environment impact coefficient, and θ lack It represents the attitude adjustment value of the underwater pipeline inspection robot fish, θ η It represents the angle adjustment value of the thruster with known angle adjustment efficiency.

[0049] The control system of the underwater pipeline inspection robot fish based on machine vision includes:

[0050] An image analysis unit, used to obtain inspection images of the underwater pipeline inspection robot fish and generate a shooting offset index;

[0051] An image judgment module is used to judge whether the shooting quality of the inspection image is normal based on the shooting offset index;

[0052] A rotation angle analysis module is used to obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index if the inspection image is offset; wherein the camera rotation angle includes the camera horizontal rotation angle and the camera vertical rotation angle;

[0053] A rotation angle judgment module is used to judge whether there is a risk of obstruction in the camera's rotation angle based on the camera's rotation angle evaluation index;

[0054] an adjustment analysis unit, wherein if the camera rotation angle is obstructed, the adjustment analysis unit is used to establish a propeller adjustment analysis model based on the shooting offset index and the camera rotation angle evaluation index to generate a propeller propulsion power adjustment value;

[0055] The adjustment control module is used to adjust the propulsion power of the underwater pipeline inspection robot fish according to the propulsion power adjustment value of the propeller.

[0056] Further technical solution: The image analysis unit specifically includes:

[0057] An image acquisition module is used to acquire inspection images of the underwater pipeline inspection robot fish and mark them as inspection images;

[0058] An image recognition module is used to obtain the features of the inspection target in the inspection image and the location of the features of the inspection target in the inspection image, and mark them as inspection target features and feature coordinates respectively;

[0059] A shooting offset index generation module is used to establish an offset analysis model based on feature coordinates and generate a shooting offset index;

[0060] The adjustment analysis unit specifically includes:

[0061] A posture adjustment analysis module is used to generate a posture adjustment value for the underwater pipeline inspection robot fish based on the shooting offset index and the camera rotation angle evaluation index;

[0062] The water flow analysis module is used to obtain underwater environmental data and generate underwater environmental impact coefficients. The underwater environmental data includes water flow impact pressure and water flow impact angle. Water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish. Water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish.

[0063] The adjustment value output module is used to establish a thruster adjustment analysis model, substitute the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient into the thruster adjustment analysis model, and generate the thruster propulsion power adjustment value.

[0064] The present invention provides a control method and system for an underwater pipeline inspection robot fish based on machine vision, which has the following advantages over the prior art:

[0065] The present invention determines whether the camera shooting angle is abnormal based on the position of the pipeline in the image. If the camera angle deviates, the robot fish's posture is automatically adjusted to realign it with the pipeline. When the camera has reached the maximum rotation angle, the system calculates precise mechanical and propeller power adjustment values ​​based on the offset position of the pipeline, the remaining rotation angle of the camera, and the impact of the water flow on the robot fish's posture, thereby optimizing the robot fish's posture adjustment and improving the stability and image quality of image acquisition. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flowchart of a control method for an underwater pipeline inspection robot fish based on machine vision provided by an embodiment of the present invention.

[0067] Figure 2 This is a flowchart of step S10 provided in an embodiment of the present invention.

[0068] Figure 3 This is a flowchart of step S50 provided in an embodiment of the present invention.

[0069] Figure 4 A schematic structural diagram of a control system for a machine vision-based underwater pipeline inspection robot fish provided in an embodiment of the present invention.

[0070] Figure 5 This is a module block diagram of an image analysis unit provided in an embodiment of the present invention.

[0071] Figure 6 This is a module block diagram of the adjustment analysis unit provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0073] The specific implementation of the present invention is described in detail below with reference to specific embodiments.

[0074] See also Figure 1 , a method for controlling an underwater pipeline inspection robot fish based on machine vision provided in one embodiment of the present invention, comprising the following steps:

[0075] Step S10: Acquire an inspection image of the underwater pipeline inspection robot fish and generate a shooting offset index;

[0076] Step S20: judging whether the shooting quality of the inspection image is normal according to the shooting offset index;

[0077] Step S30: If the inspection image is offset, obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index; wherein the camera rotation angle includes the camera horizontal rotation angle and the camera vertical rotation angle;

[0078] In this embodiment, the horizontal rotation angle of the camera refers to the rotation angle of the current camera from the midpoint to the sides in the horizontal direction, and the vertical rotation angle of the camera refers to the rotation angle of the current camera from the midpoint to the sides in the vertical direction;

[0079] Step S40: judging whether there is a risk of obstruction in the camera rotation angle according to the camera rotation angle evaluation index;

[0080] Step S50: If the camera rotation angle is blocked, a propeller adjustment analysis model is established based on the shooting offset index and the camera rotation angle evaluation index to generate a propeller propulsion power adjustment value;

[0081] It should be explained that the propeller in the propulsion power adjustment value is the propeller used by the underwater pipeline inspection robot fish for posture adjustment, which is different from the power propeller of the underwater pipeline inspection robot fish. The propeller in the present invention refers to the propeller used for posture adjustment of the underwater pipeline inspection robot fish;

[0082] Step S60: adjusting the propulsion power of the underwater pipeline inspection robot fish according to the propulsion power adjustment value of the propeller.

[0083] See also Figure 2 As a preferred embodiment of the present invention, the method for generating the shooting offset index specifically includes the following steps:

[0084] S11: Obtain an inspection image of the underwater pipeline inspection robot fish and mark it as an inspection image;

[0085] S12: Acquire features of the inspection target in the inspection image and positions of the features of the inspection target in the inspection image, and mark them as inspection target features and feature coordinates, respectively;

[0086] It should be explained that the feature coordinates refer to the position of the inspection target feature center point in the inspection image;

[0087] Specifically, a coordinate system is established based on the inspection image, and the position of the inspection target feature center point in the coordinate system is the feature coordinate;

[0088] In addition, the inspection path of the underwater pipeline inspection robot fish is a set mode, and its inspection path is set by relevant personnel in this field. For example, the target tracking mode can be set according to the underwater distribution status of the inspection target, so that the underwater pipeline inspection robot fish is always in a certain direction and height of the inspection target, thereby tracking the inspection target.

[0089] S13: establishing an offset analysis model based on the feature coordinates and generating a shooting offset index;

[0090] It should be noted that when the robotic fish is conducting inspections, due to factors such as water flow impact, the robotic fish may tilt during operation (this phenomenon refers to the fact that the robotic fish is still on the scheduled inspection route, but its posture is abnormal, for example, the robotic fish is in a flipped state, that is, the part of the robotic fish that was originally at the upper end is now at the lower end). As a result, the inspection images taken by the robotic fish's shooting equipment are different from the standard inspection images, and the greater the difference, the more obvious the tilt of the robotic fish.

[0091] As a preferred embodiment of the present invention, the expression of the offset analysis model is specifically:

[0092]

[0093] In the expression, Q deviation It represents the shooting offset index, Δx i It represents the difference between the X-axis coordinate value of the feature coordinate and the X-axis coordinate value of the center point of the inspection image, Δy i It represents the difference between the Y-axis coordinate value of the feature coordinate and the Y-axis coordinate value of the inspection image center point coordinate. max It represents the maximum offset value of the feature coordinates;

[0094] It should be noted that the distance between the feature coordinates and the image center coordinates is calculated by the Pythagorean theorem, that is, the distance between the feature coordinates and the image center coordinates in the expression of the offset analysis model. (The calculation principle of this calculation formula is the Pythagorean theorem); the distance value between the feature coordinates and the image center coordinates refers to the distance that the feature coordinates deviate from the image center coordinates;

[0095] Then, according to the distance between the coordinates of the center point of the image and the coordinates of the cross edge in the inspection image, that is, L in the expression of the offset analysis model max , the shooting offset index of the feature coordinates can be obtained; the shooting offset index refers to the degree of deviation between the position of the inspection target feature in the inspection image and the feature coordinates in the standard inspection image;

[0096] In this embodiment, the maximum offset value of the feature coordinates refers to the maximum distance between the coordinates of the center point of the image and the edge coordinates in the inspection image; for example, when the robotic fish is inspecting a pipeline, if the robotic fish is tilted, the feature coordinates in the inspection image will deviate from the feature coordinates in the standard inspection image; since the inspection image has a fixed size, when the inspection target feature cannot be identified in the inspection image (that is, the inspection target feature is not in the inspection image, which is caused by the tilt of the robotic fish and the inspection target feature is not within the shooting range of the camera), if the inspection target feature can be identified, the maximum distance that the feature coordinates can be offset is the maximum offset value of the feature coordinates;

[0097] As a preferred embodiment of the present invention, the method of judging whether the shooting quality of the inspection image is normal is specifically as follows:

[0098] comparing the shot shift index to a shot shift index threshold;

[0099] In this embodiment, the shooting offset index threshold is a set value, which is set by relevant personnel in this field, and the setting range of the shooting offset index threshold is 0-1;

[0100] When the shooting offset index is less than or equal to the shooting offset index threshold, the shooting quality of the inspection image is determined to be normal. At this time, the smaller the shooting offset index is, the smaller the offset of the feature coordinates in the inspection image is, that is, the more normal the shooting quality of the inspection image is.

[0101] When the shooting offset index is greater than the shooting offset index threshold, it is determined that the shooting quality of the inspection image is in an abnormal state; at this time, the larger the shooting offset index, the greater the offset of the feature coordinates in the inspection image, that is, the more abnormal the shooting quality of the inspection image is.

[0102] As a preferred embodiment of the present invention, the method for generating the camera rotation angle evaluation index specifically includes:

[0103] By formula:

[0104] K= i i

[0105] maxθ;

[0106] Generate camera rotation angle evaluation index K;

[0107] In the formula, maxθ represents the maximum rotation angle of the camera, θ i Indicates the current rotation angle;

[0108] It should be noted that the maximum rotation angle of the camera, maxθ, refers to the average of the maximum rotation amplitude of the camera to both sides. The current rotation angle θ i Refers to the angle of the current camera rotating from the center to a certain side; for example, if the camera's rotation angle range is 180°, then the maximum rotation range of the camera from the center to both sides is 90°, that is, the maximum rotation angle of the camera maxθ is 90°;

[0109] In this embodiment, based on the current camera rotation angle and the camera's maximum rotation angle, an analysis is performed to determine whether the current camera rotation angle has reached the camera's maximum rotation angle, thereby determining whether the camera can be rotated so that the inspection target is within the inspection image. For example, if the current camera rotation angle is 82° and the camera's maximum rotation angle is 90°, this indicates that the camera's rotation angle is about to reach its maximum rotation angle. When the maximum rotation angle is reached, the camera cannot rotate further, and the posture of the camera base (i.e., the robotic fish) must be adjusted so that the inspection image captured by the camera includes the features of the inspection target.

[0110] As a preferred embodiment of the present invention, the method of determining whether there is a risk of obstruction in the rotation angle of the camera is specifically as follows:

[0111] comparing the camera rotation angle evaluation index with a camera rotation angle evaluation index threshold;

[0112] In this embodiment, the camera rotation angle evaluation index threshold is a set value, which is set by relevant personnel in this field, and the setting range of the camera rotation angle evaluation index threshold is 0-1;

[0113] If the camera rotation angle evaluation index is less than or equal to the camera rotation angle evaluation index threshold, it is determined that the camera rotation angle does not have the risk of being obstructed. In this case, the smaller the camera rotation angle evaluation index, the larger the range that the current camera can adjust, and the smaller the risk of the camera rotation angle being obstructed. When it is determined that the camera rotation angle does not have the risk of being obstructed, the camera shooting angle can be adjusted to improve the shooting quality of the inspection image.

[0114] If the camera rotation angle evaluation index is greater than the camera rotation angle evaluation index threshold, it is determined that the camera rotation angle is at risk of being obstructed; at this time, the larger the camera rotation angle evaluation index is, the smaller the range that the current camera can adjust is, and the greater the risk of the camera rotation angle being obstructed is.

[0115] See also Figure 3 As a preferred embodiment of the present invention, the method for generating the propeller propulsion power adjustment value specifically includes the following steps:

[0116] S51: If the rotation angle of the camera is blocked, generating a posture adjustment value of the underwater pipeline inspection robot fish according to the shooting offset index and the camera rotation angle evaluation index;

[0117] S52: Acquire underwater environmental data and generate an underwater environmental impact coefficient; wherein the underwater environmental data includes water flow impact pressure and water flow impact angle; water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish; water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish;

[0118] S53: Establish a thruster adjustment analysis model, substitute the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient into the thruster adjustment analysis model, and generate a thruster propulsion power adjustment value.

[0119] As a preferred embodiment of the present invention, the posture adjustment value of the underwater pipeline inspection robot fish is generated in the following manner:

[0120] generating a shooting offset index adjustment value according to the shooting offset index and the shooting offset index threshold; wherein the shooting offset index adjustment value refers to the difference between the shooting offset index and the shooting offset index threshold, and the shooting offset index is greater than the shooting offset index threshold;

[0121] It should be explained that when the shooting offset index is less than the shooting offset index threshold, the shooting offset index adjustment value is set to 0;

[0122] Generate the camera shooting angle adjustment requirement value according to the shooting offset index adjustment value;

[0123] The camera shooting angle adjustment requirement value is generated in the following manner:

[0124] By formula:

[0125] Δθ=ΔQ*L max *ω;

[0126] Generate the camera shooting angle adjustment requirement value Δθ;

[0127] In the formula, ΔQ represents the shooting offset index adjustment value, L max It represents the maximum offset value of the feature coordinates, and ω represents the influence coefficient of the camera rotation angle on the change of the feature coordinates in the inspection image;

[0128] It should be explained that the shooting imaging adjustment coefficient ω refers to the coefficient of influence of the camera rotation angle on the change of feature coordinates in the inspection image. For example, when the shooting height is fixed at h, the camera rotation angle is n, and the coordinate change value of the feature coordinate in the inspection image is m, then the value of the shooting imaging adjustment coefficient ω is m / n.

[0129] When capturing inspection images, the inspection target is fixed. Whenever the camera rotates, the feature coordinates in the inspection image change with the camera's rotation, and this position change is the same as the pinhole imaging principle. For example, in a pinhole imaging experiment, whenever the light source moves, the position of its light source image also changes, and this changes in a fixed pattern.

[0130] Adjust the required value according to the camera shooting angle to generate the posture adjustment value of the underwater pipeline inspection robot fish;

[0131] The specific method for generating the posture adjustment value of the underwater pipeline inspection robot fish is as follows:

[0132] By formula:

[0133] θ lack =Δθ-(maxθ-θ i );

[0134] Generate the posture adjustment value θ of the underwater pipeline inspection robot fish lack ;

[0135] In the formula, Δθ represents the required value of the camera shooting angle adjustment, maxθ represents the maximum rotation angle of the camera, and θ i Indicates the current rotation angle;

[0136] In this embodiment, (maxθ-θ i ) refers to the remaining rotation angle of the camera; if it is necessary to keep the feature coordinates in the inspection image in a normal state, the camera's shooting angle needs to be adjusted. When the camera's shooting angle is adjusted to the maximum rotation angle, it is still impossible to keep the feature coordinates in the inspection image in a normal state; in this case, the posture of the mechanical fish is adjusted instead of the camera to rotate, so as to keep the feature coordinates in the inspection image in a normal state; this is because the camera is fixed to the mechanical fish. Adjusting the posture of the mechanical fish is also adjusting the posture of the camera, which is equivalent to adjusting the camera's shooting angle (the shooting angle here is relative to the inspection target);

[0137] In addition, in some practical application scenarios, when the shooting offset index adjustment value is set to 0, the camera shooting angle adjustment demand value Δθ is also 0. When the camera shooting angle adjustment demand value Δθ is 0, the posture adjustment value θ of the underwater pipeline inspection robot fish islack The value of is also set to 0.

[0138] As a preferred embodiment of the present invention, the underwater environmental impact coefficient is generated in the following manner:

[0139] Obtaining the water flow impact pressure of the underwater environment and generating a water flow velocity influence coefficient; wherein the water flow velocity influence coefficient refers to the ratio of the difference between the water flow impact pressure and the water flow impact pressure threshold to the water flow impact pressure threshold;

[0140] It should be noted that the water flow impact pressure threshold refers to the minimum water flow impact pressure that can change the posture of the underwater pipeline inspection robot fish; in addition, the pressure value corresponding to the water flow impact pressure threshold is the pressure value of the water flow vertically impacting the surface of the underwater pipeline inspection robot fish;

[0141] By formula:

[0142]

[0143] Generate underwater environmental impact coefficient σ;

[0144] In the formula, F asse It represents the water velocity influence coefficient, impactθ refers to the water impact angle, impactθ max Refers to the maximum impact angle of water flow;

[0145] In this embodiment, the maximum impact angle of the water flow is impactθ max Generally, it is 90°. There are two angles between the direction of water flow and the surface of the underwater pipeline inspection robot fish. The sum of the two angles is 180°. The smaller angle is selected as the water flow impact angle. The maximum water flow impact angle is max is 90°;

[0146] It should be noted that It represents the correction coefficient of the water flow impact angle on the water flow velocity coefficient. For example, when the water flow impact angle impactθ is 5°, at the same water flow velocity, the impact of this impact angle on the mechanical fish posture is less than the impact of the water flow impact angle impactθ on the mechanical fish posture. Therefore, the role of the water flow impact angle influence coefficient is to correct the water flow velocity influence coefficient F. asse Correction is made. Under the condition of the same water flow rate, the impact caused by different impact angles is different.

[0147] As a preferred embodiment of the present invention, the expression of the thruster adjustment analysis model is specifically:

[0148]

[0149] In the expression, ΔJ represents the propulsion power adjustment value of the thruster, J η It represents the propulsion power increase value of the propeller with a known angle adjustment efficiency, σ represents the underwater environment impact coefficient, and θ lack It represents the attitude adjustment value of the underwater pipeline inspection robot fish, θ η It represents the angle adjustment value of the thruster's known angle adjustment efficiency;

[0150] In this embodiment, the propeller known angle adjustment efficiency refers to the propulsion power increase required to adjust a certain angle within a fixed time. For example, when the propeller increases power by x, it can adjust angle r within time t. Then power x is the propulsion power increase J in the propeller known angle adjustment efficiency. η , angle r is the angle adjustment value θ in the known angle adjustment efficiency of the propeller η .

[0151] See also Figure 4 The present invention also provides a control system for an underwater pipeline inspection robot fish based on machine vision, the system comprising:

[0152] An image analysis unit 10 is used to obtain an inspection image of the underwater pipeline inspection robot fish and generate a shooting offset index;

[0153] The image judgment module 20 is used to judge whether the shooting quality of the inspection image is normal according to the shooting offset index;

[0154] A rotation angle analysis module 30 is used to obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index if the inspection image is offset; wherein the camera rotation angle includes the camera lateral rotation angle and the camera longitudinal rotation angle;

[0155] A rotation angle determination module 40 is configured to determine whether the camera rotation angle is at risk of being obstructed based on the camera rotation angle evaluation index;

[0156] An adjustment analysis unit 50 is configured to establish a propeller adjustment analysis model based on the shooting offset index and the camera rotation angle evaluation index to generate a propeller propulsion power adjustment value if the camera rotation angle is blocked;

[0157] The adjustment control module 60 is used to adjust the propulsion power of the underwater pipeline inspection robot fish according to the propulsion power adjustment value of the propeller.

[0158] See also Figure 5 As a preferred embodiment of the present invention, the image analysis unit specifically includes:

[0159] An image acquisition module 11 is used to acquire an inspection image of the underwater pipeline inspection robot fish and mark it as an inspection image;

[0160] An image recognition module 12 is used to obtain the features of the inspection target in the inspection image and the positions of the features of the inspection target in the inspection image, and mark them as inspection target features and feature coordinates respectively;

[0161] The shooting offset index generating module 13 is used to establish an offset analysis model according to the feature coordinates and generate a shooting offset index.

[0162] See also Figure 6 As a preferred embodiment of the present invention, the adjustment analysis unit specifically includes:

[0163] A posture adjustment analysis module 51 is used to generate a posture adjustment value of the underwater pipeline inspection robot fish based on the shooting offset index and the camera rotation angle evaluation index;

[0164] The water flow analysis module 52 is used to obtain underwater environmental data and generate an underwater environmental impact coefficient. The underwater environmental data includes water flow impact pressure and water flow impact angle. Water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish. Water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish.

[0165] The adjustment value output module 53 is used to establish a thruster adjustment analysis model, substitute the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient into the thruster adjustment analysis model, and generate a thruster propulsion power adjustment value.

[0166] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A control method for underwater pipeline inspection robot fish based on machine vision, characterized in that: Also includes: Acquire inspection images of underwater pipeline inspection robot fish and generate shooting offset index; According to the shooting offset index, judge whether the shooting quality of the inspection image is normal; If the inspection image is offset, obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index; the camera rotation angle includes the camera horizontal rotation angle and the camera vertical rotation angle; Based on the camera rotation angle evaluation index, determine whether there is a risk of obstruction in the camera's rotation angle; If the camera's rotation angle is blocked, a thruster adjustment analysis model is established based on the shooting offset index and the camera rotation angle evaluation index to generate the thruster propulsion power adjustment value; According to the propulsion power adjustment value of the propeller, the propulsion power of the underwater pipeline inspection robot fish is adjusted.

2. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 1 is characterized in that: The method for generating the shooting offset index specifically includes the following steps: Obtain inspection images of the underwater pipeline inspection robot fish and mark them as inspection images; Obtaining the features of the inspection target in the inspection image and the positions of the features of the inspection target in the inspection image, and marking them as inspection target features and feature coordinates respectively; According to the characteristic coordinates, an offset analysis model is established to generate the shooting offset index.

3. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 2 is characterized in that: The expression of the offset analysis model is specifically: In the expression, Q deviation It represents the shooting offset index, Δx i It represents the difference between the X-axis coordinate value of the feature coordinate and the X-axis coordinate value of the center point of the inspection image, Δy i It represents the difference between the Y-axis coordinate value of the feature coordinate and the Y-axis coordinate value of the inspection image center point coordinate. max It represents the maximum offset value of the feature coordinates.

4. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 1 is characterized in that: The camera rotation angle evaluation index is generated in the following manner: By formula: Generate camera rotation angle evaluation index K; In the formula, maxθ represents the maximum rotation angle of the camera, θ i Indicates the current rotation angle.

5. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 1 is characterized in that: The method for generating the propeller propulsion power adjustment value specifically includes the following steps: If the camera's rotation angle is blocked, the posture adjustment value of the underwater pipeline inspection robot fish is generated according to the shooting offset index and the camera rotation angle evaluation index; Acquire underwater environmental data and generate an underwater environmental impact coefficient; the underwater environmental data includes water flow impact pressure and water flow impact angle; water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish; water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish; A thruster adjustment analysis model is established, and the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient are substituted into the thruster adjustment analysis model to generate the thruster propulsion power adjustment value.

6. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 5 is characterized in that: The method for generating the posture adjustment value of the underwater pipeline inspection robot fish is specifically as follows: generating a shooting offset index adjustment value according to the shooting offset index and the shooting offset index threshold; wherein the shooting offset index adjustment value refers to the difference between the shooting offset index and the shooting offset index threshold, and the shooting offset index is greater than the shooting offset index threshold; By formula: Δθ=ΔQ*L max *oh; Generate the camera shooting angle adjustment requirement value Δθ; In the formula, ΔQ represents the shooting offset index adjustment value L max It represents the maximum offset value of the feature coordinates, and ω represents the influence coefficient of the camera rotation angle on the change of the feature coordinates in the inspection image; By formula: i lack =Δθ-(maxθ-θ i ); Generate the posture adjustment value θ of the underwater pipeline inspection robot fish lack ; In the formula, Δθ represents the required value of the camera shooting angle adjustment, maxθ represents the maximum rotation angle of the camera, and θ i Indicates the current rotation angle.

7. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 5 is characterized in that: The underwater environmental impact coefficient is generated in the following manner: Obtaining the water flow impact pressure of the underwater environment and generating a water flow velocity influence coefficient; wherein the water flow velocity influence coefficient refers to the ratio of the difference between the water flow impact pressure and the water flow impact pressure threshold to the water flow impact pressure threshold; By formula: Generate underwater environmental impact coefficient σ; In the formula, F asse It represents the water velocity influence coefficient, impactθ refers to the water impact angle, impactθ max Refers to the maximum impact angle of water flow.

8. The control method of the underwater pipeline inspection robot fish based on machine vision according to claim 1 is characterized in that: The expression of the thruster adjustment analysis model is specifically: In the expression, ΔJ represents the propulsion power adjustment value of the thruster, J η It represents the propulsion power increase value of the propeller with a known angle adjustment efficiency, σ represents the underwater environment impact coefficient, and θ lack It represents the attitude adjustment value of the underwater pipeline inspection robot fish, θ η It represents the angle adjustment value of the thruster with known angle adjustment efficiency.

9. The control system of underwater pipeline inspection robot fish based on machine vision is characterized by: The system is used to execute the control method of the underwater pipeline inspection robot fish based on machine vision according to any one of claims 1 to 8; the system comprises: An image analysis unit, used to obtain inspection images of the underwater pipeline inspection robot fish and generate a shooting offset index; An image judgment module is used to judge whether the shooting quality of the inspection image is normal based on the shooting offset index; A rotation angle analysis module is used to obtain the camera rotation angle of the underwater pipeline inspection robot fish and generate a camera rotation angle evaluation index if the inspection image is offset; wherein the camera rotation angle includes the camera horizontal rotation angle and the camera vertical rotation angle; A rotation angle judgment module is used to judge whether there is a risk of obstruction in the camera's rotation angle based on the camera's rotation angle evaluation index; an adjustment analysis unit, wherein if the camera rotation angle is obstructed, the adjustment analysis unit is used to establish a propeller adjustment analysis model based on the shooting offset index and the camera rotation angle evaluation index to generate a propeller propulsion power adjustment value; The adjustment control module is used to adjust the propulsion power of the underwater pipeline inspection robot fish according to the propulsion power adjustment value of the propeller.

10. The control system of the underwater pipeline inspection robot fish based on machine vision according to claim 9 is characterized in that: The image analysis unit specifically includes: An image acquisition module is used to acquire inspection images of the underwater pipeline inspection robot fish and mark them as inspection images; An image recognition module is used to obtain the features of the inspection target in the inspection image and the location of the features of the inspection target in the inspection image, and mark them as inspection target features and feature coordinates respectively; A shooting offset index generation module is used to establish an offset analysis model based on feature coordinates and generate a shooting offset index; The adjustment analysis unit specifically includes: A posture adjustment analysis module is used to generate a posture adjustment value for the underwater pipeline inspection robot fish based on the shooting offset index and the camera rotation angle evaluation index; The water flow analysis module is used to obtain underwater environmental data and generate underwater environmental impact coefficients. The underwater environmental data includes water flow impact pressure and water flow impact angle. Water flow impact pressure refers to the impact force of water flow on the underwater pipeline inspection robot fish. Water flow impact angle refers to the angle formed between the water flow direction and the underwater pipeline inspection robot fish. The adjustment value output module is used to establish a thruster adjustment analysis model, substitute the posture adjustment value of the underwater pipeline inspection robot fish and the underwater environment impact coefficient into the thruster adjustment analysis model, and generate the thruster propulsion power adjustment value.

Images