Building engineering progress judgment method and device based on image recognition and storage medium

Abstract

The invention discloses a building engineering progress judgment method and device based on image recognition and a storage medium, and relates to the technical field of building management. The method comprises the following steps: decomposing a target construction plan to obtain discrete construction task units; constructing a dynamic progress baseline sequence evolved along with time based on the discretized construction task unit; obtaining regularly collected multi-modal image data of the construction site, and generating a site three-dimensional live-action model located in the same coordinate system with the dynamic progress baseline sequence based on the multi-modal image data; and performing three-dimensional registration and difference analysis on the field three-dimensional real scene model and the dynamic progress baseline corresponding to the current time node to obtain a quantitative index of the project progress. The method aims at improving the precision of automatically judging the progress of the constructional engineering in a complex environment.

Description

Technical Field

[0001] This application belongs to the field of construction engineering management technology, and in particular relates to a method, equipment and storage medium for determining the progress of construction projects based on image recognition. Background Technology

[0002] In the field of construction project progress management, the traditional reliance on manual inspections for project progress assessment has long been problematic, suffering from inherent deficiencies such as high subjectivity, low efficiency, and data lag. To address these issues, a solution based on automation technology is proposed.

[0003] However, current automation-based solutions mainly include methods based on two-dimensional image recognition and combining it with Building Information Modeling (BIM). For example, convolutional neural networks mounted on drones are used to identify the presence of building components, and BIM models are introduced to compare and analyze images collected on-site with three-dimensional planning models to improve the accuracy of the judgment. However, this method also has technical limitations. On the one hand, it forces a comparison between the dynamically changing construction site and a pre-set, absolutely static ideal model, lacking a precise three-dimensional spatial correspondence; on the other hand, when the construction plan changes, the static BIM model cannot be updated in time, causing the comparison benchmark to become invalid.

[0004] In conclusion, existing automation technologies still suffer from limitations in adapting to complex environments and exhibit low accuracy when determining the progress of construction projects. Summary of the Invention

[0005] In view of this, the embodiments of this application provide a method, device and storage medium for determining the progress of construction projects based on image recognition. By converting the construction plan into a dynamic progress baseline sequence containing three-dimensional geometric information and accurately registering it with the on-site three-dimensional real scene model, the method achieves a leap from qualitative judgment to quantitative analysis of project progress, significantly improving the accuracy and automation level of progress determination.

[0006] This application provides a method for determining the progress of a construction project based on image recognition, including: The target construction plan is decomposed into discretized construction task units; A dynamic progress baseline sequence that evolves over time is constructed based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; Acquire multimodal image data of the construction site collected periodically, and generate a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data; The on-site 3D reality model is registered in 3D with the dynamic progress baseline corresponding to the current time node, and a difference analysis is performed to obtain quantitative indicators of project progress.

[0007] In one embodiment, constructing a dynamic progress baseline sequence that evolves over time based on the discretized construction task units includes: Each of the discretized construction task units is associated with a corresponding target three-dimensional component in the building information model, wherein the target three-dimensional component includes the target three-dimensional geometric information and attribute information; Based on the time nodes and task logic relationships in the construction schedule, the associated target 3D components are combined to generate target 3D models representing the completion status of different planned time nodes. The dynamic progress baseline sequence is composed of the target 3D model.

[0008] In one embodiment, the method further includes: In response to a received construction plan change instruction, the discrete construction task units of the affected time nodes are identified, and the construction plan change instruction carries the change content; Based on the changes, the target 3D components and their target 3D geometric and attribute information at the affected time nodes are regenerated or adjusted to update the dynamic progress baseline sequence.

[0009] In one embodiment, acquiring periodically collected multimodal image data of the construction site and generating a three-dimensional real-world model of the site based on the multimodal image data in the same coordinate system as the dynamic progress baseline sequence includes: Simultaneously acquire visible light and depth images of the construction site; Based on the structure-of-motion reconstruction algorithm and the depth image, the visible light image is reconstructed in three dimensions to generate a three-dimensional point cloud model containing color and geometric information. The three-dimensional point cloud model is then used as the three-dimensional real-world model of the scene.

[0010] In one embodiment, the step of performing 3D registration and difference analysis between the on-site 3D reality model and the dynamic progress baseline corresponding to the current time node includes: The on-site 3D reality model is semantically segmented to identify and separate multiple independent building entity units; Each segmented building entity unit is iteratively registered with the corresponding target 3D component in the current dynamic progress baseline; wherein, the registration process is judged to converge based on a preset registration accuracy threshold. Calculate the volume overlap and surface area difference between each building entity unit and the target three-dimensional component to quantify the completion progress of each building entity unit.

[0011] In one embodiment, the step of obtaining the quantitative indicators of project progress includes: Based on the completion progress of all the building entity units, and combined with the preset engineering quantity weights, a weighted average calculation is performed to obtain the overall project progress percentage. The overall progress percentage is compared with the planned progress percentage to obtain a quantitative indicator of the project progress.

[0012] In one embodiment, the method further includes: Collect error samples between the quantitative indicators of the project progress and the actual progress data; Based on the error samples, the accuracy threshold and / or the algorithm parameters for calculating volume overlap and surface area difference are optimized to improve the accuracy of the judgment.

[0013] A second aspect of this application provides a construction project progress determination device based on image recognition, comprising: The decomposition module is used to decompose the target construction plan into discretized construction task units; A construction module is used to construct a dynamic progress baseline sequence that evolves over time based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; The generation module is used to acquire multimodal image data of the construction site collected periodically, and generate a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data. The analysis module is used to perform three-dimensional registration and difference analysis between the on-site three-dimensional reality model and the dynamic progress baseline corresponding to the current time node to obtain quantitative indicators of project progress.

[0014] In one embodiment, the building module includes: The association unit is used to associate each of the discretized construction task units with the corresponding target three-dimensional component in the building information model, wherein the target three-dimensional component includes the target three-dimensional geometric information and attribute information; The combination unit is used to combine the associated target 3D components according to the time nodes and task logic relationships in the construction schedule to generate target 3D models representing the completion status of different planned time nodes. The constituent unit is used to construct the dynamic progress baseline sequence from the target three-dimensional model.

[0015] In one embodiment, the apparatus further includes: The identification module is used to identify the discrete construction task units of the affected time nodes in response to the received construction plan change instruction, wherein the construction plan change instruction carries the change content; The update module is used to regenerate or adjust the target three-dimensional components and their target three-dimensional geometric and attribute information at the affected time nodes based on the changes, so as to update the dynamic progress baseline sequence.

[0016] In one embodiment, the generation module includes: The acquisition unit is used to simultaneously acquire visible light images and depth images of the construction site; The reconstruction unit is used to perform three-dimensional reconstruction of the visible light image based on the motion recovery structure algorithm and the depth image, and generate a three-dimensional point cloud model containing color and geometric information, and use the three-dimensional point cloud model as the on-site three-dimensional real scene model.

[0017] In one embodiment, the analysis module includes: The segmentation unit is used to perform semantic segmentation on the on-site three-dimensional reality model, and to identify and separate multiple independent building entity units. The registration unit is used to iteratively register each segmented building entity unit with the corresponding target 3D component in the current dynamic progress baseline; wherein, the registration process is judged to converge based on a preset registration accuracy threshold. A calculation unit is used to calculate the volume overlap and surface area difference between each of the building entity units and the target three-dimensional component, so as to quantify the completion progress of each of the building entity units.

[0018] In one embodiment, the analysis module further includes: The unit is used to calculate the overall project progress percentage by performing a weighted average calculation based on the completion progress of all the building entity units and a preset engineering quantity weight. The comparison unit is used to compare the overall progress percentage with the planned progress percentage to obtain a quantitative indicator of the project progress.

[0019] In one embodiment, the apparatus further includes: The collection module is used to collect error samples between the quantitative indicators of the project progress and the actual progress data; An optimization module is used to optimize the accuracy threshold and / or the algorithm parameters for calculating volume overlap and surface area difference based on the error samples, so as to improve the accuracy of the judgment.

[0020] A third aspect of this application provides a construction project progress determination device based on image recognition, characterized in that it includes: a processor, a memory, and a computer program stored in the memory and executable on the processor; when the processor executes the computer program, it implements the steps of the method described in the first aspect above.

[0021] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described in the first aspect above.

[0022] The image recognition-based construction project progress determination method provided in this application includes: constructing a dynamic progress baseline sequence that evolves over time based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and includes the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; acquiring periodically collected multimodal image data of the construction site, and generating a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data; performing three-dimensional registration and difference analysis between the three-dimensional real-scene model of the site and the dynamic progress baseline corresponding to the current time node to obtain a quantitative indicator of the project progress. By transforming the construction plan into a dynamic progress baseline sequence containing three-dimensional geometric information and accurately registering it with the three-dimensional real-scene model of the site, a leap from qualitative judgment to quantitative analysis of project progress is achieved, significantly improving the accuracy and automation level of progress determination. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 A flowchart illustrating a construction project progress determination method based on image recognition provided in an embodiment of this application; Figure 2 A flowchart illustrating a construction project progress determination method based on image recognition, provided as another embodiment of this application; Figure 3 A schematic diagram of a construction project progress determination device based on image recognition provided in an embodiment of this application; Figure 4 This is a schematic diagram of a construction project progress determination device based on image recognition, provided in an embodiment of this application. Detailed Implementation

[0025] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0027] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0028] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0029] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0030] In the description of the embodiments of this application, the term "multiple frames" refers to two or more (including two).

[0031] In the description of the embodiments of this application, the technical terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of this application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0032] This application provides a method for determining the progress of construction projects based on image recognition. Through a multi-level data verification mechanism, it improves the reliability of data transmission in the wireless BMS system, reduces the bit error rate, and ensures the security of system operation.

[0033] Please see Figure 1 As shown, Figure 1 This is a flowchart illustrating a construction project progress determination method based on image recognition, as provided in an embodiment of this application. The construction project progress determination method based on image recognition provided in this embodiment is implemented by a construction project progress determination device based on image recognition, which may include a computer or server, etc.

[0034] Depend on Figure 1 As can be seen, the image recognition-based method for determining the progress of construction projects includes steps S110 to S140. Details are as follows: S110: Decompose the target construction plan to obtain discretized construction task units.

[0035] The target construction plan is the overall construction plan for the project. This step aims to decompose the macro-level overall project construction plan according to deliverables or construction procedures, and convert it into discrete construction task units that can be recognized and processed by computers.

[0036] In a specific implementation, for example, the construction plan of "pouring three-story slabs" can be broken down into a series of more granular, discrete construction task units, such as "erecting three-story slab formwork," "tying three-story slab reinforcement," and "pouring three-story slab concrete." Each discrete construction task unit is defined as an independent data object, and its data structure includes key fields such as its unique identifier, task description, planned start time, planned completion time, and dependencies on preceding tasks. Finally, all discrete construction task units are integrated and stored in a structured task unit list or database, providing a data foundation for subsequent processes.

[0037] S120: Construct a dynamic progress baseline sequence that evolves over time based on discrete construction task units.

[0038] Each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at that planned time node.

[0039] The core of this step is to create a visualized digital planning baseline that is tied to a timeline. The dynamic progress baseline sequence consists of a series of 3D models arranged according to planned time nodes (such as daily, weekly, or key milestone days). Each model (i.e., a baseline) precisely defines the target 3D geometric and attribute information of the set of building entities expected to be completed at the corresponding node.

[0040] For example, constructing a dynamic progress baseline sequence that evolves over time based on discretized construction task units includes: associating each discretized construction task unit with a corresponding target 3D component in a Building Information Model (BIM), whereby the target 3D component includes target 3D geometric information and attribute information; combining the associated target 3D components according to the time nodes and task logic relationships in the construction schedule to generate target 3D models representing the completion status at different planned time nodes; and constructing a dynamic progress baseline sequence from the target 3D models. Understandably, the target 3D model is a progressive state model, capable of accurately representing the intermediate or final completion state of the 3D component during the construction process.

[0041] For example, for the construction task unit "Pouring Column Z1", the target 3D components in its associated Building Information Model (BIM) might include, for instance, a "Completed 3D Component with Reinforcement Skeleton" (a 3D model containing only the reinforcement), a "Completed 3D Component with Formwork Erection" (a 3D model containing the formwork), a "Component with Concrete Poured to 50% Height" (a half-height concrete column model), and a "Component with Completed Concrete Pour" (including a complete concrete column model). This association defines the 3D geometric and attribute information of the task completion level. Each 3D component's corresponding 3D model contains precise target 3D geometric information (such as shape, size, and spatial location) and attribute information (such as component ID, material, and strength grade).

[0042] Based on the time nodes and task logic in the construction schedule, for each planned time node, the latest 3D components associated with all task units planned to be completed before (and including) that time node are aggregated and integrated into a 3D model representing the complete state the site should reach by the end of that node. This step is repeated for all key planned nodes (such as weekly or milestone nodes) to generate a sequence of 3D models arranged in chronological order, i.e., a dynamic schedule baseline sequence. This dynamic schedule baseline sequence is essentially a sequence of four-dimensional BIM models that incorporates the time dimension, clearly presenting the expected spatial state and evolution of the construction progress.

[0043] Furthermore, when generating or updating the dynamic schedule baseline sequence, the aggregated 3D model is automatically checked for spatial conflicts and logical consistency. For example, it checks whether two tasks are scheduled to complete at the same time, but their 3D models interfere with each other spatially (conflict). When a conflict is detected, an alarm is generated and the conflicting components are identified, prompting planners to make adjustments, thereby ensuring the logical correctness of the schedule baseline.

[0044] S130: Acquire multimodal image data of the construction site collected periodically, and generate a 3D real-world model of the site based on the multimodal image data in the same coordinate system as the dynamic progress baseline sequence.

[0045] By acquiring actual site condition data, we can prepare for accurate comparison with planned benchmarks.

[0046] Specifically, multimodal image data of the construction site are collected periodically and reconstructed into a three-dimensional real-world model of the site in the same coordinate system as the dynamic progress baseline sequence, providing a basis for actual status for subsequent progress comparison.

[0047] For example, visible light images and depth images of the construction site can be collected synchronously at a set period (such as daily or weekly) by a drone or mobile device equipped with a visible light camera and a depth camera (such as LiDAR) to obtain multimodal image data.

[0048] For example, acquiring multimodal image data of the construction site collected periodically, and generating a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data, includes: synchronously acquiring visible light images and depth images of the construction site; performing three-dimensional reconstruction of the visible light images based on the structure-reconstruction-motion algorithm and the depth images to generate a three-dimensional point cloud model containing color and geometric information, and using the three-dimensional point cloud model as the three-dimensional real-scene model of the site.

[0049] Specifically, visible light images provide scene texture and color information, while depth images directly provide pixel-level 3D spatial information. Based on the Structure for Motion Restoration (SfM) algorithm, feature points (such as SIFT, ORB, etc.) in a large number of overlapping visible light images are analyzed to estimate camera pose and generate sparse 3D point clouds. The depth images then provide precise geometric constraints and encryption information for this sparse point cloud, effectively compensating for reconstruction errors in weakly textured areas. Ultimately, a high-precision, high-density 3D point cloud model with realistic colors is generated, which serves as a 3D reality model of the scene.

[0050] To ensure spatial consistency with the planned model, control points with known absolute coordinates are established at the construction site. During the 3D reconstruction process, these control points serve as strong spatial constraints, ensuring that the generated 3D reality model of the site and the dynamic progress baseline sequence are in the same world coordinate system.

[0051] S140: Perform 3D registration and difference analysis between the on-site 3D reality model and the dynamic progress baseline corresponding to the current time node to obtain quantitative indicators of project progress.

[0052] In this step, high-precision 3D comparison transforms the discrepancies between the actual situation and the plan into objective data. Specifically, the 3D reality model of the site is registered and the dynamic progress baseline corresponding to the current time point is analyzed using 3D registration. This includes: semantic segmentation of the 3D reality model of the site to identify and separate multiple independent building entity units; iterative nearest-point registration of each segmented building entity unit with the corresponding target 3D component in the current dynamic progress baseline; the registration process is judged to converge based on a preset registration accuracy threshold; and the volume overlap and surface area difference between each building entity unit and the target 3D component are calculated to quantify the completion progress of each building entity unit.

[0053] For example, a pre-trained deep learning model is used to perform semantic segmentation on the on-site 3D reality model (point cloud model) to identify and separate multiple independent building entity units (such as columns, beams, walls, slabs, etc.). For each segmented building entity unit, its corresponding target 3D component is found in the dynamic progress baseline. Specifically, instead of performing fine registration directly on each building entity unit, an initial rigid body transformation matrix is ​​first calculated using the center and principal axis direction of the axial bounding box of the segmentation result to coarsely register the reality point cloud to the vicinity of the target BIM component. This step greatly improves the dependence of the Iterative Closest Point (ICP) algorithm on the initial position and effectively avoids local optima.

[0054] Based on the coarse registration described above, the Iterative Closest Point (ICP) algorithm is used for fine registration, ensuring that the actual building units and planned components are precisely aligned in space. The registration process determines whether convergence has been achieved based on a preset accuracy threshold (e.g., 1 mm).

[0055] The volume overlap ratio between each building entity unit and its corresponding target 3D component is calculated, i.e., the percentage of intersection between the actual point cloud volume and the BIM model volume, serving as the primary basis for the component's completion progress. To improve accuracy, a "visibility-based confidence weight" is introduced during the calculation. Specifically, areas with point cloud density significantly lower than the average (potentially due to occlusion) are assigned a lower weight when calculating overlap ratio to reduce errors caused by missing data. The completion status is determined based on the volume overlap ratio and the component's geometric characteristics. For example, for vertical components (columns, walls), if the ratio of point cloud height to target model height is 70%, then even with a low overall volume overlap ratio, its progress can be determined as "cast to 70% height." The calculation formula is: Construction completion progress = Volume overlap ratio × 100%. This represents a leap from a binary determination of "whether it is completed" to a continuous quantification of "how much is completed."

[0056] Simultaneously, surface area differences are calculated to identify component dimensional deviations and surface construction quality.

[0057] After obtaining the completion progress of all components, quantitative indicators are calculated. Specifically, the steps for calculating quantitative indicators include: based on the completion progress of all building entity units, combined with the engineering quantity weights obtained from the automatic quantity calculation from the BIM model, a weighted average calculation is performed to obtain the overall project progress percentage; the overall progress percentage is compared with the planned progress percentage to obtain the quantitative indicators of project progress.

[0058] Based on the automatic quantity calculation results from the BIM model, a weight is preset for each component in the total project quantity (reflecting its proportion in the total project quantity). The overall project progress percentage is calculated by weighted average, with the formula: Overall Progress Percentage = Σ(Completion progress of the i-th component × Weight of the i-th component). From the dynamic progress baseline sequence, the planned progress percentage corresponding to the planned time node on the data collection date is obtained. The calculated overall progress percentage is compared with the planned progress percentage to obtain a quantitative indicator of project progress, namely: Progress Difference Value = Overall Progress Percentage - Planned Progress Percentage. A positive value indicates that the progress is ahead of schedule, and a negative value indicates that the progress is behind schedule.

[0059] In one embodiment, to improve the accuracy of long-term operation, the method further includes a step of optimizing the loop. Specifically, during the execution of the above steps, the method further includes: collecting error samples between quantitative indicators of project progress and actual progress data; and optimizing the accuracy threshold and / or the algorithm parameters for calculating volume overlap and surface area difference based on the error samples.

[0060] Specifically, error samples are collected between the determined progress results and the actual progress data verified manually, and an error sample library is established. Based on this error sample library, the parameters in the registration accuracy threshold, volume overlap calculation algorithm, and semantic segmentation model in S140 are fed back and optimized, thereby continuously improving the accuracy and adaptability of progress determination.

[0061] As can be seen from the above analysis, the image recognition-based construction project progress determination method provided in this application includes: constructing a dynamic progress baseline sequence that evolves over time based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and includes the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; acquiring periodically collected multimodal image data of the construction site, and generating a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data; performing three-dimensional registration and difference analysis between the three-dimensional real-scene model of the site and the dynamic progress baseline corresponding to the current time node to obtain a quantitative index of the project progress. By transforming the construction plan into a dynamic progress baseline sequence containing three-dimensional geometric information and accurately registering it with the three-dimensional real-scene model of the site, a leap from qualitative judgment to quantitative analysis of project progress is achieved, significantly improving the accuracy and automation level of progress determination.

[0062] Please see Figure 2 , Figure 2 This is a flowchart illustrating a construction project progress determination method based on image recognition, provided as another embodiment of this application. This embodiment is related to... Figure 1 Compared to the illustrated embodiment, the implementation processes of S210 to S220 are the same as those of S110 to S120, and the specific implementation processes of S250 to S260 are the same as those of S130 to S140. The difference lies in the inclusion of S230 to S240 after S220. Specifically, S230 and S250 can be executed in parallel, or one of them can be executed. Details are as follows: S210: Decompose the target construction plan to obtain discretized construction task units.

[0063] S220: Construct a dynamic schedule baseline sequence that evolves over time based on discretized construction task units; wherein each baseline in the dynamic schedule baseline sequence corresponds to a planned time node and contains the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node.

[0064] S230: In response to a received construction plan change instruction, identify the discretized construction task unit of the affected time node, and the construction plan change instruction carries the change content.

[0065] To provide an accurate and efficient three-dimensional schedule baseline, it is further necessary to receive external construction plan change instructions, analyze the changes carried by the construction plan change instructions, and identify all tasks and time nodes affected by the changes.

[0066] Specifically, the system receives construction plan change instructions from integrated project management software via a pre-defined standard data interface (such as an API) and parses the structured change content carried by these instructions. For example, the structured change content includes, but is not limited to: task identifiers indicating one or more discrete construction task units requiring change; change types including schedule delays or advances; changes in logical relationships such as task A must start after task B is completed; task additions or deletions; and change parameters including specific changes such as a 2-day delay or a new start date. Based on the project plan network diagram stored within it, and the logical relationships between task components defined along the diagram (such as finish-start, start-start, and other dependencies), automated impact analysis is performed. Specifically, based on the parsed task representations, this process locates one or more task units directly affected by change in the plan network diagram. Starting from the source task, it recalculates the earliest start and earliest finish times of all subsequent tasks based on their logical relationships with subsequent tasks. Any planned time node that changes is identified as an affected time node. After the analysis is complete, a structured list of affected tasks is generated. The list clearly identifies all affected time points and their new planned times, as well as the set of discretized construction task units associated with each affected time point whose planned status (such as start, finish) or time has changed.

[0067] S240: Based on the changes, regenerate or adjust the target 3D components and their target 3D geometric and attribute information for the affected time nodes to update the dynamic progress baseline sequence.

[0068] Based on the obtained list of affected entities, for each time point requiring adjustment, the corresponding target 3D component is regenerated or adjusted. The 3D geometry of the component (such as construction status and completion percentage) and attribute information (such as planned date and task status) will be automatically updated according to the new plan data.

[0069] All updated 3D components will be integrated back into the timeline, reconstructing a new dynamic schedule baseline sequence. This sequence fully represents the latest approved construction plan. After the update, operations will be conducted based on this new dynamic schedule baseline sequence to ensure that the schedule determination benchmark remains synchronized with the latest construction plan.

[0070] S250: Acquire multimodal image data of the construction site collected periodically, and generate a 3D real-world model of the site based on the multimodal image data in the same coordinate system as the dynamic progress baseline sequence.

[0071] S260: Perform 3D registration and difference analysis between the on-site 3D reality model and the dynamic progress baseline corresponding to the current time node to obtain quantitative indicators of project progress.

[0072] As can be seen from the above analysis, the image recognition-based construction project progress determination method provided in this application includes: constructing a dynamic progress baseline sequence that evolves over time based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and includes the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; acquiring periodically collected multimodal image data of the construction site, and generating a three-dimensional real-scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data; performing three-dimensional registration and difference analysis between the three-dimensional real-scene model of the site and the dynamic progress baseline corresponding to the current time node to obtain a quantitative index of the project progress. By transforming the construction plan into a dynamic progress baseline sequence containing three-dimensional geometric information and accurately registering it with the three-dimensional real-scene model of the site, a leap from qualitative judgment to quantitative analysis of project progress is achieved, significantly improving the accuracy and automation level of progress determination.

[0073] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0074] Please see Figure 3 , Figure 3 This is a schematic diagram of a construction project progress determination device based on image recognition, provided in an embodiment of this application. The image recognition-based construction project progress determination device includes modules or units for performing... Figure 1 or Figure 2 The steps in the corresponding embodiments. Please refer to the details. Figure 1 or Figure 2 The relevant descriptions in the corresponding embodiments are shown below. For ease of explanation, only the parts relevant to this embodiment are shown. See also... Figure 3 A construction project progress determination device 300 based on image recognition includes: The decomposition module 310 is used to decompose the target construction plan into discretized construction task units; The construction module 320 is used to construct a dynamic progress baseline sequence that evolves over time based on the discretized construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains the target three-dimensional geometric information and attribute information of the set of building entities expected to be completed at the planned time node; The generation module 330 is used to acquire multimodal image data of the construction site collected periodically, and generate a three-dimensional real scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multimodal image data. The analysis module 340 is used to perform three-dimensional registration and difference analysis between the on-site three-dimensional real scene model and the dynamic progress baseline corresponding to the current time node to obtain quantitative indicators of project progress.

[0075] In one embodiment, the construction module 320 includes: The association unit is used to associate each of the discretized construction task units with the corresponding target three-dimensional component in the building information model, wherein the target three-dimensional component includes the target three-dimensional geometric information and attribute information; The combination unit is used to combine the associated target 3D components according to the time nodes and task logic relationships in the construction schedule to generate target 3D models representing the completion status of different planned time nodes. The constituent unit is used to construct the dynamic progress baseline sequence from the target three-dimensional model.

[0076] In one embodiment, the device 300 further includes: The identification module is used to identify the discrete construction task units of the affected time nodes in response to the received construction plan change instruction, wherein the construction plan change instruction carries the change content; The update module is used to regenerate or adjust the target three-dimensional components and their target three-dimensional geometric and attribute information at the affected time nodes based on the changes, so as to update the dynamic progress baseline sequence.

[0077] In one embodiment, the generation module 330 includes: The acquisition unit is used to simultaneously acquire visible light images and depth images of the construction site; The reconstruction unit is used to perform three-dimensional reconstruction of the visible light image based on the motion recovery structure algorithm and the depth image, and generate a three-dimensional point cloud model containing color and geometric information, and use the three-dimensional point cloud model as the on-site three-dimensional real scene model.

[0078] In one embodiment, the analysis module 340 includes: The segmentation unit is used to perform semantic segmentation on the on-site three-dimensional reality model, and to identify and separate multiple independent building entity units. The registration unit is used to iteratively register each segmented building entity unit with the corresponding target 3D component in the current dynamic progress baseline; wherein, the registration process is judged to converge based on a preset registration accuracy threshold. A calculation unit is used to calculate the volume overlap and surface area difference between each of the building entity units and the target three-dimensional component, so as to quantify the completion progress of each of the building entity units.

[0079] In one embodiment, the analysis module 340 further includes: The unit is used to calculate the overall project progress percentage by performing a weighted average calculation based on the completion progress of all the building entity units and a preset engineering quantity weight. The comparison unit is used to compare the overall progress percentage with the planned progress percentage to obtain a quantitative indicator of the project progress.

[0080] In one embodiment, the device 300 further includes: The collection module is used to collect error samples between the quantitative indicators of the project progress and the actual progress data; An optimization module is used to optimize the accuracy threshold and / or the algorithm parameters for calculating volume overlap and surface area difference based on the error samples, so as to improve the accuracy of the judgment.

[0081] Please see Figure 4 , Figure 4 This is a schematic diagram of a construction project progress determination device based on image recognition, provided in an embodiment of this application. Figure 4 It is understood that the image recognition-based construction project progress determination device 400 includes: a processor 410, a memory 420, and a computer program 430 stored in the memory 420 and executable on the processor 410; when the processor 410 executes the computer program 430, it implements the steps in the above-described embodiments of the image recognition-based construction project progress determination method, for example... Figure 1 The steps S110 to S140 are shown. Alternatively, when the processor 410 executes the computer program 430, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 3 The functions of modules 310 to 340 are shown.

[0082] For example, computer program 430 may be divided into one or more modules / units, one or more of which are stored in memory 420 and executed by processor 410 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of computer program 430 in a wireless communication-based energy storage system management device. For example, computer program 430 may be divided into a decomposition module, a construction module, a generation module, and an analysis module.

[0083] The image recognition-based construction project progress determination device provided in this embodiment may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that... Figure 4 This is merely an example of a construction project progress determination device based on image recognition, and does not constitute a limitation on such devices. It may include more or fewer components than shown in the illustration, or combine certain components, or different components. For example, a construction project progress determination device based on image recognition may also include input / output devices, network access devices, buses, etc.

[0084] The processor 410 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0085] The memory 420 can be an internal storage unit of the image recognition-based construction project progress determination device, such as a hard drive or memory. The memory 420 can also be an external storage device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, or flash card. Furthermore, the image recognition-based construction project progress determination device can include both internal and external storage units. The memory 420 is used to store computer programs and other programs and data required by the image recognition-based construction project progress determination device. The memory 420 can also be used to temporarily store data that has been output or will be output.

[0086] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0087] This application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the steps in any of the above method embodiments.

[0088] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0089] This application provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps described in the above-described method embodiments.

[0090] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0091] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0092] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0093] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0094] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0095] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An image recognition-based construction progress determination method, characterized by, The method comprises the following steps: decompose the target construction plan to obtain discrete construction task units; construct a dynamic progress baseline sequence evolving over time based on the discrete construction task units; each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains target three-dimensional geometric information and attribute information of a set of building entities expected to be completed at the planned time node; obtain multi-modal image data of the construction site collected periodically, and generate a three-dimensional real scene model of the construction site in the same coordinate system as the dynamic progress baseline sequence based on the multi-modal image data; perform three-dimensional registration and difference analysis between the three-dimensional real scene model and the dynamic progress baseline corresponding to the current time node to obtain a quantitative indicator of the engineering progress.

2. The image recognition-based construction progress determination method according to claim 1, wherein, The method further comprises the following steps: associate each discrete construction task unit with a corresponding target three-dimensional component in the building information model, wherein the target three-dimensional component comprises the target three-dimensional geometric information and attribute information; combine the associated target three-dimensional components according to the time node and task logic relationship in the construction progress plan to generate a target three-dimensional model representing the completion state of different planned time nodes; construct the dynamic progress baseline sequence from the target three-dimensional model.

3. The image recognition-based construction progress determination method according to claim 2, wherein, The method further comprises the following steps: in response to a received construction plan change instruction, identify the discrete construction task units of the affected time nodes, wherein the construction plan change instruction carries the change content; based on the change content, regenerate or adjust the target three-dimensional components and their target three-dimensional geometric information and attribute information at the affected time nodes to update the dynamic progress baseline sequence.

4. The image recognition-based construction progress determination method according to claim 1, wherein, The method further comprises the following steps: synchronously obtain visible light images and depth images of the construction site; based on the motion recovery structure algorithm and the depth images, perform three-dimensional reconstruction on the visible light images to generate a three-dimensional point cloud model containing color and geometric information, and use the three-dimensional point cloud model as the three-dimensional real scene model of the construction site.

5. The image recognition-based construction progress determination method according to claim 4, wherein, The method further comprises the following steps: perform semantic segmentation on the three-dimensional real scene model to identify and separate multiple independent building entity units; perform iterative closest point registration between each segmented building entity unit and the corresponding target three-dimensional component in the current dynamic progress baseline; wherein the registration process is judged whether to converge according to a preset registration accuracy threshold; calculate the volume overlap and surface area difference between each building entity unit and the target three-dimensional component to quantify the completion progress of each building entity unit.

6. The image recognition-based construction progress determination method according to claim 5, wherein The method further comprises the following steps: based on the completion progress of all the building entity units, combine the preset engineering quantity weight to perform weighted average calculation and obtain the overall progress percentage of the project; The whole progress percentage is compared with the planned progress percentage to obtain the quantitative index of the engineering progress.

7. The image recognition-based construction progress determination method according to claim 6, wherein, The method further comprises: collecting error samples between the quantitative index of the engineering progress and actual progress data; based on the error samples, optimizing the precision threshold and / or algorithm parameters for calculating the volume coincidence degree and surface area difference to improve the accuracy of the determination.

8. A construction project progress determination device based on image recognition, characterized in that, comprise: a decomposition module configured to decompose a target construction plan to obtain discrete construction task units; a construction module configured to construct a dynamic progress baseline sequence evolving over time based on the discrete construction task units; wherein each baseline in the dynamic progress baseline sequence corresponds to a planned time node and contains target three-dimensional geometric information and attribute information of a set of building entities expected to be completed at the planned time node; a generation module configured to obtain multi-modal image data of a construction site collected at regular intervals and generate a three-dimensional real scene model of the site in the same coordinate system as the dynamic progress baseline sequence based on the multi-modal image data; an analysis module configured to perform three-dimensional registration and difference analysis between the three-dimensional real scene model and a dynamic progress baseline corresponding to a current time node to obtain a quantitative index of the engineering progress.

9. A construction project progress determination device based on image recognition, characterized in that, comprise: a processor, a memory, and a computer program stored in the memory and executable on the processor; the processor implements the steps of the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. the computer program implements the steps of the method according to any one of claims 1 to 7 when executed by the processor.

Images