A method, apparatus, equipment, and medium for calculating engineering quantities

Abstract

This invention provides a method, apparatus, device, and medium for calculating engineering quantities. The method includes: acquiring video data corresponding to the construction process in a construction area; determining the operating space corresponding to each piece of equipment based on the video data; the operating space including the equipment boundary and the operating scene of the equipment; cropping the video data to obtain a target video corresponding to the operating space; determining the operating trajectory of the equipment in the target video; determining the number of times the equipment performs a complete operation based on the operating trajectory; and determining the engineering quantity based on the number of times the equipment performs a complete operation. This invention, by determining the operating space including equipment boundaries and the operating scene, and cropping the video data to generate target videos corresponding to each piece of equipment, accurately extracts the operating trajectory and counts the number of operations. This eliminates the dependence on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of engineering quantity calculation.

Description

Technical Field

[0001] This invention relates to the field of engineering quantity calculation technology, and in particular to an engineering quantity calculation method and an engineering quantity calculation device. Background Technology

[0002] In existing technologies, determining the quantity of work requires classifying temporal postures based on predefined action categories to count the number of work cycles. This makes the action recognition module reliant on a manually defined and fixed category system, which is difficult to adapt to complex and ever-changing actual working conditions and has weak generalization ability. In addition, small errors in the detection or tracking stages can be propagated and amplified along the pipeline, leading to inaccurate final quantity statistics and poor overall system robustness. Summary of the Invention

[0003] In view of the above problems, embodiments of the present invention are proposed to provide a method, apparatus, device and medium for calculating engineering quantities that overcomes or at least partially solves the above problems.

[0004] According to a first aspect of the present invention, a method for calculating engineering quantities is provided, the method comprising: Acquire video data corresponding to the construction process in the construction area; Based on the video data, the operating space corresponding to each piece of work equipment in the construction area is determined; the operating space includes the equipment boundary of the work equipment and the work scene of the work equipment; the work scene includes a local construction area associated with the construction task currently being performed by the work equipment; The video data is cropped to obtain the target video corresponding to the operation space; Determine the operating trajectory of the operating equipment in the target video; Based on the work trajectory, determine the number of times the work equipment performs a complete work process; The workload is determined based on the number of times the equipment in the construction area completes the entire work process.

[0005] Optionally, acquiring video data corresponding to the construction process in the construction area includes: Acquire video data of the construction area captured by multiple acquisition devices; The video data of the construction area captured by the multiple acquisition devices is converted from the pixel coordinate system to the world coordinate system; Multiple video data transformed to the world coordinate system are fused to obtain the first fused video data in the world coordinate system. The fused video data in the world coordinate system is transformed to the pixel coordinate system to obtain the second fused video data corresponding to the construction process in the construction area.

[0006] Optionally, determining the operating space corresponding to each piece of work equipment in the construction area based on the video data includes: Identify each operating device in the video data and assign a device identifier to each operating device; The operation space corresponding to each piece of equipment in the construction area is determined by tracking each piece of equipment based on the equipment identifier.

[0007] Optionally, the step of tracking each working device based on the device identifier and determining the operating space corresponding to each working device in the construction area includes: During the tracking of each working device based on the device identifier, the device boundary of each working device in the video data and the working scene of the working device are determined; Based on the equipment boundaries and the operating scenarios of the operating equipment, the operating space corresponding to each operating device in the construction area is determined.

[0008] Optionally, determining the operating space corresponding to each piece of equipment in the construction area based on the equipment boundaries and the operating scenario of the equipment includes: Determine the equipment boundary compensation value corresponding to the operating scenario of the operating equipment; The equipment boundaries corresponding to each working device are expanded outwards according to the equipment boundary compensation value to determine the operating space corresponding to each working device in the construction area.

[0009] Optionally, determining the number of times the work equipment in the construction area performs a complete work process based on the work trajectory includes: The evaluation rules for a complete operation process of the equipment are obtained; the evaluation rules include at least one of action sequence, spatial movement trajectory and operation space context constraint; the action sequence is a number of time-sequential sub-actions required for a complete operation process; the movement trajectory is a continuous path from the starting position through intermediate work points to the ending position required for a complete operation process; the operation space context constraint is the existence state and position state required for the operation object in a complete operation process. The model determines the number of complete operation processes in the operation trajectory based on the operation trajectory in the target video by using the number of iterations, and then determines the number of complete operation processes that meet the evaluation rules from all the operation processes.

[0010] Optionally, the method further includes: Obtain the historical complete operation process count for the corresponding operating equipment; The model is determined by optimizing the number of iterations based on the number of complete historical workflows.

[0011] Optionally, determining the workload based on the number of times the work equipment in the construction area performs a complete work process includes: Obtain the rated parameters, historical average operating efficiency, and total operating time for each of the operating devices; Determine the product of the number of complete operation cycles, rated parameters, and historical average operating efficiency for each operating device; The target workload is determined based on the ratio of the product to the total operation time.

[0012] Optionally, before acquiring the video data corresponding to the construction process in the construction area, the method further includes: Receive the user's instruction to retrieve the amount of work completed; Based on the quantity acquisition instruction, the construction area for which the quantity of work to be calculated is determined, as well as the construction process of the construction area.

[0013] According to a second aspect of the present invention, an engineering quantity calculation apparatus is provided, the apparatus comprising: The video data acquisition module is used to acquire video data corresponding to the construction process in the construction area; The operation space determination module is used to determine the operation space corresponding to each working device in the construction area based on the video data; the operation space includes the device boundary of the working device and the working scene of the working device; the working scene includes a local construction area associated with the construction task currently being performed by the working device; The target video acquisition module is used to crop the video data to obtain the target video corresponding to the operation space; The operation trajectory determination module is used to determine the operation trajectory of the operation equipment in the target video; The number of times determination module is used to determine the number of times the work equipment performs the complete work process based on the work trajectory; The quantity determination module is used to determine the quantity of work based on the number of times the working equipment in the construction area performs a complete work process.

[0014] Optionally, the video data acquisition module includes: The video data acquisition submodule is used to acquire video data of the construction area captured by multiple acquisition devices; The coordinate transformation submodule is used to transform the video data of the construction area captured by the multiple acquisition devices from the pixel coordinate system to the world coordinate system; The first fused video data determination submodule is used to fuse multiple video data transformed to the world coordinate system to obtain the first fused video data in the world coordinate system. The second fused video data determination submodule is used to transform the fused video data in the world coordinate system to the pixel coordinate system to obtain the second fused video data corresponding to the construction process in the construction area.

[0015] Optionally, the operating space determination module includes: The device identifier allocation submodule is used to identify each working device in the video data and assign a device identifier to each working device. The operating space determination submodule is used to track each working device based on the device identifier and determine the operating space corresponding to each working device in the construction area.

[0016] Optionally, the operation space determination submodule includes: The work scene determination unit is used to determine the device boundary of each work device and the work scene of the work device in the video data during the process of tracking each work device according to the device identifier; The operating space determination unit is used to determine the operating space corresponding to each piece of equipment in the construction area based on the equipment boundary and the operating scenario of the operating equipment.

[0017] Optionally, the operating space determination unit includes: The compensation value determination subunit is used to determine the equipment boundary compensation value corresponding to the operating scenario of the operating equipment. The operating space determination subunit is used to expand the equipment boundary corresponding to each working device in all directions according to the equipment boundary compensation value, thereby determining the operating space corresponding to each working device in the construction area.

[0018] Optionally, the number determination module includes: The evaluation rule acquisition submodule is used to acquire the evaluation rules for a complete operation process of the working equipment; the evaluation rules include at least one of action sequence, spatial movement trajectory, and operation space context constraint; the action sequence is a series of time-sequential sub-actions required for a complete operation process; the movement trajectory is a continuous path from the starting position through intermediate work points to the ending position required for a complete operation process; the operation space context constraint is the existence state and position state required for the work object in a complete operation process. The number of iterations determination submodule is used to determine all the operation processes in the operation trajectory in the target video based on the operation trajectory through the loop number determination model, and to determine the number of complete operation processes that meet the evaluation rules from all the operation processes.

[0019] Optionally, the device further includes: The historical data acquisition unit is used to acquire the number of complete historical operation processes corresponding to the operating equipment; The model optimization unit is used to determine the model by optimizing the number of cycles based on the number of times the historical complete job process has been completed.

[0020] Optionally, the quantity determination module includes: The parameter acquisition submodule is used to acquire the rated parameters, historical average operating efficiency, and total operating time of each operating device. The quantity determination submodule is used to determine the product of the number of complete operation processes, rated parameters, and historical average operating efficiency for each operating device. The target workload is determined based on the ratio of the product to the total operation time.

[0021] Optionally, the device further includes: The instruction acquisition module is used to receive the user's instructions for acquiring project quantities; The construction process determination module is used to determine the construction area for which the quantity of work to be calculated, and the construction process of the construction area, based on the quantity of work acquisition instruction.

[0022] According to a third aspect of the present invention, an electronic device is provided, the electronic device comprising: a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the steps of the engineering quantity calculation method as described in any of the preceding claims.

[0023] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the engineering quantity calculation method as described in any of the preceding claims.

[0024] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention provides a method, apparatus, device, and medium for calculating engineering quantities. It involves acquiring video data corresponding to the construction process in a construction area; determining the operating space for each piece of equipment based on the video data; the operating space including the equipment boundary and the operating scene of the equipment; cropping the video data to obtain a target video corresponding to the operating space; determining the operating trajectory of the equipment in the target video; determining the number of times the equipment performs a complete operation based on the operating trajectory; and determining the engineering quantity based on the number of times the equipment performs a complete operation. This invention, by determining the operating space including equipment boundaries and the operating scene, and cropping the video data to generate target videos corresponding to each equipment, accurately extracts the operating trajectory and counts the number of operations. This eliminates reliance on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of engineering quantity calculation. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the steps of a method for calculating engineering quantities provided in an embodiment of the present invention; Figure 2 This is a flowchart of another method for calculating engineering quantities provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a scenario for a method for calculating engineering quantities provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of an engineering quantity calculation device provided in an embodiment of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] One of the core concepts of this invention is that by determining the operating space including the equipment boundary and the work scene, video data is cropped to generate target videos corresponding to each equipment, so as to accurately extract the work trajectory and count the number of operations. This eliminates the dependence on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of engineering quantity calculation.

[0028] Reference Figure 1 The diagram illustrates a flowchart of a method for calculating engineering quantities according to an embodiment of the present invention. The method may specifically include the following steps: Step 101: Obtain video data corresponding to the construction process in the construction area; For example, data captured by cameras at different angles, covering all images of the construction process within the construction area, can serve as the most intuitive, continuous, and spatiotemporally rich carrier of construction activities, comprehensively recording the position, posture, movement trajectory, and operational behavior of the equipment. Analyzing this video data can avoid the problems of subjectivity, low efficiency, and easy omissions inherent in traditional manual inspections or log entries.

[0029] Step 102: Based on the video data, determine the operating space corresponding to each piece of work equipment in the construction area; the operating space includes the equipment boundary of the work equipment and the work scene of the work equipment; the work scene includes a local construction area associated with the construction task currently being performed by the work equipment; For example, a work scenario represents the context of equipment operation, such as functional areas like excavation areas, unloading points, support faces, or transportation channels. Their spatial location and semantic attributes together form important criteria for judging the effectiveness of equipment operations. Equipment tracking is a key link connecting static identification and dynamic analysis. Simply identifying equipment without tracking it fails to obtain its continuous movement trajectory, making it impossible to determine whether a complete work process has been completed. Cross-frame association using equipment identifiers allows for the reconstruction of the movement path of each piece of equipment in space and time. Based on this, its "operating space"—the functional spatial area jointly formed by the equipment itself (equipment boundary) and its surrounding work environment (work scenario)—is further determined. This operating space not only reflects the physical existence of the equipment but also its operational impact range. For example, the operating space of an excavator includes not only its vehicle outline but also the earthwork area within its excavation radius. Clearly defining this space helps to accurately define the "effective working area," avoiding misjudging equipment passing by or in standby mode as operational behavior, thereby improving the accuracy and semantic rationality of quantity calculations.

[0030] Step 103: Crop the video data to obtain the target video corresponding to the operation space; For example, raw video data typically covers the entire construction area and contains a large amount of background information unrelated to the operation of specific equipment (such as other equipment, personnel, fixed structures, etc.). Analyzing the behavior of a single piece of equipment directly on the entire map is not only computationally expensive but also susceptible to interference leading to false positives or false negatives. By precisely cropping the video data according to the operating space of each piece of equipment, a "target video" containing only the equipment and its key operating area can be generated, achieving data focusing and noise reduction. This cropping strategy effectively isolates visual interference between multiple devices, improving the accuracy and stability of subsequent target detection, attitude estimation, and trajectory extraction. At the same time, the cropped video has a relatively higher resolution and more compact content, significantly reducing the processing burden on the algorithm, which is beneficial for engineering calculation scenarios with high real-time requirements. Furthermore, using the operating space as the basis for cropping ensures that the target video completely covers the equipment itself and the area affected by the operation, avoiding the loss of key action information due to overly narrow cropping, which is a prerequisite for achieving high-precision and high-efficiency operation behavior analysis.

[0031] Step 104: Determine the operating trajectory of the operating equipment in the target video; For example, the work trajectory is a core temporal feature reflecting the dynamic behavior of equipment, recording the equipment's position, speed, and movement path within the operating space over time. By extracting this trajectory from the target video, the actual working process of the equipment can be objectively and quantitatively reconstructed, providing direct evidence for determining whether it has performed effective work. For instance, the typical work trajectory of an excavator presents a periodic closed loop of "digging-lifting-reversing-returning," while a transport vehicle exhibits a linear path traveling back and forth between two points. Only by acquiring continuous and complete trajectories can these patterns be identified and the effective work distinguished from non-operational states such as idling or standby. Furthermore, trajectory data can also be used to analyze work efficiency, path optimization, and abnormal behavior detection. Therefore, determining the work trajectory is a crucial step from visual data to semantic understanding of engineering quantities, and an indispensable intermediate link in achieving automated and seamless engineering measurement.

[0032] Step 105: Based on the work trajectory, determine the number of times the work equipment performs the complete work process; For example, the amount of work (such as earthwork excavation, concrete pouring, and number of muck transport trips) is usually positively correlated with the number of times the equipment completes a "complete work process". For instance, each time an excavator completes a "dig-lift-dump-return" cycle, it represents the transfer of a certain volume of earth; each time a muck truck completes a "load-transport-unload-return" cycle, it is counted as one effective transport shift. Therefore, by analyzing the spatiotemporal characteristics of the work trajectory (such as trajectory shape, stopping points, directional changes, periodicity, etc.), the number of such complete work cycles can be automatically identified and counted. This process transforms continuous visual motion data into discrete, measurable engineering events, realizing the leap from "seeing the action" to "calculating the work quantity". Compared with traditional methods that rely on manual recording or predefined action classification, trajectory-based cycle counting is more objective, adaptable, and robust, and can effectively cope with complex and changing on-site conditions. This is the core logic of this invention for achieving high-precision automated work quantity calculation.

[0033] Step 106: Determine the workload based on the number of times the work equipment in the construction area completes the entire work process.

[0034] For example, the volume of work (such as earthwork excavation, concrete pouring, and waste soil transportation) is essentially directly related to the number of effective work cycles completed by the equipment. For instance, each complete work cycle completed by an excavator typically corresponds to the transfer of a fixed volume of earth; each complete work cycle completed by a dump truck represents a standard transport shift. Therefore, using the "number of complete work cycles" as the core input for work volume calculation can transform abstract visual behavior into quantifiable and cumulative production indicators. Compared to traditional methods that rely on manual reporting or sensor-accumulated working hours, this approach is more objective and accurate, and can effectively distinguish between effective work and non-productive states such as idling or standby. By using the actual completed work cycles as the unit of measurement, not only is the accuracy of work volume statistics improved, but a highly reliable data foundation is also provided for progress monitoring, cost accounting, and resource scheduling, making it a key link in achieving intelligent and automated construction management.

[0035] This invention provides a method for calculating engineering quantities. The method involves acquiring video data corresponding to the construction process in a construction area; determining the operating space for each piece of equipment based on the video data; the operating space including the equipment boundary and the operating scene of the equipment; cropping the video data to obtain a target video corresponding to the operating space; determining the operating trajectory of the equipment in the target video; determining the number of times the equipment performs a complete operation based on the operating trajectory; and determining the engineering quantity based on the number of times the equipment performs a complete operation. This invention, by determining the operating space including equipment boundaries and the operating scene, and cropping the video data to generate target videos corresponding to each piece of equipment, accurately extracts the operating trajectory and counts the number of operations. This eliminates the reliance on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of engineering quantity calculation.

[0036] Reference Figure 2 The diagram illustrates a flowchart of a method for calculating engineering quantities according to an embodiment of the present invention. The method may specifically include the following steps: Step 201: Obtain video data corresponding to the construction process in the construction area; For example, data captured by cameras at different angles, covering all images of the construction process within the construction area, can serve as the most intuitive, continuous, and spatiotemporally rich carrier of construction activities, comprehensively recording the position, posture, movement trajectory, and operational behavior of the equipment. Analyzing this video data can avoid the problems of subjectivity, low efficiency, and easy omissions inherent in traditional manual inspections or log entries.

[0037] In one embodiment, step 201 includes the following sub-steps: Sub-step S11: Acquire video data of the construction area captured by multiple acquisition devices; For example, a single camera's limited field of view is susceptible to occlusion, lighting changes, or blind spots, making it difficult to comprehensively cover all operational equipment and their operational details in large or complex construction scenarios. By deploying multiple acquisition devices (such as fixed surveillance cameras, drones, and mobile terminals) to simultaneously acquire video from different angles, heights, and positions, the limitations of a single viewpoint can be effectively overcome, enabling multi-dimensional stereoscopic observation of the construction area. Multi-view data not only improves the completeness and accuracy of equipment detection and tracking but also reduces trajectory interruptions caused by occlusion through complementary perspectives. Furthermore, multi-source video provides redundant information for subsequent spatial fusion and 3D reconstruction, enhancing the system's perception robustness in complex dynamic environments and ensuring that critical operational behaviors are not missed, thus laying a solid data foundation for high-precision engineering quantity calculations.

[0038] Sub-step S12: Convert the video data of the construction area captured by the multiple acquisition devices from the pixel coordinate system to the world coordinate system; For example, each acquisition device images independently, and its video data is based on its own pixel coordinate system, lacking a unified spatial reference benchmark, making direct geometric alignment or content fusion impossible. However, engineering quantity calculations require device positioning, trajectory analysis, and operational space modeling in real physical space, necessitating a unified measurement system. Therefore, by using camera calibration and spatial mapping techniques to transform pixels in each video stream to a common world coordinate system (such as an East-North-Elevation coordinate system with a point on the construction site as the origin), precise alignment of multi-view observation results in real three-dimensional space can be achieved. This transformation makes the device positions, boundaries, and motion trajectories from different cameras comparable and consistent, providing a necessary prerequisite for subsequent cross-view data fusion and global situational awareness.

[0039] Sub-step S13 involves fusing multiple video data transformed to the world coordinate system to obtain the first fused video data in the world coordinate system. For example, under a unified world coordinate system, although multiple video streams possess spatial consistency, they remain dispersed data streams with varying perspectives. Directly processing them separately can easily lead to the same device being counted repeatedly or its trajectory being broken from different viewpoints. By fusing multi-source video in the world coordinate system (e.g., based on voxel grids, point cloud accumulation, or semantic map fusion), a comprehensive, blind-spot-free, and highly complete "first fused video data" (which can be understood as virtual panoramic video or a spatiotemporally consistent semantic map) can be generated. This fused data integrates the advantages of each perspective, effectively eliminating occlusions, filling gaps, and forming a global, continuous, and consistent digital representation of the construction area. This provides high-quality, unambiguous input for subsequent accurate extraction of equipment operating space and work trajectories, significantly improving the overall perception capability of the system.

[0040] Sub-step S14: Convert the fused video data in the world coordinate system to the pixel coordinate system to obtain the second fused video data corresponding to the construction process in the construction area.

[0041] For example, although fused data in the world coordinate system possesses good spatial consistency, most computer vision algorithms (such as object detection, instance segmentation, and optical flow estimation) still need to operate efficiently in the two-dimensional image domain (i.e., pixel coordinate system). Furthermore, the final visualization results (such as labeled videos and trajectory playback) also need to be presented in a standard video format. Therefore, projecting the fused data in the world coordinate system back to the pixel coordinate system of one or more preferred viewpoints to generate "second fused video data" preserves the information integrity and spatial accuracy of the multi-view fusion while remaining compatible with existing image processing pipelines. This video can be considered the "optimal synthetic viewpoint," possessing a wide field of view, low occlusion, and high resolution, facilitating stable and efficient execution of subsequent operations such as device cropping and trajectory extraction within a standard image analysis framework, thus achieving a balance between algorithmic practicality and engineering applicability.

[0042] For example, in key construction areas such as slope excavation, slag removal, support, and backfilling, fixed high-definition cameras and mobile intelligent inspection robots are deployed to achieve full-time and full-space video stream acquisition of the entire construction process. Timestamps and spatial coordinate labels are added to the video data to form a high-fidelity original data source. At the same time, a joint calibration method based on ground control points (GCPs) and camera intrinsic parameter calibration is introduced. At least six high-precision GCPs with coordinates obtained by RTK (Real-Time Kinematic) measurement can be set up in the construction area. Combined with multi-angle calibration board images, OpenCV is used to solve the camera intrinsic parameters (focal length, principal point, distortion coefficient), and nonlinear optimization algorithms are used to solve the extrinsic parameters (rotation and translation parameters). A mapping function from pixel coordinates to real-world three-dimensional coordinates is constructed. After verification with a test set, it provides an accurate spatial-visual fusion foundation for subsequent engineering quantity calculations.

[0043] Step 202: Identify each operating device in the video data and assign a device identifier to each operating device; For example, in complex construction scenarios, multiple types of similar work equipment (such as multiple excavators, dump trucks, etc.) often coexist. If they cannot be distinguished and uniquely identified, subsequent trajectory tracking, work behavior analysis, and workload statistics will not be accurately attributed to specific equipment, leading to measurement confusion or even double counting. Therefore, it is first necessary to locate and classify various types of work equipment from video data using target detection and recognition algorithms (such as YOLO, Mask R-CNN, etc.), and assign a unique equipment identifier (such as an ID number) to each detected equipment. This identifier serves as the equipment's "identity tag" throughout the video sequence, ensuring that it can be continuously tracked in subsequent frames, which is a prerequisite for achieving refined equipment-level management. Only by establishing a "one machine, one code" correspondence can we support the counting of work cycles by equipment and the calculation of individual workloads, thereby providing a reliable basis for construction scheduling, cost accounting, and performance evaluation.

[0044] Step 203: Track each working device according to the device identifier to determine the operating space corresponding to each working device in the construction area; the operating space includes the device boundary and the working scene of the working device; For example, a work scenario represents the context of equipment operation, such as functional areas like excavation areas, unloading points, support faces, or transportation channels. Their spatial location and semantic attributes together form important criteria for judging the effectiveness of equipment operations. Equipment tracking is a key link connecting static identification and dynamic analysis. Simply identifying equipment without tracking it fails to obtain its continuous movement trajectory, making it impossible to determine whether a complete work process has been completed. Cross-frame association using equipment identifiers allows for the reconstruction of the movement path of each piece of equipment in space and time. Based on this, its "operating space"—the functional spatial area jointly formed by the equipment itself (equipment boundary) and its surrounding work environment (work scenario)—is further determined. This operating space not only reflects the physical existence of the equipment but also its operational impact range. For example, the operating space of an excavator includes not only its vehicle outline but also the earthwork area within its excavation radius. Clearly defining this space helps to accurately define the "effective working area," avoiding misjudging equipment passing by or in standby mode as operational behavior, thereby improving the accuracy and semantic rationality of quantity calculations.

[0045] In one embodiment, step 203 includes the following sub-steps: Sub-step S21: During the process of tracking each working device according to the device identifier, determine the device boundary of each working device in the video data and the working scene of the working device. For example, equipment boundaries (usually obtained through instance segmentation) accurately describe the equipment's shape and position in the image, providing the fundamental geometric information for cropping target videos and extracting trajectories. The work scenario, on the other hand, reflects the current construction context of the equipment, such as "operating in a slope excavation area" or "in a backfilling area." This can be achieved by using a scene semantic segmentation model to automatically classify areas in the video into functional types such as "excavation area," "pouring area," or "transportation channel," or by combining pre-defined and semantically labeled construction functional areas (such as "backfilling area" or "support area"). Then, the corresponding work scenario is determined by matching the equipment's real-time position in the video to its location. Simultaneously determining both during tracking allows for dynamic perception of the interaction between the equipment and its environment.

[0046] Sub-step S22: Based on the equipment boundary and the operating scenario of the operating equipment, determine the operating space corresponding to each operating device in the construction area.

[0047] For example, different types of operating equipment (such as excavators, tower cranes, concrete pump trucks, and dump trucks) differ significantly in structure, size, degrees of freedom of movement, and typical operating modes, and therefore require different operating spaces. By pre-establishing a "equipment type - operating space template" mapping library (which can be built based on design parameters, historical data, or simulation models), the system can directly call the corresponding standardized operating space model after identifying the equipment type, and dynamically adapt it in combination with the current equipment boundaries (such as boom angle and telescopic length).

[0048] In one embodiment, sub-step S22 includes the following sub-steps: Sub-step S221: Determine the equipment boundary compensation value corresponding to the operating scenario of the operating equipment; For example, different operational scenarios have different requirements for safety distances and operational radii: equipment near blasting zones requires a larger safety buffer, while equipment in narrow passageways requires smaller compensation to avoid ineffective cutting. By pre-setting or learning a mapping rule for "operation scenario → compensation value" (such as lookup table or neural network regression), a reasonable outward extension distance can be automatically allocated to each piece of equipment in the current scenario. This compensation value is not a fixed constant, but a parameter that is dynamically adjusted according to the scenario, reflecting "site-specific" engineering intelligence. It ensures that the operating space covers the necessary activities and safety areas without excessive redundancy, thereby optimizing computing resources and improving the signal-to-noise ratio of subsequent trajectory extraction and operation determination while ensuring the integrity of the analysis.

[0049] Sub-step S222: Expand the equipment boundary corresponding to each working device in all directions according to the equipment boundary compensation value to determine the operating space corresponding to each working device in the construction area.

[0050] For example, the equipment boundary itself only represents the static area occupied by the equipment, while in actual operation, its influence range far exceeds the main body (such as boom swing, bucket digging, personnel safety distance, etc.). By uniformly or asymmetrically expanding the compensation value around the equipment boundary, a minimum circumscribed area (such as a rectangle, polygon, or ellipse) surrounding the equipment and its operational influence area can be formed, i.e., the operating space. This operating space is directly used to guide video cropping: cropping area = equipment boundary + compensation margin. In this way, the resulting target video can clearly present the main body of the equipment and fully include the key areas where its operational actions occur, avoiding trajectory breakage or action loss due to overly tight cropping, and also preventing irrelevant background interference from overly wide cropping. This spatial expansion strategy based on physical meaning is a key technical guarantee for achieving high-precision and highly robust engineering quantity calculations.

[0051] Reference Figure 3 The diagram illustrates a scenario of an engineering quantity calculation device provided by an embodiment of the present invention. In the diagram, d represents the equipment boundary compensation value; the work scenario is the physical environment in which the equipment performs its tasks in the construction area; the equipment boundary represents the outline range of the working equipment in the image; and the operating space is the enclosed area formed by extending outward by a distance d based on the equipment boundary, which is used to completely cover the equipment body and its work influence area, ensuring that video cropping can accurately capture effective work behavior.

[0052] Step 204: Crop the video data to obtain the target video corresponding to the operation space; For example, raw video data typically covers the entire construction area and contains a large amount of background information unrelated to the operation of specific equipment (such as other equipment, personnel, fixed structures, etc.). Analyzing the behavior of a single piece of equipment directly on the entire map is not only computationally expensive but also susceptible to interference leading to false positives or false negatives. By precisely cropping the video data according to the operating space of each piece of equipment, a "target video" containing only the equipment and its key operating area can be generated, achieving data focusing and noise reduction. This cropping strategy effectively isolates visual interference between multiple devices, improving the accuracy and stability of subsequent target detection, attitude estimation, and trajectory extraction. At the same time, the cropped video has a relatively higher resolution and more compact content, significantly reducing the processing burden on the algorithm, which is beneficial for engineering calculation scenarios with high real-time requirements. Furthermore, using the operating space as the basis for cropping ensures that the target video completely covers the equipment itself and the area affected by the operation, avoiding the loss of key action information due to overly narrow cropping, which is a prerequisite for achieving high-precision and high-efficiency operation behavior analysis.

[0053] Step 205: Determine the operating trajectory of the operating equipment in the target video; For example, the work trajectory is a core temporal feature reflecting the dynamic behavior of equipment, recording the equipment's position, speed, and movement path within the operating space over time. By extracting this trajectory from the target video, the actual working process of the equipment can be objectively and quantitatively reconstructed, providing direct evidence for determining whether it has performed effective work. For instance, the typical work trajectory of an excavator presents a periodic closed loop of "digging-lifting-reversing-returning," while a transport vehicle exhibits a linear path traveling back and forth between two points. Only by acquiring continuous and complete trajectories can these patterns be identified and the effective work distinguished from non-operational states such as idling or standby. Furthermore, trajectory data can also be used to analyze work efficiency, path optimization, and abnormal behavior detection. Therefore, determining the work trajectory is a crucial step from visual data to semantic understanding of engineering quantities, and an indispensable intermediate link in achieving automated and seamless engineering measurement.

[0054] Step 206: Based on the work trajectory, determine the number of times the work equipment performs the complete work process; For example, the amount of work (such as earthwork excavation, concrete pouring, and number of muck transport trips) is usually positively correlated with the number of times the equipment completes a "complete work process". For instance, each time an excavator completes a "dig-lift-dump-return" cycle, it represents the transfer of a certain volume of earth; each time a muck truck completes a "load-transport-unload-return" cycle, it is counted as one effective transport shift. Therefore, by analyzing the spatiotemporal characteristics of the work trajectory (such as trajectory shape, stopping points, directional changes, periodicity, etc.), the number of such complete work cycles can be automatically identified and counted. This process transforms continuous visual motion data into discrete, measurable engineering events, realizing the leap from "seeing the action" to "calculating the work quantity". Compared with traditional methods that rely on manual recording or predefined action classification, trajectory-based cycle counting is more objective, adaptable, and robust, and can effectively cope with complex and changing on-site conditions. This is the core logic of this invention for achieving high-precision automated work quantity calculation.

[0055] In one embodiment, step 206 includes the following sub-steps: Sub-step S31: Obtain the evaluation rules for a complete operation process of the working equipment; the evaluation rules include at least one of action sequence, spatial movement trajectory, and operation space context constraint; the action sequence is a series of time-sequential sub-actions required for a complete operation process; the movement trajectory is a continuous path from the starting position through intermediate work points to the ending position required for a complete operation process; the operation space context constraint is the existence state and position state required for the work object in a complete operation process. For example, traditional methods rely on predefined action categories (such as "digging" and "pouring"), which are difficult to adapt to complex and ever-changing on-site conditions. However, the evaluation rules of this invention can comprehensively characterize the essential features of the complete work process from three dimensions based on the target video corresponding to each working device: the action sequence ensures that sub-actions occur in the correct sequence (e.g., "positioning → digging → lifting → rotating → unloading"); the spatial movement trajectory verifies whether the equipment completes a closed-loop movement along a reasonable path (e.g., from a full-load state at the digging point, rotating to the unloading point, and then returning to an empty vehicle); and the operational space context constraint introduces environmental semantics, requiring the work object (e.g., a pile of soil, a silo) to be in the correct state and position at key nodes (e.g., in a valid work process, a pile of soil is dug from one location and piled up to another; the change in the pile state between the two locations determines whether it is a valid work process). At least one of the three conditions must be met to form a multimodal and strongly constrained definition of "what is an effective operation", which significantly improves the accuracy and generalization ability of loop identification, avoids the defect of single feature being easily interfered with, and enables the system to adapt to different equipment types, construction stages and site layouts, laying a logical foundation for high-reliability engineering quantity calculation.

[0056] Sub-step S32: Based on the operation trajectory in the target video, the model determines all operation processes in the operation trajectory by the number of iterations, and determines the number of complete operation processes that meet the evaluation rules from all operation processes.

[0057] For example, the original trajectory contains all movement behaviors, mixed with non-productive activities such as posture adjustments. The model first segments all potential "work flow" fragments (such as each round trip, each rotation cycle) from the trajectory, and then performs multi-dimensional verification on each of these fragments: Does it meet the preset action sequence logic? Does it follow a typical path topology? Is it executed in the correct operating space context? Only any flow that meets the evaluation rules such as action sequence, movement trajectory, and context constraints is recognized as a "complete work flow". This rule-based filtering mechanism effectively eliminates noise and interference, ensuring that the statistical results truly reflect production activities. This process realizes a leap from "seeing movement" to "understanding the job", so that the calculation of engineering quantities no longer relies on manual intervention or fixed action templates, but is based on intelligent judgment of the actual behavior of the equipment, which greatly improves the system's automation level, robustness, and adaptability to complex scenarios.

[0058] Sub-step S33: Obtain the number of historical complete operation processes corresponding to the operating equipment; For example, historical data reflects the typical operating efficiency, cycle time, and behavior patterns of the equipment under similar working conditions (e.g., a certain model of excavator completes an average of 15 digging and loading cycles per hour). This information can be used to verify the reasonableness of the current identification results in real time: if the currently identified cycle count is much higher or lower than the historical average, it may indicate an identification error, equipment failure, or sudden change in working conditions. Furthermore, historical counts can also be used for equipment performance evaluation, capacity prediction, and resource scheduling optimization. More importantly, it provides a supervisory signal for model adaptive optimization—by comparing historical patterns with the current output, the thresholds or weights of the evaluation rules can be dynamically adjusted, enabling the system to continuously adapt to long-term evolutionary factors such as equipment aging and changes in operating habits while maintaining stability, thereby ensuring the long-term consistency and reliability of engineering quantity statistics.

[0059] Sub-step S34: Optimize the number of cycles to determine the model based on the number of times the historical complete job process has been completed.

[0060] For example, the initial model's evaluation rules (such as trajectory length thresholds and action interval tolerances) are usually based on prior knowledge and may deviate when faced with new equipment, new working conditions, or differences in operating habits. By introducing historical job counts as feedback signals, a closed-loop optimization mechanism can be constructed: if the model's recent output of the number of cycles continuously deviates from the historical average of the equipment (excluding changes in working conditions), the model parameters are automatically fine-tuned (such as relaxing trajectory matching tolerance and adjusting the action sequence error tolerance window) to bring the recognition results back to a reasonable range. This self-correction mechanism based on historical performance significantly enhances the model's robustness and generalization ability, avoiding overfitting or underfitting problems caused by fixed thresholds. Ultimately, the system can not only accurately identify the current job but also continuously optimize its performance over time, achieving intelligent engineering quantity measurement that becomes more accurate with use.

[0061] Step 207: Determine the workload based on the number of times the work equipment in the construction area completes the entire work process.

[0062] For example, the volume of work (such as earthwork excavation, concrete pouring, and waste soil transportation) is essentially directly related to the number of effective work cycles completed by the equipment. For instance, each complete work cycle completed by an excavator typically corresponds to the transfer of a fixed volume of earth; each complete work cycle completed by a dump truck represents a standard transport shift. Therefore, using the "number of complete work cycles" as the core input for work volume calculation can transform abstract visual behavior into quantifiable and cumulative production indicators. Compared to traditional methods that rely on manual reporting or sensor-accumulated working hours, this approach is more objective and accurate, and can effectively distinguish between effective work and non-productive states such as idling or standby. By using the actual completed work cycles as the unit of measurement, not only is the accuracy of work volume statistics improved, but a highly reliable data foundation is also provided for progress monitoring, cost accounting, and resource scheduling, making it a key link in achieving intelligent and automated construction management.

[0063] In one embodiment, step 207 includes the following sub-steps: Sub-step S41: Obtain the rated parameters, historical average operating efficiency, and total operating time corresponding to each operating device; For example, the working capabilities of different equipment models vary significantly (e.g., the earthmoving volume per cycle differs by several times between a small excavator and a large excavator), making it impossible to accurately calculate the workload based solely on the number of cycles. Rated parameters provide a theoretical upper limit, historical average operating efficiency reflects performance under actual working conditions, and total operating time is used to verify the reasonableness of the number of cycles (e.g., to exclude abnormally high-frequency misjudgments). Combining these three factors allows for refined modeling of individual equipment, avoiding a crude "one-size-fits-all" estimation. Furthermore, historical efficiency data can compensate for performance fluctuations caused by equipment aging, operator skill level, or site conditions, making the workload calculation closer to the actual site conditions and significantly improving the accuracy and reliability of the results.

[0064] Sub-step S42: Determine the product between the number of complete operation processes, rated parameters, and historical average operating efficiency for each operating device; For example, the number of complete operation cycles indicates how many effective operations the equipment has performed; rated parameters (such as bucket capacity) provide the theoretical maximum output of a single operation; while historical average operating efficiency (such as actual filling rate, effective operation ratio) dynamically adjusts the theoretical value, reflecting the effective output ratio under actual working conditions (e.g., theoretical bucket capacity 1.5m³, but actual average loading is only 1.2m³). The product of these three can be regarded as the "equivalent standard operating output" of the equipment in the current period, which not only retains the individual characteristics of the equipment but also integrates historical experience data, effectively overcoming the overestimation caused by relying solely on rated parameters or the lag caused by using only historical efficiency. This product, as an intermediate quantity, provides a scientific and robust input for subsequent standardized engineering quantity calculations.

[0065] Sub-step S43: Determine the target workload based on the ratio of the product to the total operation time.

[0066] For example, by calculating the ratio of the aforementioned product (i.e., equivalent standard work output) to the total work duration, the effective engineering output rate per unit time is essentially calculated. This rate is then used to calibrate the final work quantity, eliminating deviations caused by work interruptions, inefficient operation, or identification errors. For instance, if a piece of equipment identifies 10 cycles in one hour, but the total work duration is only 30 minutes (the remaining time is spent offline or idling), directly using 10 cycles might lead to an overestimation. However, by using "output / total work duration," a more realistic work intensity can be obtained, and a reasonable work quantity can be deduced accordingly. Furthermore, this ratio can also be used to compare the work efficiency of different equipment or work groups, assisting management decisions. More importantly, introducing a time dimension enables dynamic normalization of work quantity calculations, enhancing the system's adaptability to complex scenarios such as discontinuous operations and intermittent construction, ensuring that the final result reflects both the actual completed workload and meets the requirements of engineering measurement standards.

[0067] In one embodiment, before acquiring video data corresponding to the construction process of the construction area, the method further includes: receiving a user's engineering quantity acquisition instruction; and determining, based on the engineering quantity acquisition instruction, the construction area for which the engineering quantity is to be calculated, and the construction process of the construction area.

[0068] For example, construction sites typically cover a wide area with numerous work areas. Indiscriminately collecting and analyzing video data across the entire site would waste computing resources, increase processing delays, and potentially introduce irrelevant interference. By allowing users to proactively initiate quantity acquisition commands (such as selecting a slope excavation section or a foundation pit support area), the system can clearly define the target area and time period of interest, retrieving only video streams from relevant cameras or inspection equipment, focusing on specific construction processes (such as "earthwork excavation" or "concrete pouring"). This not only improves the targeting and efficiency of data processing but also ensures that the quantity calculation results align with the user's actual management needs, avoiding the problem of "comprehensive but irrelevant calculations." Simultaneously, this mechanism supports flexible query and verification scenarios, enhancing the system's practicality and human-machine collaboration capabilities, and is a crucial preliminary step towards achieving intelligent, on-demand engineering measurement services.

[0069] This invention provides a method for calculating engineering quantities. The method involves acquiring video data corresponding to the construction process in a construction area; determining the operating space for each piece of equipment based on the video data; the operating space including the equipment boundary and the operating scene of the equipment; cropping the video data to obtain a target video corresponding to the operating space; determining the operating trajectory of the equipment in the target video; determining the number of times the equipment performs a complete operation based on the operating trajectory; and determining the engineering quantity based on the number of times the equipment performs a complete operation. This invention, by determining the operating space including equipment boundaries and the operating scene, and cropping the video data to generate target videos corresponding to each piece of equipment, accurately extracts the operating trajectory and counts the number of operations. This eliminates the reliance on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of engineering quantity calculation.

[0070] Considering the diverse needs for quantity calculation at construction sites, often expressed in natural language, traditional methods require manual coding or parameter configuration to adapt to different tasks, which is not only inefficient but also prone to errors. Furthermore, directly using a large language model to process all stages may introduce semantic ambiguity or calculation bias, affecting the accuracy of quantity statistics. Therefore, in one embodiment, this application simultaneously introduces a large language model and specialized small models: the large language model is used to compile natural language requirements into structured event analysis files, breaking down complex, open natural language instructions into executable and verifiable sub-task sequences; while deterministic small models such as text classification, construction machinery detection and tracking, and spatiotemporal deduplication are used for task routing, event boundary determination, and quantity calculation. By combining the semantic understanding capabilities of the large language model with the closed-loop measurement capabilities of the small models, this application can automatically convert the user's natural language description into a structured instantiated configuration file, thereby driving the subsequent quantity calculation process. Specifically, before executing step 201 to obtain video data corresponding to the construction process in the construction area, the method further includes the following steps: S301. Based on the large language model, the acquired natural language description data is converted into event definition files in a predefined format, and the event definition template and task category labels are determined based on the event definition files; It should be noted that the large language model analyzes the core intent and key constraints in natural language descriptions (e.g., analyzing the excavation volume of the slope section from K12+300 to K12+700) and converts them into event definition files in a predefined format, such as JSON (JavaScript Object Notation). After manual verification and improvement, these are saved as fixed event definition templates, and a unique task category label is created for each verified and improved event definition template. The event definition template includes, but is not limited to, the following key fields: task and event identifier (defining the task type and event name), spatial range (delineating the physical boundary of the event, including regional coordinates, spatial range description and coordinate system information, such as mileage, station number and slope three-dimensional boundary), object of interest and its change characteristics (listing the core objects and their attributes used to judge the event, such as the type of construction machinery used for auxiliary verification, and explaining how to measure its change), event logic definition (including the event triggering conditions, time sequence boundaries and quantitative output requirements for each independent event, such as excavation volume, duration, etc.), required data modality (declaring the data types that need to be collected to execute the task, such as video frames), and accuracy and error constraints (hard indicator requirements for the final quantitative results, such as time boundary error <1 hour, as a standard for algorithm acceptance).

[0071] S302. Based on the event definition file, determine short text samples, and train the initial text classification model based on the short text samples and their corresponding task category labels to obtain a fully trained text classification model. It should be noted that the text classification model is trained as a lightweight model for short text classification, such as a distilled Chinese BERT (Bidirectional Encoder Representations from Transformers) or TextCNN (Text Convolutional Neural Network). The input of the text classification model is a series of labeled short text samples. Each short text sample is generated from the key fields (such as task type, event name, and spatial range) of the event definition file (such as a JSON file) through template combination, synonym rewriting, and sentence transformation, which are natural language that the user may input. For example, "Analyze the excavation volume of the slope from K12+300 to K12+700" or "Help me see how many cubic meters of this slope have been excavated". The output of the text classification model during training is the task category label of this short text sample, i.e., ID (Identity).

[0072] S303. Based on a well-trained text classification model, perform text classification on the input natural language description data to be processed, and obtain the target task category label and its corresponding target event definition file template. It should be noted that the fully trained text classification model takes the natural language description data to be processed as input and outputs the task category label and its corresponding confidence score. If the confidence score is not lower than the threshold, it is used as the target task category label to determine the target event definition file template to be activated. If the confidence score is lower than the threshold, the natural language description to be processed is manually matched to the corresponding target event definition file template.

[0073] S304. Based on the slot filling model, dynamic parameters are extracted from the natural language description data to be processed and filled into the target event definition file template to obtain an instantiated configuration file. It should be noted that the slot filling model uses named entity recognition or rules. For example, specific dynamic parameters such as mileage station number and time range are extracted from the natural language description data to be processed, and injected into the corresponding fields of the target event definition file template to generate an instantiated configuration file.

[0074] S305. Based on the scheduling model, the configuration file is parsed and executed to implement the engineering quantity calculation method of this application.

[0075] Specifically, the scheduling model performs the workload calculations according to the following process: Step 1: The scheduling model reads the "Required Data Modality" field from the instantiated configuration file, iterates through each data modality listed in this field (e.g., "Fixed Monitoring Video Frames"), and checks whether the corresponding data service is available. If any required data modality is unavailable, an error message is returned and the task is aborted; if all required data modalities are available, the model calls the data interface to retrieve the corresponding raw video and camera calibration parameters (e.g., camera intrinsic and extrinsic parameters, distortion coefficients, etc.) based on the area coordinates (e.g., mileage marker range) and coordinate system information within the "Spatial Range" in the configuration file, combined with the time range in the user command.

[0076] It should be noted that fixed high-definition cameras and mobile intelligent inspection robots are deployed in key construction areas such as slope excavation, slag removal, support, and backfilling. This deployment enables the full-time and full-space video stream acquisition of the entire slope construction process, and the original video is obtained by combining timestamps and spatial coordinate labels.

[0077] Step 2: The scheduling model reads the "Construction Machinery" object from the "Objects of Interest" to obtain the construction machinery types for auxiliary verification (e.g., excavators, loaders). The scheduling model calls a target detection model, such as YOLO (You Only LookOnce), to detect each frame of video, identifying the specified type of machinery and its bounding box. Then, it calls a multi-object tracking model, such as DeepSORT (Simple Online and Realtime Tracking with a DeepAssociation Metric), to assign a unique identifier to each machine, generate its continuous trajectory on the timeline, and record the appearance time and dwell time of each machine. Next, based on the detected machine bounding box, a local image patch containing the machine and its surrounding environment is cropped from the original video data according to a compensation value (e.g., expanding outwards by 20%), and uniformly scaled to a fixed size (e.g., 640 by 640 pixels) to form an image sequence, i.e., the target video.

[0078] It should be noted that the small model (i.e., the target detection and tracking model) focuses on the detection and tracking of construction machinery. By cropping out the image of a single excavator, it effectively avoids interference with the quantity statistics caused by multiple excavators operating simultaneously.

[0079] Step 3: The scheduling model reads the "Quantitative Output Requirements" field in the configuration file, confirms the need to output "excavation volume," and then triggers the multimodal large model (i.e., the loop count determination model). The cropped target video is input into the multimodal large model, accompanied by natural language prompts (e.g., "Count the number of complete 'dig-lift-pour-back' loops completed by this excavator"). The multimodal large model analyzes the operation trajectory in the target video, identifies the excavator's action sequence and spatial movement trajectory, and combines the operational space context constraints to determine each complete operation process. When at least one of the above evaluation rules is satisfied, it is counted as a valid load. Finally, the multimodal large model outputs the number of loads within the entire video segment, which is the number of complete operation processes in this application. Then, the scheduling model obtains the rated parameters of the machine (i.e., rated bucket capacity, queried from the equipment parameter database) and the historical average operating efficiency (i.e., historical average loading rate, obtained from historical statistical data), and calculates the product of the number of complete operation processes, rated parameters, and historical average operating efficiency. At the same time, based on the timestamps of the machine's first and last appearance in Step 2, the total operation time is determined. The target workload is determined based on the ratio of the product to the total operation time: Target workload = (Number of complete operation processes × Rated parameters × Historical average operation efficiency) / Total operation time. If the configuration file also requires the output of "Event start and end time" and "Duration", then the event start and end time and duration are encapsulated together. Finally, the calculated target workload, event start and end time, duration, etc., are encapsulated according to the format required by the configuration file and returned as the workload statistics result.

[0080] It should be noted that the multimodal large model directly processes the cropped target video. By jointly analyzing the mechanical posture, environmental context and related objects, it achieves zero-sample calculation of the number of loading times of construction machinery without the need for predefined action categories, thereby improving the system's generalization ability and adaptability.

[0081] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0082] Reference Figure 4 The diagram shows a structural block diagram of an engineering quantity calculation device provided by an embodiment of the present invention, which may specifically include the following modules: The video data acquisition module 301 is used to acquire video data corresponding to the construction process in the construction area. The operation space determination module 302 is used to determine the operation space corresponding to each working device in the construction area based on the video data; the operation space includes the device boundary of the working device and the working scene of the working device; the working scene includes a local construction area associated with the construction task currently being performed by the working device; The target video acquisition module 303 is used to crop the video data to obtain the target video corresponding to the operation space; The operation trajectory determination module 304 is used to determine the operation trajectory of the operation equipment in the target video; The number of times determination module 305 is used to determine the number of times the work equipment performs the complete work process based on the work trajectory; The quantity determination module 306 is used to determine the quantity of work based on the number of times the working equipment in the construction area performs a complete work process.

[0083] In one embodiment, the video data acquisition module includes: The video data acquisition submodule is used to acquire video data of the construction area captured by multiple acquisition devices; The coordinate transformation submodule is used to transform the video data of the construction area captured by the multiple acquisition devices from the pixel coordinate system to the world coordinate system; The first fused video data determination submodule is used to fuse multiple video data transformed to the world coordinate system to obtain the first fused video data in the world coordinate system. The second fused video data determination submodule is used to transform the fused video data in the world coordinate system to the pixel coordinate system to obtain the second fused video data corresponding to the construction process in the construction area.

[0084] In one embodiment, the operating space determination module includes: The device identifier allocation submodule is used to identify each working device in the video data and assign a device identifier to each working device. The operating space determination submodule is used to track each working device based on the device identifier and determine the operating space corresponding to each working device in the construction area.

[0085] In one embodiment, the operation space determination submodule includes: The work scene determination unit is used to determine the device boundary of each work device and the work scene of the work device in the video data during the process of tracking each work device according to the device identifier; The operating space determination unit is used to determine the operating space corresponding to each piece of equipment in the construction area based on the equipment boundary and the operating scenario of the operating equipment.

[0086] In one embodiment, the operating space determination unit includes: The compensation value determination subunit is used to determine the equipment boundary compensation value corresponding to the operating scenario of the operating equipment. The operating space determination subunit is used to expand the equipment boundary corresponding to each working device in all directions according to the equipment boundary compensation value, thereby determining the operating space corresponding to each working device in the construction area.

[0087] In one embodiment, the number determination module includes: The evaluation rule acquisition submodule is used to acquire the evaluation rules for a complete operation process of the working equipment; the evaluation rules include at least one of action sequence, spatial movement trajectory, and operation space context constraint; the action sequence is a series of time-sequential sub-actions required for a complete operation process; the movement trajectory is a continuous path from the starting position through intermediate work points to the ending position required for a complete operation process; the operation space context constraint is the existence state and position state required for the work object in a complete operation process. The number of iterations determination submodule is used to determine all the operation processes in the operation trajectory in the target video based on the operation trajectory through the loop number determination model, and to determine the number of complete operation processes that meet the evaluation rules from all the operation processes.

[0088] In one embodiment, the device further includes: The historical data acquisition unit is used to acquire the number of complete historical operation processes corresponding to the operating equipment; The model optimization unit is used to determine the model by optimizing the number of cycles based on the number of times the historical complete job process has been completed.

[0089] In one embodiment, the quantity determination module includes: The parameter acquisition submodule is used to acquire the rated parameters, historical average operating efficiency, and total operating time of each operating device. The quantity determination submodule is used to determine the product of the number of complete operation processes, rated parameters, and historical average operating efficiency for each operating device. The target workload is determined based on the ratio of the product to the total operation time.

[0090] In one embodiment, the device further includes: The instruction acquisition module is used to receive the user's instructions for acquiring project quantities; The construction process determination module is used to determine the construction area for which the quantity of work to be calculated, and the construction process of the construction area, based on the quantity of work acquisition instruction.

[0091] This invention provides a work quantity calculation device that acquires video data corresponding to the construction process in a construction area; determines the operating space corresponding to each piece of equipment based on the video data; the operating space includes the equipment boundary and the operating scene of the equipment; crops the video data to obtain a target video corresponding to the operating space; determines the operating trajectory of the equipment in the target video; determines the number of times the equipment performs a complete work process based on the operating trajectory; and determines the work quantity based on the number of times the equipment performs a complete work process. This invention, by determining the operating space including equipment boundaries and the operating scene, crops the video data to generate target videos corresponding to each piece of equipment, thereby accurately extracting the operating trajectory and counting the number of operations. This eliminates the dependence on predefined action categories, effectively suppresses error propagation, and significantly improves the accuracy and robustness of work quantity calculation.

[0092] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0093] This invention also provides an electronic device, comprising: It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described engineering quantity calculation method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0094] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described engineering quantity calculation method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0095] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0096] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0097] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0098] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0099] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0100] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0101] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0102] The foregoing has provided a detailed description of the engineering quantity calculation method and the engineering quantity calculation device provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for calculating engineering quantities, characterized in that, The method includes: Acquire video data corresponding to the construction process in the construction area; Based on the video data, the operating space corresponding to each piece of work equipment in the construction area is determined; the operating space includes the equipment boundary of the work equipment and the work scene of the work equipment; the work scene includes a local construction area associated with the construction task currently being performed by the work equipment; The video data is cropped to obtain the target video corresponding to the operation space; Determine the operating trajectory of the operating equipment in the target video; Based on the work trajectory, determine the number of times the work equipment performs a complete work process; The workload is determined based on the number of times the equipment in the construction area completes the entire work process.

2. The method for calculating engineering quantities according to claim 1, characterized in that, The acquisition of video data corresponding to the construction process in the construction area includes: Acquire video data of the construction area captured by multiple acquisition devices; The video data of the construction area captured by the multiple acquisition devices is converted from the pixel coordinate system to the world coordinate system; Multiple video data transformed to the world coordinate system are fused to obtain the first fused video data in the world coordinate system. The fused video data in the world coordinate system is transformed to the pixel coordinate system to obtain the second fused video data corresponding to the construction process in the construction area.

3. The method for calculating engineering quantities according to claim 1, characterized in that, The step of determining the operating space corresponding to each piece of work equipment in the construction area based on the video data includes: Identify each operating device in the video data and assign a device identifier to each operating device; The operation space corresponding to each piece of equipment in the construction area is determined by tracking each piece of equipment based on the equipment identifier.

4. The method for calculating engineering quantities according to claim 3, characterized in that, The step of tracking each piece of work equipment based on the equipment identifier and determining the operating space corresponding to each piece of work equipment in the construction area includes: During the tracking of each working device based on the device identifier, the device boundary of each working device in the video data and the working scene of the working device are determined; Based on the equipment boundaries and the operating scenarios of the operating equipment, the operating space corresponding to each operating device in the construction area is determined.

5. The method for calculating engineering quantities according to claim 4, characterized in that, The step of determining the operating space corresponding to each piece of equipment in the construction area based on the equipment boundary and the operating scenario of the equipment includes: Determine the equipment boundary compensation value corresponding to the operating scenario of the operating equipment; The equipment boundaries corresponding to each working device are expanded outwards according to the equipment boundary compensation value to determine the operating space corresponding to each working device in the construction area.

6. The method for calculating engineering quantities according to claim 1, characterized in that, The step of determining the number of times the working equipment in the construction area performs a complete work process based on the work trajectory includes: The evaluation rules for a complete operation process of the equipment are obtained; the evaluation rules include at least one of action sequence, spatial movement trajectory and operation space context constraint; the action sequence is a number of time-sequential sub-actions required for a complete operation process; the movement trajectory is a continuous path from the starting position through intermediate work points to the ending position required for a complete operation process; the operation space context constraint is the existence state and position state required for the operation object in a complete operation process. The model determines the total number of work processes in the work trajectory based on the work trajectory in the target video by using the number of iterations, and then determines the number of complete work processes that meet the evaluation rules from all the work processes.

7. The method for calculating engineering quantities according to claim 6, characterized in that, The method further includes: Obtain the historical complete operation process count for the corresponding operating equipment; The model is determined by optimizing the number of iterations based on the number of complete historical workflows.

8. The method for calculating engineering quantities according to claim 1, characterized in that, The determination of the workload based on the number of times the work equipment in the construction area performs a complete work process includes: Obtain the rated parameters, historical average operating efficiency, and total operating time for each of the operating devices; Determine the product of the number of complete operation cycles, rated parameters, and historical average operating efficiency for each operating device; The target workload is determined based on the ratio of the product to the total operation time.

9. The method for calculating engineering quantities according to claim 1, characterized in that, Before acquiring the video data corresponding to the construction process in the construction area, the method further includes: Receive the user's instruction to retrieve the amount of work completed; Based on the quantity acquisition instruction, the construction area for which the quantity of work to be calculated is determined, as well as the construction process of the construction area.

10. A device for calculating engineering quantities, characterized in that, The device includes: The video data acquisition module is used to acquire video data corresponding to the construction process in the construction area; The operation space determination module is used to determine the operation space corresponding to each working device in the construction area based on the video data; the operation space includes the device boundary of the working device and the working scene of the working device; the working scene includes a local construction area associated with the construction task currently being performed by the working device; The target video acquisition module is used to crop the video data to obtain the target video corresponding to the operation space; The operation trajectory determination module is used to determine the operation trajectory of the operation equipment in the target video; The number of times determination module is used to determine the number of times the work equipment performs the complete work process based on the work trajectory; The quantity determination module is used to determine the quantity of work based on the number of times the working equipment in the construction area performs a complete work process.

11. An electronic device, characterized in that, include: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the steps of the quantity calculation method as described in any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the quantity calculation method as described in any one of claims 1-9.

Images