Device and method for automatically measuring compaction settlement by adopting binocular vision

Abstract

The invention provides a device and a method for automatically measuring compaction settlement by adopting binocular vision, and relates to the technical field of computer vision, the compaction settlement is calculated by calling and analyzing interactive data of an equipment scheduling execution module, a visual system construction module, a rammer identification and positioning module, an image tracking and collecting module and a coordinate calculation execution module; and a compaction settlement change curve is obtained. The technical problems that in the prior art, due to the fact that compaction settlement monitoring depends on manual measurement or uses a contact type sensor to conduct compaction settlement monitoring, timeliness of compaction settlement measurement is insufficient, the measurement result is prone to being influenced by manual work, and construction risks in the foundation compaction process are increased are solved. Reliable technical support is provided for real-time monitoring of the foundation compaction effect in the rammer construction process, the automation degree and precision of compaction settlement measurement are achieved, and the technical effects of providing reliable guarantee for the foundation compaction quality in the construction process are achieved.

Description

Technical Field

[0001] The invention relates to the technical field of computer vision, and in particular to a device and method for automatically measuring tamping amount using binocular vision. Background Art

[0002] The existing technology mainly relies on manual operation or contact sensors for monitoring during the tamping amount measurement process, which brings many thorny problems, especially when the construction environment is complex and high-frequency measurements are required.

[0003] Firstly, the manual measurement method is inefficient, requiring construction workers to enter the site to conduct point-to-point ground settlement measurements after the rammer strikes. This is not only time-consuming and labor-intensive, but also presents a greater safety risk. Especially in a construction environment with continuous strikes, workers need to frequently approach the construction area, increasing the possibility of personal injury.

[0004] In addition, the results of manual measurement often produce errors due to differences in the operator's technical level, the accuracy of the measuring tools, and subjective judgment, and the consistency and stability of the data cannot be guaranteed.

[0005] On the other hand, the traditional method that relies on contact sensors also has limitations. Contact sensors usually need to be installed on the ground or on a rammer, which increases the complexity of equipment layout. The sensor itself is easily affected by environmental factors such as construction vibration, dirt, and dust, resulting in inaccurate measurements or equipment damage.

[0006] More importantly, contact sensors cannot provide real-time feedback on settlement data, and there is a lag in data acquisition, making it difficult for construction workers to grasp the foundation compaction effect in a timely manner, resulting in the inability to effectively control construction quality, increasing construction risks during the foundation compaction process, and affecting overall construction efficiency and project quality. Summary of the invention

[0007] The present application provides an automatic measurement device and method for tamping amount using binocular vision, which is used to solve the technical problem that the prior art relies on manual measurement or uses contact sensors to monitor the tamping amount, resulting in insufficient timeliness in the measurement of the tamping amount and the measurement results are easily affected by human factors, thereby increasing the construction risk of the foundation compaction process.

[0008] In view of the above problems, the present application provides an automatic measurement device and method for tamping amount using binocular vision.

[0009] In a first aspect of the present application, a device for automatically measuring tamping amount using binocular vision is provided, the device comprising: an equipment scheduling execution module, for scheduling equipment according to construction environment information of a construction site, and obtaining a first industrial camera and a second industrial camera; a visual system construction module, for locating a target tamping area at the construction site, and after installing the first industrial camera and the second industrial camera to the target tamping area, calibrating the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of a binocular vision system; a tamping hammer identification and positioning module, for locating and embedding tamping hammer identification points according to the structural characteristics of the target tamping hammer, and obtaining K tamping hammer identification points, wherein K is a positive integer greater than or equal to 3; an image tracking and acquisition module, for synchronously tracking and acquiring images of the target tamping hammer through the binocular vision system during the operation of the target tamping hammer, and obtaining a stereoscopic image pair sequence; a coordinate calculation execution module, for calculating a three-dimensional coordinate change sequence of the K tamping hammer identification points based on the stereoscopic image pair sequence; a data calculation execution module, for calculating the tamping amount based on the three-dimensional coordinate change sequence, and obtaining a tamping amount change curve.

[0010] The second aspect of the present application provides an automatic measurement method for tamping amount using binocular vision, the method comprising: performing equipment scheduling according to construction environment information of the construction site to obtain a first industrial camera and a second industrial camera; locating a target tamping area at the construction site, and after installing the first industrial camera and the second industrial camera in the target tamping area, calibrating the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of a binocular vision system; locating and embedding tamping hammer identification points according to the structural characteristics of the target tamping hammer to obtain K tamping hammer identification points, wherein K is a positive integer greater than or equal to 3; during the operation of the target tamping hammer, synchronously tracking and collecting images of the target tamping hammer through the binocular vision system to obtain a stereoscopic image pair sequence; calculating a three-dimensional coordinate change sequence of the K tamping hammer identification points based on the stereoscopic image pair sequence; calculating the tamping amount based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] The device provided in the embodiment of the present application is used to perform equipment scheduling according to the construction environment information of the construction site, and obtain the first industrial camera and the second industrial camera through an equipment scheduling execution module; a visual system construction module is used to locate the target ramming area at the construction site, and after installing the first industrial camera and the second industrial camera to the target ramming area, the first industrial camera and the second industrial camera are calibrated with internal and external parameters to complete the construction of the binocular vision system; a rammer identification and positioning module is used to locate and embed the rammer identification point according to the structural characteristics of the target rammer, and obtain K rammer identification points, wherein K is a positive integer greater than or equal to 3; an image tracking and acquisition module is used to synchronously track and acquire images of the target rammer through the binocular vision system during the operation of the target rammer, and obtain a stereo image pair sequence; a coordinate calculation execution module is used to calculate the three-dimensional coordinate change sequence of the K rammer identification points based on the stereo image pair sequence; a data calculation execution module is used to calculate the tamping amount based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve. It provides reliable technical support for real-time monitoring of the foundation compaction effect during rammer construction, achieves the automation and accuracy of tamping volume measurement, and provides reliable technical guarantee for the foundation compaction quality during construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A schematic diagram of the structure of the automatic measurement device for tamping amount using binocular vision provided in this application;

[0014] Figure 2 A schematic flow chart of the method for automatically measuring tamping amount using binocular vision provided in this application.

[0015] Explanation of the reference numerals: equipment scheduling execution module 11, visual system construction module 12, rammer identification and positioning module 13, image tracking and acquisition module 14, coordinate calculation execution module 15, data calculation execution module 16. DETAILED DESCRIPTION

[0016] The present application provides an automatic measurement device and method for tamping amount using binocular vision, which is used to solve the technical problem that the existing technology relies on manual measurement or uses contact sensors to monitor tamping amount, resulting in insufficient timeliness of tamping amount measurement and the measurement results are easily affected by manual labor, which increases the construction risk of the foundation compaction process. It provides reliable technical support for real-time monitoring of the foundation compaction effect during tamping hammer construction, achieves the degree of automation and accuracy of tamping amount measurement, and provides a reliable guarantee for the foundation compaction quality during construction.

[0017] Below, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all of them.

[0018] Embodiment 1, as Figure 1 As shown, the present application provides an automatic measuring device for tamping amount using binocular vision, the device comprising:

[0019] The equipment scheduling execution module 11 is used to perform equipment scheduling according to the construction environment information of the construction site to obtain the first industrial camera and the second industrial camera.

[0020] The visual system construction module 12 is used to locate the target impact area at the construction site, and after installing the first industrial camera and the second industrial camera at the target impact area, calibrate the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of the binocular vision system.

[0021] The rammer identification and positioning module 13 is used to locate and embed the rammer identification points according to the structural characteristics of the target rammer, and obtain K rammer identification points, where K is a positive integer greater than or equal to 3.

[0022] The image tracking and acquisition module 14 is used to synchronously track and acquire images of the target rammer through the binocular vision system during the operation of the target rammer to obtain a stereoscopic image pair sequence.

[0023] The coordinate calculation execution module 15 is used to calculate and obtain a three-dimensional coordinate change sequence of the K rammer identification points based on the stereoscopic image pair sequence.

[0024] The data calculation execution module 16 is used to calculate the tamping amount based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve.

[0025] In one embodiment, the device scheduling execution module 11 further includes the following operation steps:

[0026] Preset camera-associated environmental indicators, and decompose the construction environment information based on the camera-associated environmental indicators to obtain environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit; perform protection level conversion based on the environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit to obtain dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements; use the dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements as equipment scheduling constraints, and schedule to obtain the first industrial camera and the second industrial camera.

[0027] In one embodiment, the visual system construction module 12 further includes the following operation steps:

[0028] The camera parameters are set according to the striking range of the target rammer in the construction site to obtain the first installation parameters and the second installation parameters; a real-time construction model is constructed according to the construction progress information of the construction site; the striking range of the target rammer in the construction site obtained interactively is fitted to the real-time construction model to select the striking area and locate the target rammer area; a coverage analysis is performed on the target rammer area according to the first installation parameters and the second installation parameters to obtain the camera position optimization parameters; the first industrial camera and the second industrial camera are installed in the target rammer area with the first installation parameters, the second installation parameters and the camera position optimization parameters as installation constraints.

[0029] In one embodiment, the visual system construction module 12 further includes the following operation steps:

[0030] Taking the first installation parameter and the second installation parameter as installation constraints, construct the first camera model and the second camera model of the first industrial camera and the second industrial camera; perform camera perspective simulation in the process of randomly moving the first camera model and the second camera model in the target ramming area to obtain a first perspective coverage sequence and a second perspective coverage sequence; after randomly arranging and combining the first perspective coverage sequence and the second perspective coverage sequence, superimpose the perspective coverage areas to obtain a perspective overlapping area sequence; use the target ramming area to traverse the perspective overlapping area sequence to locate the first position parameter and the second position parameter; the first position parameter and the second position parameter constitute the camera position optimization parameters.

[0031] In one embodiment, the rammer identification and positioning module 13 further includes the following operation steps:

[0032] Calling a rammer model library, wherein the rammer model library includes multiple sample rammer models and multiple sample identification point arrays; obtaining a target model by modeling according to the structural features of the target rammer; using the target model to traverse and align multiple sample rammer models in the rammer model library to locate the target sample model with the highest overlap; calling the target identification point array of the target sample model in the multiple sample identification point arrays; after fitting the target identification point array to the target model, embedding the K rammer identification points in the target rammer according to the relative position relationship of the target identification point array in the target model.

[0033] In one embodiment, the data calculation execution module 16 further includes the following operation steps:

[0034] The vibration displacement pair sequence of the binocular vision system is called according to the image acquisition time interval of the stereoscopic image pair sequence; mapping displacement compensation of the three-dimensional coordinate change sequence is performed according to the vibration displacement pair sequence to obtain a compensated coordinate change sequence; and tamping amount is calculated based on the compensated coordinate change sequence to obtain a tamping amount change curve.

[0035] In one embodiment, the coordinate calculation execution module 15 further includes the following operation steps:

[0036] Mark the K rammer identification points in the stereo image pair sequence, and perform frequency statistics of the rammer identification points based on the marked stereo image pair sequence to obtain K overlapping recurrence frequencies; preset an application reliable frequency threshold, and obtain M rammer identification points based on the K overlapping recurrence frequencies and the application reliable frequency threshold, wherein K≥M≥3 and M is a positive integer; and calculate based on the stereo image pair sequence to obtain a three-dimensional coordinate change sequence of the M rammer identification points.

[0037] Embodiment 2, based on the same inventive concept as the automatic measuring device for tamping amount using binocular vision in the above embodiment, Figure 2 As shown, the present application provides a method for automatically measuring tamping amount using binocular vision, the method comprising:

[0038] A100: Perform equipment scheduling according to the construction environment information of the construction site to obtain the first industrial camera and the second industrial camera.

[0039] In one embodiment, equipment scheduling is performed according to the construction environment information of the construction site to obtain a first industrial camera and a second industrial camera. Step A100 of the method provided in this application includes:

[0040] Preset camera-associated environmental indicators, and decompose the construction environment information based on the camera-associated environmental indicators to obtain environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit; perform protection level conversion based on the environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit to obtain dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements; use the dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements as equipment scheduling constraints, and schedule to obtain the first industrial camera and the second industrial camera.

[0041] Specifically, in this embodiment, the camera-related environmental indicators are environmental indicators related to camera performance, including dust, humidity, vibration, temperature and light intensity. The construction environment information is decomposed based on the camera-related environmental indicators to provide a basis for equipment selection, and specifically obtains the environmental dust limit (the upper limit of dust concentration in the construction environment), the environmental humidity limit (the upper limit of humidity in the construction environment), the environmental vibration limit (the upper limit of vibration intensity in the construction environment), the environmental temperature limit (the temperature range in the construction environment), and the environmental light intensity limit (the range of light intensity in the construction environment).

[0042] The purpose of protection level conversion is to convert environmental indicators into specific protection level requirements so as to select a camera that meets these requirements. This embodiment performs protection level conversion based on the environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit to obtain dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements, wherein the dustproof level requirement is to determine the dustproof level of the camera according to the dust limit (such as IP5X or IP6X), the waterproof level requirement is to determine the waterproof level of the camera according to the humidity limit (such as IPX5 or IPX7), the shockproof level requirement is to determine the shockproof level of the camera according to the vibration limit (such as complying with MIL-STD-810G standard), the temperature resistance level requirement is to determine the temperature resistance range of the camera according to the temperature limit, and the light resistance level requirement is to determine the light resistance performance of the camera according to the light resistance limit.

[0043] Then, according to the protection level requirements obtained by the conversion, the constraints of the equipment scheduling are set to query the camera models that meet all protection level requirements in the equipment library, and then two industrial cameras are called based on the camera models to obtain the first industrial camera and the second industrial camera.

[0044] This embodiment achieves the technical effect of improving the compatibility of binocular vision device components with application scenarios.

[0045] A200: After locating the target impact area at the construction site and installing the first industrial camera and the second industrial camera at the target impact area, calibrate the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of the binocular vision system.

[0046] In one embodiment, a target impact area is located at the construction site, and the first industrial camera and the second industrial camera are installed at the target impact area. Step A200 of the method provided in the present application includes:

[0047] The camera parameters are set according to the striking range of the target rammer in the construction site to obtain the first installation parameters and the second installation parameters; a real-time construction model is constructed according to the construction progress information of the construction site; the striking range of the target rammer in the construction site obtained interactively is fitted to the real-time construction model to select the striking area and locate the target rammer area; a coverage analysis is performed on the target rammer area according to the first installation parameters and the second installation parameters to obtain the camera position optimization parameters; the first industrial camera and the second industrial camera are installed in the target rammer area with the first installation parameters, the second installation parameters and the camera position optimization parameters as installation constraints.

[0048] In one embodiment, coverage analysis is performed on the target tamping area according to the first installation parameter and the second installation parameter to obtain camera position optimization parameters. The method step A200 provided in the present application includes:

[0049] Taking the first installation parameter and the second installation parameter as installation constraints, construct the first camera model and the second camera model of the first industrial camera and the second industrial camera; perform camera perspective simulation in the process of randomly moving the first camera model and the second camera model in the target ramming area to obtain a first perspective coverage sequence and a second perspective coverage sequence; after randomly arranging and combining the first perspective coverage sequence and the second perspective coverage sequence, superimpose the perspective coverage areas to obtain a perspective overlapping area sequence; use the target ramming area to traverse the perspective overlapping area sequence to locate the first position parameter and the second position parameter; the first position parameter and the second position parameter constitute the camera position optimization parameters.

[0050] Specifically, in this embodiment, based on the striking range of the target rammer, the first industrial camera and the second industrial camera need to be parameterized, and the first installation parameters include the camera's focal length, resolution, shooting angle, viewing angle range and other basic parameters, the camera's physical installation position, installation height, angle from the target and other installation parameters, and the first installation parameters are consistent with the second installation parameters. By setting the first installation parameters and the second installation parameters, it is ensured that the camera can cover the target striking area and meet subsequent image acquisition requirements.

[0051] A real-time construction model is constructed based on the construction progress information of the construction site. The real-time construction model is a dynamic digital model of the construction site, which reflects the current construction status and progress information. It can help understand the actual striking range of the rammer, construction path and other information, and ensure that the camera can adapt to construction changes on site when installed.

[0052] The striking range of the target rammer in the construction site obtained interactively is fitted to the real-time construction model to better define the ramming area, thereby optimizing the installation position of the camera. Based on the fitted real-time construction model, the actual striking area of ​​the rammer can be determined and the target ramming area can be accurately located.

[0053] A coverage analysis is performed in the target ramming area according to the first installation parameter and the second installation parameter. The core of the coverage analysis is to ensure that the perspectives of the two industrial cameras can overlap as much as possible in the target area and form an accurate stereoscopic vision. In order to find the best camera position, it is necessary to perform perspective coverage simulation and camera position optimization. The specific method is as follows:

[0054] Using the first installation parameters and the second installation parameters as installation constraints, construct a "virtual model" of the first industrial camera and the second industrial camera to obtain the first camera model and the second camera model. By adjusting the position and angle of the first camera model and the second camera model, the actual industrial camera viewing angle can be simulated.

[0055] In the target impact area, the camera model is randomly moved to simulate the viewing angle coverage under different installation positions. Through this simulation, the "first viewing angle coverage sequence" and "second viewing angle coverage sequence" can be generated.

[0056] Next, the two view sequences are randomly combined to simulate the overlapping of the views of the two cameras at different positions. By superimposing different view angles, the overlapping area of ​​the camera views can be determined and a view angle overlapping area sequence can be obtained.

[0057] By traversing the entire ramming area and analyzing the overlapping area of ​​the perspective, the optimal installation position of the camera is finally located to obtain the first position parameter and the second position parameter. The industrial camera is deployed based on the first position parameter and the second position parameter to obtain a perspective that completely covers the target ramming area (maximizing the overlapping coverage of the perspective). The first position parameter and the second position parameter constitute the final camera position optimization parameters.

[0058] The first installation parameter, the second installation parameter and the camera position optimization parameter are used as installation constraints, and the first industrial camera and the second industrial camera are installed in the target ramming area.

[0059] This embodiment achieves the technical effect of selecting the optimal camera installation position based on the dynamic information of the construction site and the actual striking range of the rammer, ensuring that the binocular vision system can accurately cover the entire tamping area, and providing reliable visual data for subsequent image acquisition and tamping amount measurement.

[0060] This embodiment uses the first installation parameter, the second installation parameter and the camera position optimization parameter as installation constraints. After installing the first industrial camera and the second industrial camera to the target tamping area, the first industrial camera and the second industrial camera are calibrated with internal parameters respectively, so as to correct the optical distortion and pixel geometric relationship of the camera, ensure that the first industrial camera and the second industrial camera can accurately reflect the real geometric information in the image when working alone, and perform collaborative external parameter calibration on the first industrial camera and the second industrial camera, so that the fields of view of the two cameras can be accurately aligned to form an overlapping field of view of binocular vision. The preferred method is to control the first industrial camera and the second industrial camera to simultaneously capture the same mark point image, so as to determine the external parameters between the first industrial camera and the second industrial camera by reference to calculate the relative position and angle of the two cameras in space. After completing the internal and external parameter calibration, the binocular vision system is formally constructed, and the binocular vision system obtains the three-dimensional coordinates of the target tamping hammer through parallax calculation, forming accurate spatial perception capabilities, and providing a basis for subsequent motion tracking and tamping amount calculation.

[0061] A300: Position and embed the rammer identification points according to the structural characteristics of the target rammer, and obtain K rammer identification points, where K is a positive integer greater than or equal to 3.

[0062] In one embodiment, the rammer identification points are located and embedded according to the structural features of the target rammer to obtain K rammer identification points. The method step A300 provided in the present application includes:

[0063] Calling a rammer model library, wherein the rammer model library includes multiple sample rammer models and multiple sample identification point arrays; obtaining a target model by modeling according to the structural features of the target rammer; using the target model to traverse and align multiple sample rammer models in the rammer model library to locate the target sample model with the highest overlap; calling the target identification point array of the target sample model in the multiple sample identification point arrays; after fitting the target identification point array to the target model, embedding the K rammer identification points in the target rammer according to the relative position relationship of the target identification point array in the target model.

[0064] Specifically, in this embodiment, a rammer model library is called online, and the rammer model library contains multiple rammer models and corresponding identification point arrays, specifically including multiple sample rammer models and multiple sample identification point arrays, each rammer model represents rammers of different types, shapes or sizes, and each model has predefined identification points, and the identification points are usually selected at key structural parts of the rammer, such as edges, protrusions or surface marks.

[0065] A 3D model of the target rammer (target model) is created based on the structural features of the target rammer, such as its shape, size, weight, surface details, etc. The target model will be used to compare with the sample models in the rammer model library.

[0066] By comparing the 3D model of the target rammer with the sample models in the model library, the model closest to the target rammer structure is found. This process is completed by calculating the geometric overlap between the two, and finally locating the target sample model with the highest overlap.

[0067] Call the target recognition point array of the target sample model in the multiple sample recognition point arrays, and "fit" the target recognition point array to the target model of the target rammer. This requires considering the subtle differences in the structures of the two, and aligning the recognition point array with the surface features of the target rammer through geometric transformations (such as scaling, rotation, and translation). After fitting the target recognition point array to the target model, embed the K rammer recognition points in the target rammer according to the relative position relationship of the target recognition point array in the target model. The so-called embedding refers to marking K rammer recognition points on the surface of the target rammer. The K rammer recognition points will be used for subsequent tracking and image processing of the binocular vision system, where K is a positive integer greater than or equal to 3, meaning that at least 3 feature points can be tracked by the binocular vision system.

[0068] This embodiment models the structural features of the target rammer, combines the sample models in the rammer model library, finds the most suitable identification point array, and embeds these identification points on the target rammer, thereby achieving the technical effect of ensuring that the rammer can be accurately captured and tracked by the system during movement.

[0069] A400: During the operation of the target rammer, the image of the target rammer is synchronously tracked and collected by the binocular vision system to obtain a stereoscopic image pair sequence.

[0070] Specifically, in this embodiment, the binocular vision system tracks the movement of the target rammer in real time and synchronously collects images of the rammer during its operation. Specifically, the first industrial camera and the second industrial camera in the binocular vision system synchronously capture the images of the target rammer during its movement. Since the cameras shoot from different perspectives, a pair of stereo image pairs (stereoimage pairs) are formed, and the stereo image pairs contain the movement information of the rammer in space. In order to ensure that the images can be accurately superimposed, the cameras need to be time synchronized, and time synchronization can be achieved through a hardware trigger system or software clock synchronization.

[0071] During the operation of the rammer, the binocular camera will continuously capture images to generate a series of stereo image pair sequences. Each image pair represents the image of the rammer from two different perspectives at the same time point. The subsequent three-dimensional coordinate calculation will be performed based on the stereo image pair sequence.

[0072] A500: Calculate and obtain a three-dimensional coordinate change sequence of the K rammer identification points based on the stereoscopic image pair sequence.

[0073] In one embodiment, the three-dimensional coordinate change sequence of the K rammer identification points is calculated based on the stereoscopic image pair sequence. The method step A500 provided in the present application includes:

[0074] Mark the K rammer identification points in the stereo image pair sequence, and perform frequency statistics of the rammer identification points based on the marked stereo image pair sequence to obtain K overlapping recurrence frequencies; preset an application reliable frequency threshold, and obtain M rammer identification points based on the K overlapping recurrence frequencies and the application reliable frequency threshold, wherein K≥M≥3 and M is a positive integer; and calculate based on the stereo image pair sequence to obtain a three-dimensional coordinate change sequence of the M rammer identification points.

[0075] Specifically, in step A400 of this embodiment, a sequence of stereoscopic image pairs of the target rammer during operation has been obtained through synchronous shooting with a binocular camera, and each pair of stereoscopic images records the state of the rammer at the same time point.

[0076] Based on the K rammer identification points embedded in the previous A300 step, an image processing algorithm (such as Harris corner detection, SIFT or SURF feature matching) is used to identify and mark the K rammer identification points in the stereo image pair.

[0077] In each pair of stereo images, two cameras will capture feature points from different angles, and these points will be binocularly matched through image processing algorithms to ensure that the feature points marked in the left and right images are the same physical points, avoiding repeated statistics of overlapping recurrence frequencies.

[0078] Due to environmental factors (such as occlusion, light changes, etc.), some feature points are not successfully identified in some images. Therefore, for the stereo image pair sequence, check whether each feature point is successfully identified in all stereo image pairs, count the frequency of occurrence of each feature point, and count their overlapping recurrence frequencies in different frames, that is, the number of times the feature point appears in the entire image sequence, and obtain K overlapping recurrence frequencies.

[0079] According to the overlapping recurrence frequency of each feature point, the feature points below the application reliable frequency threshold are eliminated, and finally M reliable feature points are screened out, where K≥M≥3, and M is a positive integer greater than or equal to 3.

[0080] For the M reliable feature points selected, the disparity is calculated through the stereo image pair of the binocular camera. The three-dimensional coordinates of the M reliable feature points in space can be calculated by combining the calibrated internal and external parameters of the camera with the disparity. The larger the disparity, the closer the object is to the camera; the smaller the disparity, the farther the object is from the camera.

[0081] The three-dimensional coordinates of the M reliable feature points will change over time. By calculating the three-dimensional coordinates of the M reliable feature points in each frame of the image, the time series of the three-dimensional coordinates of the M reliable feature points forms a three-dimensional coordinate change sequence of the rammer identification point.

[0082] The three-dimensional coordinate change sequence not only reflects the spatial position change of each feature point, but also can deduce the motion trajectory, velocity, acceleration and other information of the rammer.

[0083] This embodiment ensures that the final three-dimensional coordinate change sequence has a high degree of accuracy and reliability through multiple screening and calculation, achieving the technical effect of providing data support for subsequent tamping amount calculation.

[0084] A600: Calculate the tamping amount based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve.

[0085] In one embodiment, the tamping amount is calculated based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve. The method step A600 provided in the present application includes:

[0086] The vibration displacement pair sequence of the binocular vision system is called according to the image acquisition time interval of the stereoscopic image pair sequence; mapping displacement compensation of the three-dimensional coordinate change sequence is performed according to the vibration displacement pair sequence to obtain a compensated coordinate change sequence; and tamping amount is calculated based on the compensated coordinate change sequence to obtain a tamping amount change curve.

[0087] Specifically, in this embodiment, first, the time interval of image acquisition is determined according to the stereo image pair sequence previously acquired by the binocular vision system, and the vibration displacement pair sequence within the time interval is called. The vibration displacement pair sequence records the relative displacement information between the ground and the rammer during the rammer impact process, reflecting the tiny displacement error caused by the vibration.

[0088] Next, the vibration displacement sequence is mapped to the three-dimensional coordinate change sequence for displacement compensation. This compensation process matches the vibration data with the three-dimensional coordinate change data to eliminate the influence of the ground vibration on the measurement data when the rammer hits, thereby obtaining a more accurate compensation coordinate change sequence.

[0089] Based on the coordinate change sequence after compensation, the motion trajectory of the tamping hammer in each frame of the image is analyzed, with particular attention paid to its displacement change in the Z-axis direction (vertical direction). The actual tamping amount after each tamping is calculated by analyzing the vertical displacement difference of the compensated three-dimensional coordinates.

[0090] Finally, the tamping settlement change results after each tamping hammer blow are accumulated and plotted into a tamping settlement change curve. The tamping settlement change curve not only shows the real-time settlement effect of the tamping hammer during the construction process, but also reflects the trend of the ground being gradually compacted after multiple blows.

[0091] This embodiment uses a binocular vision system combined with vibration displacement compensation to automatically and accurately measure the tamping amount, thereby providing reliable technical support for real-time monitoring of the foundation compaction effect during tamping hammer construction, achieving the degree of automation and accuracy in tamping amount measurement, and providing a reliable technical effect for the foundation compaction quality during construction.

[0092] Any of the methods or steps described above may be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs may be recognized by various types of computer processors to implement any of the methods or steps described above.

[0093] Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the patent protection scope of the present invention.

Claims

1. The automatic measuring device of tamping amount using binocular vision is characterized in that: The device comprises: An equipment scheduling execution module is used to perform equipment scheduling according to the construction environment information of the construction site, and obtain the first industrial camera and the second industrial camera; A visual system construction module, used for locating a target impact area at the construction site, and after installing the first industrial camera and the second industrial camera at the target impact area, calibrating the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of a binocular vision system; A rammer identification and positioning module is used to locate and embed rammer identification points according to the structural characteristics of the target rammer, and obtain K rammer identification points, where K is a positive integer greater than or equal to 3; An image tracking and acquisition module is used to synchronously track and acquire images of the target rammer through the binocular vision system during the operation of the target rammer to obtain a stereoscopic image pair sequence; A coordinate calculation execution module, used for calculating and obtaining a three-dimensional coordinate change sequence of the K rammer identification points based on the stereoscopic image pair sequence; The data calculation execution module is used to calculate the tamping amount based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve.

2. The automatic measuring device for tamping amount using binocular vision according to claim 1, characterized in that: The device scheduling execution module also includes the following operation steps: Preset camera-associated environmental indicators, and decompose the construction environment information based on the camera-associated environmental indicators to obtain environmental dust limit values, environmental humidity limit values, environmental vibration limit values, environmental temperature limit values, and environmental light limit values; The protection level is converted based on the environmental dust limit, environmental humidity limit, environmental vibration limit, environmental temperature limit and environmental light limit to obtain dustproof level requirements, waterproof level requirements, shockproof level requirements, temperature resistance level requirements and light resistance level requirements; The first industrial camera and the second industrial camera are obtained by scheduling based on the dustproof level requirement, waterproof level requirement, shockproof level requirement, temperature resistance level requirement and light resistance level requirement as equipment scheduling constraints.

3. The automatic measuring device for tamping amount using binocular vision according to claim 1, characterized in that: The visual system building module also includes the following operating steps: Setting camera parameters according to the striking range of the target rammer in the construction site to obtain first installation parameters and second installation parameters; Building a real-time construction model based on the construction progress information of the construction site; Fitting the striking range of the target rammer in the construction site obtained interactively to the real-time construction model to select the striking area, and locate the target striking area; Performing coverage analysis on the target tamping area according to the first installation parameter and the second installation parameter to obtain camera position optimization parameters; The first installation parameter, the second installation parameter and the camera position optimization parameter are used as installation constraints, and the first industrial camera and the second industrial camera are installed in the target ramming area.

4. The automatic measuring device for tamping amount using binocular vision according to claim 3, characterized in that: The visual system building module also includes the following operation steps: Taking the first installation parameter and the second installation parameter as installation constraints, constructing a first camera model and a second camera model of the first industrial camera and the second industrial camera; Performing camera perspective simulation in the process of randomly moving the first camera model and the second camera model in the target tamping area to obtain a first perspective coverage sequence and a second perspective coverage sequence; After randomly arranging and combining the first viewing angle coverage sequence and the second viewing angle coverage sequence, superimposing viewing angle coverage areas to obtain a viewing angle overlapping area sequence; Using the target tamping area to traverse the view angle overlapping area sequence to locate the first position parameter and the second position parameter; The first position parameter and the second position parameter constitute the camera position optimization parameter.

5. The automatic measuring device for tamping amount using binocular vision according to claim 1, characterized in that: The rammer identification and positioning module further includes the following operating steps: Calling a rammer model library, wherein the rammer model library includes a plurality of sample rammer models and a plurality of sample identification point arrays; Modeling according to the structural characteristics of the target rammer to obtain a target model; Using the target model to traverse and align multiple sample rammer models in the rammer model library, and locate the target sample model with the highest overlap; Calling a target recognition point array of the target sample model from the plurality of sample recognition point arrays; After fitting the target recognition point array to the target model, the K rammer recognition points are embedded in the target rammer according to the relative position relationship of the target recognition point array in the target model.

6. The automatic measuring device for tamping amount using binocular vision according to claim 1, characterized in that: The data calculation execution module also includes the following operation steps: calling the vibration displacement pair sequence of the binocular vision system according to the image acquisition time interval of the stereoscopic image pair sequence; Performing mapping displacement compensation on the three-dimensional coordinate change sequence according to the vibration displacement sequence to obtain a compensated coordinate change sequence; The tamping amount is calculated based on the compensation coordinate change sequence to obtain a tamping amount change curve.

7. The automatic measuring device for tamping amount using binocular vision according to claim 1, characterized in that: The coordinate calculation execution module also includes the following operation steps: Marking the K rammer identification points in the stereo image pair sequence, and performing frequency statistics of the rammer identification points based on the marked stereo image pair sequence to obtain K overlapping recurrence frequencies; Preset an application reliable frequency threshold, and select and obtain M rammer identification points according to the K overlapping recurrence frequencies and the application reliable frequency threshold, wherein K≥M≥3, and M is a positive integer; A three-dimensional coordinate change sequence of the M rammer identification points is obtained by calculation based on the stereoscopic image pair sequence.

8. The automatic measurement method of tamping amount using binocular vision is characterized in that: The method comprises: Perform equipment scheduling according to the construction environment information of the construction site to obtain the first industrial camera and the second industrial camera; After locating the target impact area at the construction site and installing the first industrial camera and the second industrial camera at the target impact area, calibrating the internal and external parameters of the first industrial camera and the second industrial camera to complete the construction of the binocular vision system; Positioning and embedding the rammer identification points according to the structural characteristics of the target rammer, and obtaining K rammer identification points, where K is a positive integer greater than or equal to 3; During the operation of the target rammer, the image of the target rammer is synchronously tracked and collected by the binocular vision system to obtain a stereoscopic image pair sequence; Obtaining a three-dimensional coordinate change sequence of the K rammer identification points by calculation based on the stereoscopic image pair sequence; The tamping amount is calculated based on the three-dimensional coordinate change sequence to obtain a tamping amount change curve.

Images