A method and system for detecting deformation of a ship's cargo hold inner shell

Abstract

This application provides a ship cargo hold hull deformation detection system and method. The system includes: multiple detection points located on the surface of the cargo hold hull; at least two camera groups, with fixed relative poses between cameras within the same group; and a data processing unit connected to the camera groups. The fields of view of at least two camera groups overlap, and they can photograph marker points set on each other's groups. Each detection point is captured by at least two cameras from different locations. The data processing unit is configured to: synchronously acquire images from each camera; and, based on the sub-pixel coordinates of the detection points in the images, combined with parameters characterizing the relative pose relationships between the camera groups, calculate the three-dimensional displacement of all detection points in the ship's coordinate system through adjustment. This technical solution enables real-time monitoring of the deformation of the hull structure of large ships such as liquefied gas carriers during dynamic processes.

Description

Technical Field

[0001] This application belongs to the field of marine engineering technology, and in particular relates to a method and system for detecting deformation of the inner shell of a ship's cargo hold. Background Technology

[0002] The construction of liquefied natural gas (LNG) carriers demands high precision and presents significant technical challenges. Considering shipyard conditions and timelines, the construction process requires various relocations and site changes, potentially including transfers from shore to floating docks. During this process, cargo compartments may undergo multiple construction or testing phases, such as shore relocation, transshipment, harbor basin suspension and buoyancy testing, and dockside strength testing. Therefore, advanced precision measurement technologies are needed for high-precision real-time deformation detection to obtain key parameters of the hull deformation during each process or phase, providing precise technical guidance for hull structural design. Summary of the Invention

[0003] In view of the shortcomings of the prior art described above, this application provides a method and system for detecting deformation of the inner shell of a ship's cargo hold, aiming to achieve real-time, high-precision, full-field three-dimensional deformation monitoring of the inner shell of the ship's cargo hold during dynamic processes.

[0004] To achieve the above objectives, in a first aspect, this application provides a ship cargo hold inner hull deformation detection system, comprising:

[0005] Multiple inspection points are located on the inner surface of the cargo hold shell;

[0006] At least two camera groups are fixedly mounted on the inner shell surface of the cargo hold; each camera group includes a rigidly connected mounting base and at least two cameras fixed thereon, with the relative poses of the cameras within the same group being fixed;

[0007] The data processing unit is communicatively connected to all of the cameras;

[0008] The fields of view of the at least two camera groups overlap, and they can photograph each other's marker points.

[0009] Each detection point was captured by at least two cameras from different locations;

[0010] The data processing unit is configured to: synchronously acquire images from each camera; based on the sub-pixel coordinates of the detection points in the images, combined with parameters characterizing the relative pose relationship between the camera groups, uniformly calculate the three-dimensional displacement of all detection points in the ship's coordinate system through adjustment calculation.

[0011] Secondly, this application provides a method for detecting deformation of the inner shell of a ship's cargo hold, using the system described above, and including the following steps:

[0012] S1: The detection point and at least two camera groups are deployed in the inner shell of the cargo hold to ensure that the fields of view of each camera group overlap and that they can photograph the marker points on each other's groups, and to ensure that each detection point is observed by at least two cameras.

[0013] S2: System initialization, establishing initial transformation relationships between various coordinate systems;

[0014] S3: During the detection process, all cameras are triggered to capture images simultaneously;

[0015] S4: Perform sub-pixel analysis on the acquired image and extract the coordinates of each detection point on the corresponding camera image plane;

[0016] S5: Construct an error equation based on the collinearity equation of photogrammetry, where the unknowns include the three-dimensional displacement of all detection points, the displacement and attitude change of each camera group; use the relative pose relationship parameters calculated from the mutual shooting data of the camera groups to jointly solve the unknowns and obtain the three-dimensional deformation data of each detection point under the unified ship coordinate system.

[0017] The beneficial effects of the technical solution in this application are as follows:

[0018] By using a distributed deployment of multiple camera groups and a mutual shooting network, the dependence of traditional methods on line-of-sight conditions and complex wiring is overcome; through sub-pixel interpretation and a unified adjustment model, high-precision (millimeter level), high synchronization (error <0.1 seconds), and real-time (≥1Hz) full-field three-dimensional deformation monitoring is achieved; the system is flexibly deployed and can further calculate overall torsional deformation, providing an effective tool for ship construction and structural safety assessment. Attached Figure Description

[0019] Figure 1 The intention is to set up detection points and cameras for a section of the ship's inner hull.

[0020] Figure 2 The intention is to set up detection points and cameras when viewed from above the ship's inner hull.

[0021] Figure 3 This is a schematic diagram illustrating the principle of dual-camera measurement.

[0022] Figure 4 for Figure 1 A schematic diagram of the camera group at position 2#.

[0023] Figure 5 for Figure 1 A schematic diagram of the structure of monitoring point #24 in the middle section.

[0024] Figure 6 for Figure 2 A schematic diagram of the magnetic attraction component at position 3#.

[0025] Figure 7This is a schematic diagram of the ship's coordinate system.

[0026] Figure 8 This is a schematic diagram of the camera coordinate system.

[0027] Figure 9 Define the camera coordinate system.

[0028] Figure 10 This is a flowchart of a method for detecting deformation of the inner shell of a ship's cargo hold. Detailed Implementation

[0029] Currently, the main methods used for detecting deformation of the ship's internal hull include total station measurement and displacement gauge deformation measurement.

[0030] Total station surveying method: Multiple total station sites are selected within the ship's hold. Control measurements are pre-conducted at these sites using a control network method to establish a coordinate system. Based on the requirements for hull deformation detection, multiple detection points are set up, ensuring that each detection point is measured at least once by a total station at one site. To guarantee the speed and synchronization of data acquisition, the number of detection points allocated to each site should be as equal as possible. It is estimated that completing one cycle of the above measurement will take approximately 5-20 minutes.

[0031] Displacement gauge deformation measurement method: Based on the requirements for detecting deformation of the inner shell, all detection points are set up. These detection points are then combined in pairs or in a one-to-many combination. The type of displacement gauge is selected based on the distance between the combinations of detection points to measure the relative displacement between them.

[0032] The aforementioned traditional methods suffer from drawbacks such as high line-of-sight requirements, long measurement cycles, complex equipment wiring, and poor data synchronization. These limitations make it difficult to achieve efficient and high-precision real-time deformation monitoring in construction environments with obstacles like scaffolding within the ship's cabin. For example, during deformation monitoring, if construction within the ship's cabin is not yet complete, numerous scaffolding and other obstacles may obstruct total station measurements, making it difficult to meet the line-of-sight requirements between the monitoring and station points, and even necessitating the abandonment of some monitoring points. Furthermore, due to the limited number of total stations available, each station needs to measure a large number of monitoring points within a single measurement cycle, resulting in a long cycle completion time and severely impacting the synchronization of monitoring point measurements. Finally, when using displacement gauge deformation measurement, the significant differences in relative distances between monitoring points necessitate the selection of various types of displacement gauges, complicating the selection process. For example, since displacement gauges require power supply and data transmission for real-time measurement, using a large number of displacement gauges for deformation detection of the ship's inner hull results in numerous connecting wires between the gauges, increasing the workload and wiring difficulty of gauge deployment. These shortcomings and drawbacks of existing deformation measurement technologies make these methods difficult to implement in practice.

[0033] Therefore, this application provides a method and system for detecting deformation of the inner shell of a ship's cargo hold. By employing multi-camera group collaborative shooting and sub-pixel image processing technology, a photogrammetry network covering the entire hold is constructed to achieve synchronous measurement of three-dimensional displacement of all detection points.

[0034] First, this application provides a ship cargo hold inner shell deformation detection system, including:

[0035] Multiple inspection points are located on the inner surface of the cargo hold shell;

[0036] At least two camera groups are fixedly mounted on the inner shell surface of the cargo hold; each camera group includes a rigidly connected mounting base and at least two cameras fixed thereon, with the relative poses of the cameras within the same group being fixed;

[0037] The data processing unit is communicatively connected to all of the cameras;

[0038] The fields of view of the at least two camera groups overlap, and they can photograph each other's marker points.

[0039] Each detection point was captured by at least two cameras from different locations;

[0040] The data processing unit is configured to: synchronously acquire images from each camera; based on the sub-pixel coordinates of the detection points in the images, combined with parameters characterizing the relative pose relationships between the camera group, and through adjustment calculations, uniformly calculate the three-dimensional displacement of all detection points in the ship's coordinate system. The above technical solution constructs a full-field deformation detection system adaptable to complex cabin environments. By distributing the cameras, it completely avoids the requirement of long-distance unobstructed line-of-sight in traditional optical measurements, allowing the measuring equipment to be deployed close to the cabin walls, maintaining only a partial line of sight with the luminous marker points, greatly improving the feasibility of construction in complex environments.

[0041] In some implementations, the detection point is an actively luminous marker, comprising an LED light source and a magnetic component for magnetic fixation. In the complex lighting conditions of a ship's cabin, the actively luminous marker offers an extremely high signal-to-noise ratio, and magnetic fixation is particularly suitable for steel-structured ship cabin environments.

[0042] In some implementations, the system also includes at least one single camera, independent of the camera group setup, for capturing detection points not fully covered by the camera group. This enhances the system's deployment flexibility and adaptability to complex spatial structures, ensuring comprehensive monitoring of all critical corners within the cabin, particularly at the edges of the group's field of view, such as the top and front.

[0043] In some embodiments, the mounting base includes a base plate and a support surface angled to the base plate, with some cameras fixed to the support surface to observe detection points in different directions. Simultaneous multi-directional coverage of a local space can be achieved using a minimal number of cameras.

[0044] In some embodiments, the number of camera groups is 4, and each camera group includes 9 cameras, of which 6 cameras are mounted on the curved panel for observing the corner points of the end face inside the cabin, and 3 cameras are mounted on the base plate for observing the other camera groups.

[0045] This application also provides a method for detecting deformation of the inner shell of a ship's cargo hold, using the system provided by the above technical solution, including the following steps:

[0046] S1: The detection point and at least two camera groups are deployed in the inner shell of the cargo hold to ensure that the fields of view of each camera group overlap and that they can photograph the marker points on each other's groups, and to ensure that each detection point is observed by at least two cameras.

[0047] S2: System initialization, establishing initial transformation relationships between various coordinate systems;

[0048] S3: During the detection process, all cameras are triggered to capture images simultaneously;

[0049] S4: Perform sub-pixel analysis on the acquired image and extract the coordinates of each detection point on the corresponding camera image plane;

[0050] S5: Construct an error equation based on the collinearity equation of photogrammetry, where the unknowns include the three-dimensional displacement of all detection points, the displacement and attitude change of each camera group; use the relative pose relationship parameters calculated from the mutual shooting data of the camera groups to jointly solve the unknowns and obtain the three-dimensional deformation data of each detection point under the unified ship coordinate system.

[0051] In some implementations, step S2 includes:

[0052] S21: Based on the images captured by the camera group and the coordinates of the detection points, a spatial resection algorithm combining forward and backward intersection is used to calculate the initial position parameters (d) of each camera group coordinate system relative to the ship's coordinate system. x d y d z ) and initial attitude parameters (ω) x ω y ω z );

[0053] S22: Based on the parameters calculated in step S21 and the fixed installation relationship of each camera within the camera group, determine the initial transformation matrix between the camera coordinate system and the ship coordinate system for each camera. This method can determine the initial position and attitude of each camera group in the ship coordinate system with high precision using relatively low-precision design coordinates and images captured by the system itself, significantly reducing the initial accuracy requirements and engineering costs of system installation.

[0054] In some implementations, the error equation in step S5 is: V = B * d X –L;

[0055] Where V is the observation residual vector, B is the design matrix, and its elements consist of the direction vector between the camera and the detection point and the coordinate transformation matrix, d X Let L be the vector of unknowns containing all the displacement and attitude change parameters to be determined, and let L be the vector of observed values ​​consisting of the changes in image point coordinates; the deformation displacement d of the detection point is obtained through least squares adjustment. X .

[0056] In some implementations, in step S5, the displacement and attitude changes of each camera group obtained by joint calculation are used to calculate the overall torsional deformation parameters of the inner hull of the ship.

[0057] In some implementations, in step S5, the calculation is performed at a frequency of not less than 1 Hz, and the synchronization time error of all camera shots is less than 0.1 seconds, which is suitable for the performance boundary of dynamic process monitoring, enabling it to capture the deformation of the ship in quasi-static or slow dynamic processes such as displacement, buoyancy, and loading tests.

[0058] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0059] It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0060] The present application will be described in detail below with reference to the accompanying drawings and embodiments.

[0061] 1) Camera group and detection point layout:

[0062] This embodiment uses the No. 1 cargo hold of a certain type of ship (approximately 32.1m x 22.9m x 29.4m) as an example for illustration, but this application is not limited to this specific size.

[0063] See Figure 1 and Figure 2 The ship's inner hull is designed with 60 detection points, 4 camera groups, and 10 individual cameras. All 60 detection points are located at the corners of various surfaces on the top of the inner hull. The 4 camera groups coincide with the locations of detection points 1#, 2#, 3#, and 4#, respectively, with each group containing 9 cameras. Individual cameras only photograph the detection points directly in front of the top surface; 10 individual cameras are simultaneously positioned at detection points 5# to 14#.

[0064] 2) Camera parameter settings are as follows:

[0065] The camera has a focal length of f=120mm, a photo resolution of 4024*3036, and a pixel accuracy of no more than 0.7mm / pixel at a maximum viewing distance of about 35 meters.

[0066] 3) Summary of detection methods:

[0067] To ensure detection accuracy, its field of view is designed to ensure:

[0068] a. Each detection point needs to be photographed and measured by two or more cameras. These cameras can be single cameras or cameras in a group (the cameras taking the photos are not required to be in the same group).

[0069] b. Subpixel interpretation technology is used for detection points and image interpretation.

[0070] c. By using two or more camera groups to photograph the same detection point, the relative attitude relationship of the camera groups is established, and the overall torsional deformation of the cabin is calculated.

[0071] d. Display the displacement of the detection point and its real-time changes through pseudo-3D display and digital annotation.

[0072] e. The original data is saved so that the deformation of each detection point can be recalculated later.

[0073] See Figure 3 The principle is illustrated below. Based on the principle of close-range photogrammetry, the image of the detected point P in two photos within the same coordinate system can be used to calculate the position of point P in the natural coordinate system through the transformation relationship between the camera coordinate system and the natural coordinate system. The specific structure of the camera group is as follows: Figure 4 As shown, the system includes a mounting base with a base plate and a curved support surface. The camera is secured by screws, and its relative position and orientation remain constant during measurement. This design allows cameras within a group to observe detection points in different directions and other groups.

[0074] This solution uses mutual measurements between cameras or camera groups to solve for the coordinate and orientation differences caused by inconsistencies in the camera coordinate systems. This greatly reduces the initial accuracy requirements for the position and orientation of the cameras.

[0075] Employing sub-pixel photo interpretation technology, the interpretation accuracy of detection points within a photo can reach 0.5 pixels. By controlling the field of view, the pixel accuracy is made less than 1mm / pixel. Therefore, theoretically, the detection accuracy can reach 0.5mm without considering other influencing factors.

[0076] 4) The design of the camera group and detection point markers provided in this application is as follows:

[0077] Each camera group consists of 9 cameras, 1 detection point light source, and 1 microcomputer. The 9 cameras in the camera group are fixed on specially designed mounting bases, and their relative positions and orientation remain unchanged throughout the operation of the detection system.

[0078] The mounting base includes a base plate with four connectors at its four corners for magnetically securing the base to the detection position. A curved panel perpendicular to the base plate houses six cameras with adjustable angles to ensure each camera's axis is aligned with one of the six corner observation points on the end face. These cameras work in conjunction with another camera group at the top to ensure each corner point is observed simultaneously by two cameras. Three cameras are mounted on the base plate, also with adjustable angles to ensure simultaneous observation of the other three camera groups on the top face. The overall measurement field is constructed through cross-measuring among the four camera groups. The cameras are connected to the mounting base using screws. Each camera group on the base plate has a red omnidirectional light source detection point, and each measurement point consists of two unidirectional light source monitoring points.

[0079] Considering that the ambient light inside the ship's cabin is relatively uniform and does not change much between day and night, red light sources are used to mark the detection points.

[0080] The testing point consists of an acrylic lampshade, LED beads, an aluminum lamp post, a base connecting plate, an aluminum lamp holder, and a neodymium magnet, which is magnetically attached to the position to be tested by the neodymium magnet.

[0081] 5) The data processing method provided in this application is detailed below:

[0082] See the definition of the hull coordinate system. Figure 7 The origin of the hull coordinate system is set at the center of the ship's bottom plate. The X-axis is located on the longitudinal centerline of the ship's compartment, pointing towards the bow. The Y-axis is perpendicular to the X-axis to the right, and the Z-axis is perpendicular to the top. The hull coordinate system is denoted as O-XYZ and is a left-handed coordinate system. The elements of the hull coordinate system are defined in the initial state of the hull and this definition does not change with the deformation of the hull.

[0083] See camera group coordinate system definition. Figure 8 A virtual 3D spatial coordinate system, called the camera group coordinate system, is established on the camera group mounting base. During camera group deployment and installation, the initial state of the camera group coordinate system is made as consistent as possible with the ship's coordinate system. At this point, the three coordinate axes of the camera group coordinate system are basically coincident with the ship's coordinate axes. A center point is set on the camera mounting base, and the coordinates of this center point in the ship's coordinate system are the coordinates of the camera group's center point. The camera group coordinate system is still denoted as O-XYZ.

[0084] The camera coordinate system has its origin at the intersection of the camera's principal optical axis and the photograph, with the vertical and horizontal axes of the photograph designated as U and V, respectively, and the W axis perpendicular to the photograph's direction outward from the lens. Pixels are used as the coordinate unit, forming an O-UVW left-handed spatial coordinate system. Through camera calibration and correction, the origin of the camera coordinate system can be corrected to the center of the photograph.

[0085] Let the pixel precision of U and V be 1. ,get:

[0086] (1)

[0087] At this point, the unit of measurement for u and v is the meter. and Both are camera coordinates, the former in pixels and the latter in meters.

[0088] 6) The relationship between the camera coordinate system and the camera group coordinate system provided in this application is detailed below:

[0089] The camera is fixedly mounted on the base plate of the camera group, so the relationship between the camera coordinate system and the coordinate system of the camera group remains unchanged.

[0090] The camera mounting position is designed so that the origin of the camera coordinate system is located at the center of the camera group, that is, the origin of the camera coordinate system (0, 0, 0) coincides with the origin of the camera group coordinate system (x0, y0, z0).

[0091] Therefore, the origins of the camera coordinate systems of different cameras in the same camera group coincide, but due to the different camera mounting orientations, the coordinate axes of the camera coordinate systems of different cameras are different.

[0092] The relationship between the camera coordinate system of each camera and the camera group coordinate system can be represented by three coordinate rotation angles (α, β, γ), where (α, β, γ) are the rotation angles of the camera coordinate system around the u, w, and v axes of the camera coordinate system in sequence.

[0093] According to Euler's formula, we get:

[0094] (2)

[0095]

[0096] 7) The relationship between the camera group coordinate system and the ship's coordinate system provided in this application is detailed below:

[0097] By installing the camera group using appropriate methods, the initial state of the camera group coordinate system is the same as that of the ship's coordinate system. When the ship deforms, the definition of the ship's coordinate system remains unchanged. At this time, the camera group undergoes displacement and rotation within the ship's coordinate system, thus resulting in a slight translation and rotation between the camera group coordinate system and the ship's coordinate system. The translation amount is set as... Rotation amount set to Each camera group has a total of 6 parameters.

[0098] Considering , Since these are all minute quantities, after the hull deformation, the coordinates X, Y, Z in the hull coordinate system and the coordinates x, y, z in the camera group coordinate system have the following transformation relationship:

[0099] (3)

[0100] 8) The initial coordinates of the detection points provided in this application are detailed below:

[0101] The detection points include the center point of the camera group. All detection points are marked with dedicated markers. Before the ship's relocation, the installation positions of the detection points are designed and installed according to the design drawings. After installation, measurement equipment is used to determine their initial installation positions. The coordinate measurement accuracy of the initial position is 0.5m. After camera-based joint measurement, the coordinate accuracy of the initial position can be improved to 0.1m. Considering that the purpose of deformation detection is to determine the relative displacement of the detection points, the error of this initial position has a negligible impact on the accuracy of the relative displacement calculation.

[0102] 9) Perform system initialization:

[0103] After the pre-ship relocation detection system is installed, it is activated to capture images of the detection points (including the center point of the camera array). The results (U, V) of the detection points on the images are directly read. The distance L between the detection points and the camera array center point is calculated based on the initial coordinates measured on the design drawing. The calibrated distance is then obtained from L. u and v are calculated using equation (1).

[0104] Recalculate based on the calculated w. Used as the initial camera coordinate system coordinates for the detection points.

[0105] The calculated Substitute into equation (3) to calculate the (x, y, z) of the detection point, which will be used as the initial camera group coordinates of the detection point.

[0106] 10) The direction of the camera's main axis in the compartment coordinate system is detailed below:

[0107] See camera coordinate system definition. Figure 9 Camera coordinate system coordinate axis orientation setting: Measurement value numbered k If the camera position (group) number is i and the measurement point position number is j, then the camera's principal optical axis vector is... (From camera to detection point):

[0108]

[0109] The vector of the U-axis in the compartment coordinate system Determined as:

[0110]

[0111] Then the V-axis vector Determined as:

[0112]

[0113] When the camera is positioned on the O-XY plane (W3≈0), the camera's bottom surface is parallel to the O-XY plane, and the U-axis is perpendicular to the O-XY plane, meaning the U-axis is parallel to the OZ-axis. If the U-axis is oriented in the same direction as the Z-axis at this time... = (0, 0, 1), otherwise = (0, 0, -1), then:

[0114] or

[0115] When the camera is positioned on the O-YZ plane (W1≈0), the camera's bottom surface is parallel to the O-YZ plane, and the U-axis is perpendicular to the O-YZ plane, meaning the U-axis is parallel to the OX axis. If the U-axis is oriented in the same direction as the X-axis at this time... = (1, 0, 0), conversely = (-1, 0, 0), then:

[0116]

[0117] When the camera is positioned on the O-ZX plane (W2≈0), the camera's bottom surface is parallel to the O-ZX plane, and the U-axis is perpendicular to the O-ZX plane, meaning the U-axis is parallel to the OY axis. If the U-axis is oriented in the same direction as the Y-axis at this time... = (0, 1, 0), otherwise = (0, -1, 0), then:

[0118]

[0119] 11) The impact of camera attitude changes on the measurement results is detailed below:

[0120] Due to the overall deformation of the compartment, the camera attitude changes accordingly at different locations, and this change varies depending on its location. Camera attitude changes are generally represented by three attitude angles, namely ( These are the rotation angles around the X, Y, and Z axes of the compartment coordinate system, respectively. At this point, the vector... The general formula for the influence of changes in camera status is:

[0121]

[0122] because( Since all angles are small, the above formula can be simplified to:

[0123]

[0124] have to:

[0125]

[0126]

[0127]

[0128] in, This can be seen as the displacement of the target caused by changes in camera posture.

[0129] Changes in camera pose also affect the image coordinate system. , The vector changes, but considering ( Since all angles are small, this effect can be ignored.

[0130] Since the attitude changes of cameras in the same camera group are the same, the formula for calculating the influence is also the same.

[0131] Transformation relationship between the compartment coordinate system and the camera coordinate system

[0132] Since the deformation of the detection point obtained by camera measurement belongs to the camera coordinate system and the deformation of the detection point belongs to the compartment coordinate system, a mutual transformation system should be established between the two.

[0133] Let the transformation relationship from the compartment coordinate system to the camera coordinate system be:

[0134]

[0135] Given that the unit vectors of the three coordinate axes of the compartment coordinate system are:

[0136]

[0137] Right now:

[0138]

[0139] have to:

[0140]

[0141] 12) The adjustment model for deformation calculation is detailed below:

[0142] Let camera number t, camera group number i, detection point number j, and measurement value number k. The location of camera group number i is also the location of detection point i.

[0143] In the camera coordinate system, the measured value is defined as... The change in camera measurements due to compartment deformation is quantitatively defined as: .

[0144] The camera position deformation is defined as follows: ;

[0145] The deformation of the detection point position is defined as follows: ;

[0146] The influence of camera pose change on the measurement results is defined as follows: .

[0147] Therefore, the relationship between the change in camera measurements and the influence of changes in the detection point and camera attitude is as follows:

[0148]

[0149]

[0150]

[0151] set up:

[0152]

[0153] have to:

[0154]

[0155] 13) Error equation, let:

[0156]

[0157]

[0158]

[0159]

[0160]

[0161] The error equation is then obtained as follows:

[0162] In the above settings, m represents the total number of detection points, n represents the total number of measured values, dX represents the deformation displacement of the detection point, L represents the actual measured value from the camera, V represents the correction value of the measured value, and B represents the coefficient of the error equation.

[0163] 14) Perform adjustment calculations:

[0164] a. Weighting of measured values

[0165] The weights of the measured values ​​u and v are the reciprocals of the distance from the camera to the detection point, that is:

[0166] b. Solution of normal equations

[0167]

[0168]

[0169]

[0170] The dX calculated using the above formula is the deformation displacement of the detection point.

[0171] In summary, the solution provided in this application adopts the basic principle of close-range photogrammetry, using multiple cameras to form a camera group, with multiple groups measuring simultaneously, ensuring a large overlap in the measurement area, and mutual verification; the relative deformation measurement resolution of the ship's inner hull is not less than 1mm, the overall deformation measurement accuracy of the ship's inner hull is not less than 2mm, the output frequency of the measurement data is 1Hz, and the data synchronization time error is less than 0.1 seconds; camera deformation detection is adopted, during deformation detection, the camera and the detection point are close to the inner wall of the ship's hull, and are not affected by obstacles such as scaffolding that have not been removed inside the ship's hull, making the selection of the location of the detection point and the camera deployment point more flexible and simple. The system employs camera deformation detection, enabling precise synchronization of camera images with a synchronization time error of less than 0.1 seconds. Its simple, flexible, and easily expandable design allows it to adapt to the shapes and sizes of different ship hulls. Only the cameras require power and data transmission, minimizing deployment workload and improving inspection efficiency. The system utilizes multi-camera groups and mutual image capture measurements, simultaneously calculating relative deformation between detection points and overall deformation of the ship's inner hull. It features complete real-time data recording and, with software functionality, enables 1:1 post-processing playback.

[0172] Therefore, the technical solution provided in this application has high industrial application value because it effectively overcomes the various shortcomings of the prior art.

[0173] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A ship cargo hold inner hull deformation detection system, characterized in that, include: Multiple inspection points are located on the inner surface of the cargo hold shell; At least two camera groups are fixedly mounted on the inner shell surface of the cargo hold; each camera group includes a rigidly connected mounting base and at least two cameras fixed thereon, with the relative poses of the cameras within the same group being fixed; The data processing unit is communicatively connected to all of the cameras; The fields of view of the at least two camera groups overlap, and they can photograph each other's marker points. Each detection point was captured by at least two cameras from different locations; The data processing unit is configured to: synchronously acquire images from each camera; based on the sub-pixel coordinates of the detection points in the images, combined with parameters characterizing the relative pose relationship between the camera groups, uniformly calculate the three-dimensional displacement of all detection points in the ship's coordinate system through adjustment calculation.

2. The ship cargo hold inner hull deformation detection system according to claim 1, characterized in that, The detection point is an active light-emitting marker, including an LED light source and a magnetic component for magnetic fixation.

3. The ship cargo hold inner hull deformation detection system according to claim 1, characterized in that, The system also includes at least one single camera, independent of the camera group setup, for capturing images of detection points not fully covered by the camera group.

4. The ship cargo hold inner hull deformation detection system according to claim 1, characterized in that, The mounting base includes a base plate and a support surface set at an angle to the base plate, and some cameras are fixed on the support surface to observe detection points in different directions.

5. The ship cargo hold inner hull deformation detection system according to claim 1, characterized in that, The number of camera groups is 4, and each camera group includes 9 cameras. Six cameras are mounted on the curved panel to observe the corner points of the end face inside the cabin, and three cameras are mounted on the base plate to observe the other camera groups.

6. A method for detecting deformation of the inner hull of a ship's cargo hold, employing the system as described in any one of claims 1 to 5, characterized in that, Includes the following steps: S1: The detection point and at least two camera groups are deployed in the inner shell of the cargo hold to ensure that the fields of view of each camera group overlap and that they can photograph the marker points on each other's groups, and to ensure that each detection point is observed by at least two cameras. S2: System initialization, establishing initial transformation relationships between various coordinate systems; S3: During the detection process, all cameras are triggered to capture images simultaneously; S4: Perform sub-pixel analysis on the acquired image and extract the coordinates of each detection point on the corresponding camera image plane; S5: Construct an error equation based on the collinearity equation of photogrammetry, where the unknowns include the three-dimensional displacement of all detection points, the displacement and attitude change of each camera group; use the relative pose relationship parameters calculated from the mutual shooting data of the camera groups to jointly solve the unknowns and obtain the three-dimensional deformation data of each detection point under the unified ship coordinate system.

7. The method for detecting deformation of the inner hull of a ship's cargo hold according to claim 6, characterized in that, Step S2 includes: S21: Based on the images captured by the camera group and the coordinates of the detection points, a spatial resection algorithm combining forward and backward intersection is used to calculate the initial position parameters (d) of each camera group coordinate system relative to the ship's coordinate system. x d y d z ) and initial attitude parameters (ω) x ω y ω z ); S22: Based on the parameters calculated in step S21 and the fixed installation relationship of each camera in the camera group, determine the initial transformation matrix between the camera coordinate system and the ship coordinate system for each camera.

8. The method for detecting deformation of the inner hull of a ship's cargo hold according to claim 6, characterized in that, In step S5, the error equation is: V = B*d X – L; Where V is the observation residual vector, B is the design matrix, and its elements consist of the direction vector between the camera and the detection point and the coordinate transformation matrix, d X Let L be the vector of unknowns containing all the displacement and attitude change parameters to be determined, and let L be the vector of observed values ​​consisting of the changes in image point coordinates; the deformation displacement d of the detection point is obtained through least squares adjustment. X .

9. The method for detecting deformation of the inner hull of a ship's cargo hold according to claim 6, characterized in that, In step S5, the displacement and attitude changes of each camera group are obtained through joint calculation, and then the overall torsional deformation parameters of the inner shell of the ship are calculated.

10. The method for detecting deformation of the inner hull of a ship's cargo hold according to claim 6, characterized in that, In step S5, the calculation is performed at a frequency of not less than 1 Hz, and the synchronization time error of all camera shots is less than 0.1 seconds.

Images