Traffic construction safety risk identification method and device, electronic equipment and storage medium

Abstract

The present application relates to the technical field of traffic infrastructure construction period site safety control, and particularly relates to a traffic construction safety risk identification method and device, an electronic device and a storage medium, wherein the method first acquires a first coordinate queue and a first image queue; then performs standardization processing on a plurality of first images in the first image queue to obtain a second image queue; then extracts a moving track of a known work unit in an image coordinate system from the second image queue, and constructs a first coordinate mapping table of image coordinate system conversion to entity coordinate system according to the moving track and the first coordinate queue; finally, identifies a non-known work unit from the video image, obtains a first entity coordinate according to the image coordinate of the non-known work unit and the first coordinate mapping table, and performs early warning on the work unit according to the first entity coordinate. The present application combines visual full-quantity identification and positioning record data to accurately identify non-known work units, and fills the regulatory blind spot of the control system.

Description

Technical Field

[0001] This invention relates to the field of on-site safety management and control during the construction period of transportation infrastructure, and in particular to a method, device, electronic equipment and storage medium for identifying safety risks during transportation construction. Background Technology

[0002] Currently, transportation engineering construction sites widely adopt technologies such as Beidou high-precision positioning and UWB positioning, equipping registered workers and construction machinery with positioning terminals. By collecting real-time physical coordinates, safety control measures such as electronic fences and boundary crossing alarms are achieved, which is the mainstream technical means for safety management at construction sites.

[0003] However, the scope of control for this type of technology is strictly limited to known work units that have been equipped with positioning terminals. For unknown work entities that do not have positioning terminals, such as temporary social vehicles, external equipment, and unregistered mobile personnel, it is impossible to obtain their location and movement information, forming a significant regulatory blind spot, which can easily lead to safety accidents such as cross collisions and unauthorized intrusions.

[0004] To fill management blind spots, video surveillance has been introduced in some areas to assist in management, but it largely relies on manual monitoring and analysis, resulting in poor real-time performance and a high rate of missed detections. Existing machine vision-based automatic detection solutions are susceptible to interference from various working conditions, such as dust obstruction, fluctuating day and night lighting, and complex and changing backgrounds, leading to insufficient stability in target detection and trajectory extraction. Furthermore, most vision solutions lack a unified spatial coordinate reference with the positioning system, making it difficult to integrate visually recognized targets into the existing positioning and management system for accurate spatial early warning. Camera spatial calibration requires manual target placement, resulting in high deployment and maintenance costs, making it difficult to adapt to the dynamic adjustment needs of construction scenarios, and failing to meet the intelligent safety management requirements of full target coverage and highly reliable operation during the construction period.

[0005] Therefore, it is necessary to develop a method for identifying traffic construction safety risks. Summary of the Invention

[0006] The present invention provides a method, device, electronic device and storage medium for identifying traffic construction safety risks, which is used to solve the problem that visually recognized targets are difficult to incorporate into the existing positioning and control system to achieve accurate spatial early warning in the prior art.

[0007] In a first aspect, embodiments of the present invention provide a method for identifying safety risks during traffic construction, comprising: Obtain a first coordinate queue and a first image queue, wherein the first coordinate queue corresponds to a known job unit, the first coordinate queue includes multiple first coordinates sorted according to time nodes, and each first coordinate corresponds to a first image in the first image queue; The first images in the first image queue are standardized to obtain the second image queue. Extract the movement trajectory of the known work unit in the image coordinate system from the second image queue, and construct a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system based on the movement trajectory and the first coordinate queue; Identify unknown work units from video images, obtain first entity coordinates based on the image coordinates of the unknown work units and the first coordinate mapping table, and issue an early warning to the work units based on the first entity coordinates.

[0008] In one possible implementation, the standardization process of multiple first images in the first image queue to obtain a second image queue includes: Multiple first images in the first image queue are converted to grayscale to obtain multiple first grayscale images; Extract the mean and standard deviation of each first grayscale image and construct a grayscale array; Based on multiple grayscale arrays, the inter-frame differences of the multiple first grayscale images are verified to obtain the inter-frame difference index; If the inter-frame difference index is higher than the index threshold, then each first grayscale image is adjusted according to the plurality of grayscale arrays; Otherwise, the plurality of first grayscale images are arranged in order of time nodes to obtain the second image queue.

[0009] In one possible implementation, adjusting each first grayscale image according to the plurality of grayscale arrays includes: Extract multiple grayscale means and multiple standard deviations from the multiple grayscale arrays; Based on the multiple grayscale means and the multiple standard deviations, determine the target mean and the target standard deviation; For each first grayscale image, perform the following steps: Based on the standard deviation of the first grayscale image and the target standard deviation, the pixel values ​​of the first grayscale image are scaled to obtain the first intermediate grayscale image. Pixel value compensation is performed based on the mean of the first intermediate grayscale image and the target mean to obtain the adjusted first grayscale image.

[0010] In one possible implementation, extracting the movement trajectory of the known work unit in the image coordinate system from the second image queue includes: Each image in the second image queue is segmented according to a preset segmentation template; Image data blocks corresponding to the same segmentation position are constructed into a data block array; Perform trajectory trace analysis on each data block array, and use the data block array containing trajectory traces as the trajectory array; Extract the time nodes corresponding to the image blocks containing the trajectory from the trajectory array, and use them as trajectory time nodes; Determine the coordinates of the trajectory image based on the segmentation positions corresponding to the trajectory array; The trajectory image coordinates are arranged according to the order of the trajectory time nodes to obtain the movement trajectory.

[0011] In one possible implementation, the step of performing trajectory trace analysis on each data block array, using the data block array containing trajectory traces as the trajectory array, includes: Perform the following steps for each data block array: Arrange the image data blocks in the data block array according to the corresponding time nodes to obtain the image block arrangement; Subtract two adjacent image data blocks to obtain multiple difference blocks; Calculate the sum of the absolute values ​​of multiple values ​​within each difference block, and use this as the block difference; Calculate the mean difference between multiple block differences; If there is a block difference value that deviates from the mean difference value by more than the deviation threshold, then the data block array is used as the trajectory array.

[0012] In one possible implementation, identifying unknown work units from video images and obtaining the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table includes: Identify multiple first moving targets from video images; Based on the number of the plurality of first moving targets and the plurality of second moving targets, it is determined whether there is an unknown work unit, wherein the second moving targets are targets determined based on the positioning system, the targets determined based on the positioning system are known work units, and each second moving target corresponds to a second entity coordinate; If there are unknown work units, perform the following steps: For each first moving target, extract the coordinates in the image as the first coordinates; For each first coordinate, the third entity coordinate is obtained by mapping and transformation through the first coordinate mapping table; For each second entity coordinate, find the nearest coordinate from multiple third entity coordinates and use it as the matching coordinate; Use the coordinates of the third entity that has not yet been matched as the coordinates of the first entity.

[0013] In one possible implementation, the step of issuing a warning to the work unit based on the first entity coordinates includes: Based on the coordinates of the first entity and the speed and direction of the unknown work unit, plan a predicted trajectory line; The coordinates of a fourth entity that is less than a first distance threshold from the predicted trajectory line are used as warning targets, wherein the coordinates of the fourth entity are the coordinates of a known work unit, and the entity coordinates of the known work unit are determined based on the positioning system. A warning message is sent to the warning target indicating that an unknown work unit is approaching.

[0014] In a second aspect, embodiments of the present invention provide a traffic construction safety risk identification device for implementing the traffic construction safety risk identification method as described in the first aspect or any possible implementation thereof, the traffic construction safety risk identification device comprising: The work unit coordinate acquisition module is used to acquire a first coordinate queue and a first image queue. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted by time nodes. Each first coordinate corresponds to a first image in the first image queue. An image normalization processing module is used to normalize multiple first images in the first image queue to obtain a second image queue. The coordinate mapping module is used to extract the movement trajectory of the known work unit in the image coordinate system from the second image queue, and construct a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system based on the movement trajectory and the first coordinate queue. as well as, The risk warning module is used to identify unknown work units from video images, obtain the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table, and issue a warning to the work units based on the first entity coordinates.

[0015] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect.

[0016] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first aspect or any possible implementation thereof.

[0017] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows: This invention discloses a method for identifying safety risks during traffic construction. First, a first coordinate queue and a first image queue are acquired. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted by time nodes. Each first coordinate corresponds to a first image in the first image queue. Then, the multiple first images in the first image queue are standardized to obtain a second image queue. Next, the movement trajectory of the known work unit in the image coordinate system is extracted from the second image queue, and a first coordinate mapping table is constructed based on the movement trajectory and the first coordinate queue to transform the image coordinate system into an entity coordinate system. Finally, non-known work units are identified from video images, and first entity coordinates are obtained based on the image coordinates of the non-known work units and the first coordinate mapping table. An early warning is then issued to the work units based on the first entity coordinates.

[0018] This invention relies on known work units equipped with positioning terminals to simultaneously collect entity coordinates and visual trajectories to automatically construct a coordinate mapping table. Camera spatial calibration can be completed without the need for manual deployment of calibration targets, significantly reducing on-site deployment and maintenance costs. It can also adapt to dynamic scenarios such as changes in monitoring perspective and site movement during construction.

[0019] This invention effectively counteracts image illumination fluctuations caused by daylight, dust obstruction, and changes in weather by using grayscale inter-frame difference verification and adaptive standardization processing. This significantly improves the robustness of target detection and trajectory extraction under complex working conditions and reduces false detections and missed detections caused by environmental interference.

[0020] This invention employs a segmented difference trajectory trace analysis method, which transforms global computation into local segmented processing, thereby reducing computational overhead while ensuring detection accuracy and adapting to real-time operation of edge devices.

[0021] This invention integrates full visual recognition with location registration data to accurately identify unknown work units without location terminals, filling the regulatory blind spots of the traditional management and control system; combined with motion trajectory prediction, it enables early warning of risks, shifting safety management from post-event handling to pre-event prevention, effectively reducing the risk of site collisions and unauthorized entry accidents, and comprehensively improving the efficiency of intelligent management and control of construction safety. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1This is a flowchart of the traffic construction safety risk identification method provided by the embodiments of the present invention; Figure 2 This is a schematic diagram of the motion trajectory trace extraction process provided by the embodiments of the present invention; Figure 3 This is a functional block diagram of the traffic construction safety risk identification device provided in the embodiments of the present invention; Figure 4 This is a functional block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0024] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0025] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0026] The embodiments of the present invention will be described in detail below. This example is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes. However, the protection scope of the present invention is not limited to the following embodiments.

[0027] Figure 1 A flowchart of a traffic construction safety risk identification method provided for an embodiment of the present invention.

[0028] like Figure 1 As shown, a flowchart illustrating the implementation of the traffic construction safety risk identification method provided by an embodiment of the present invention is presented, and is described in detail below: In step 101, a first coordinate queue and a first image queue are obtained. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted according to time nodes. Each first coordinate corresponds to a first image in the first image queue.

[0029] For example, the following provides a detailed description of the traffic construction safety risk identification method provided by the embodiments of the present invention. This method is applicable to all-weather safety management and control of construction sites for traffic infrastructure such as highways and municipalities. Through video analysis and high-precision positioning technology, it realizes automatic identification and early warning of risks for non-registered work entities, effectively making up for the shortcomings of traditional positioning management and control, such as low efficiency and many blind spots.

[0030] The acquisition of the first coordinate queue and the first image queue provides the basic data source for spatiotemporal alignment of subsequent coordinate system mapping calibration and target matching. Among them, the known operating units refer to construction machinery, fixed work teams, or on-site workers that have been included in the site safety management and control system and are equipped with high-precision positioning terminals. Their location data can be continuously reported through systems such as Beidou RTK and UWB indoor positioning at a fixed sampling frequency.

[0031] The first coordinate queue is the coordinate sequence of the known work unit in the entity coordinate system (construction local coordinate system or geodetic coordinate system), containing multiple first coordinates arranged in ascending order of time nodes, each carrying a unified timestamp of the corresponding sampling time. The first image queue is a continuous video frame sequence collected by fixed monitoring cameras deployed at key points in the construction site. The camera's acquisition frame rate and the sampling frequency of the positioning terminal are strictly time-synchronized to ensure that each time node corresponds to a first image frame and corresponds one-to-one with the first coordinates, achieving precise alignment in the spatiotemporal dimensions.

[0032] For example, a registered asphalt paver is a known operating unit. Its vehicle-mounted Beidou RTK terminal sampling frequency is set to 1Hz, and the corresponding site monitoring camera synchronously collects images of the operating area at a frame rate of 1 frame per second. After continuous collection for 60 seconds, a first coordinate queue containing 60 sets of planar coordinates and a first image queue containing 60 frames of monitoring images can be obtained. Each set of coordinates and images corresponds to a unified timestamp, providing a reliable matching benchmark for subsequent coordinate mapping calibration.

[0033] In step 102, the multiple first images in the first image queue are standardized to obtain the second image queue.

[0034] In some implementations, the standardization process of multiple first images in the first image queue to obtain a second image queue includes: Multiple first images in the first image queue are converted to grayscale to obtain multiple first grayscale images; Extract the mean and standard deviation of each first grayscale image and construct a grayscale array; Based on multiple grayscale arrays, the inter-frame differences of the multiple first grayscale images are verified to obtain the inter-frame difference index; If the inter-frame difference index is higher than the index threshold, then each first grayscale image is adjusted according to the plurality of grayscale arrays; Otherwise, the plurality of first grayscale images are arranged in order of time nodes to obtain the second image queue.

[0035] In some implementations, adjusting each first grayscale image based on the plurality of grayscale arrays includes: Extract multiple grayscale means and multiple standard deviations from the multiple grayscale arrays; Based on the multiple grayscale means and the multiple standard deviations, determine the target mean and the target standard deviation; For each first grayscale image, perform the following steps: Based on the standard deviation of the first grayscale image and the target standard deviation, the pixel values ​​of the first grayscale image are scaled to obtain the first intermediate grayscale image. Pixel value compensation is performed based on the mean of the first intermediate grayscale image and the target mean to obtain the adjusted first grayscale image.

[0036] For example, the lighting conditions in the construction site are easily affected by factors such as changes in day and night, weather fluctuations, and dust obstruction, resulting in significant differences in the grayscale distribution of different frames of images. Directly using these images for trajectory extraction would introduce a large amount of noise and reduce detection accuracy.

[0037] This invention unifies the grayscale distribution between frames through grayscale normalization, thereby improving the stability of subsequent trajectory analysis. The specific implementation process is as follows: Grayscale processing: Perform grayscale processing on each frame of RGB color image in the first image queue, for example, by using weighted grayscale conversion, and using a grayscale calculation formula that matches the visual sensitivity of the human eye. In the above formula, , , These represent the red, green, and blue pixel values ​​of the color image. Grayscale conversion reduces the three-channel data to a single channel, significantly reducing subsequent computation while preserving the image's brightness characteristics, ultimately yielding multiple first grayscale images that correspond one-to-one with the first image.

[0038] Gray-level feature extraction: Calculate the gray-level mean and standard deviation of all pixel values ​​in each first gray-level image frame by frame. The gray-level mean represents the overall brightness level of a single frame image, and the gray-level standard deviation represents the dispersion of pixel values; together, they constitute the gray-level distribution characteristics of a single frame image. Using the [gray-level mean, gray-level standard deviation] of each frame as feature units, arrange them in chronological order to construct a gray-level array.

[0039] Gray-level difference verification: Based on the gray-level arrays of all frames, calculate the inter-frame difference index of the gray-level distribution. The inter-frame difference index can be calculated by weighting the gray-level mean dispersion coefficient and the standard deviation dispersion coefficient. The lower the index, the closer the gray-level distribution of each frame image is, and the smaller the illumination fluctuation. The system presets an index threshold (e.g., a typical value of 0.15) to determine whether to perform gray-level normalization adjustment. If the inter-frame difference index is lower than or equal to the index threshold, it indicates that the overall illumination of the field area is stable and the inter-frame difference is small. At this time, the output is the second image queue. If the inter-frame difference index is higher than the index threshold, it indicates that a significant change in illumination has occurred in the field (such as strong dust storms or clouds blocking the sun), and the image needs to be standardized.

[0040] Grayscale image standardization adjustment: When the inter-frame differences do not meet the requirements, first extract the mean and standard deviation of all grayscale values ​​from all grayscale arrays, and then take the median as the target mean and target standard deviation. Using the median can avoid the interference of individual abnormal frames on the overall standard and improve the robustness of the algorithm.

[0041] Then, two adjustments are performed frame-by-frame for each first grayscale image: The first step, standard deviation scaling: Subtract the mean gray level of the frame from each pixel value of the first grayscale image, and then multiply by (target standard deviation / standard deviation of the frame) to normalize and scale the grayscale dispersion, obtaining the first intermediate grayscale image. The corresponding formula is: in, These are the pixel values ​​of the original grayscale image. , The first The grayscale mean and standard deviation of the frame. The target standard deviation is denoted as .

[0042] The second step is mean compensation: The target mean is superimposed on all pixel values ​​of the first intermediate grayscale image to compensate for the brightness level shift, resulting in the adjusted first grayscale image. The corresponding formula is: in, The target mean.

[0043] For example, the grayscale mean of a first grayscale image in a certain frame is 120, and the standard deviation is 30. The calculated target mean is 110, and the target standard deviation is 25. For a pixel with an original pixel value of 150, the value after standard deviation scaling is (150-120)×(25 / 30)=25. After mean compensation, the final value is 25+110=135. Through this adjustment, the grayscale distribution of all frames will be unified to the same baseline, effectively eliminating the interference of illumination fluctuations on subsequent trajectory detection.

[0044] In step 103, the movement trajectory of the known work unit in the image coordinate system is extracted from the second image queue, and a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system is constructed based on the movement trajectory and the first coordinate queue.

[0045] In some implementations, extracting the movement trajectory of the known work unit in the image coordinate system from the second image queue includes: Each image in the second image queue is segmented according to a preset segmentation template; Image data blocks corresponding to the same segmentation position are constructed into a data block array; Perform trajectory trace analysis on each data block array, and use the data block array containing trajectory traces as the trajectory array; Extract the time nodes corresponding to the image blocks containing the trajectory from the trajectory array, and use them as trajectory time nodes; Determine the coordinates of the trajectory image based on the segmentation positions corresponding to the trajectory array; The trajectory image coordinates are arranged according to the order of the trajectory time nodes to obtain the movement trajectory.

[0046] In some implementations, the step of performing trajectory trace analysis on each data block array, using the data block array containing trajectory traces as the trajectory array, includes: Perform the following steps for each data block array: Arrange the image data blocks in the data block array according to the corresponding time nodes to obtain the image block arrangement; Subtract two adjacent image data blocks to obtain multiple difference blocks; Calculate the sum of the absolute values ​​of multiple values ​​within each difference block, and use this as the block difference; Calculate the mean difference between multiple block differences; If there is a block difference value that deviates from the mean difference value by more than the deviation threshold, then the data block array is used as the trajectory array.

[0047] For example, traditional camera calibration requires manual placement of calibration targets, which is costly to maintain and has poor adaptability in dynamic construction scenarios. This invention automatically completes coordinate system mapping calibration using known real-time positioning data and visual trajectories of the work units, requiring no manual intervention and adapting to changes in the monitoring perspective of the construction site.

[0048] Movement trajectory extraction: like Figure 2 As shown, this invention uses a block trajectory trace analysis method to extract movement trajectories, transforming global detection into local block change detection, balancing detection accuracy and computational efficiency. The specific process is as follows: Image segmentation: A preset segmentation template is used to evenly divide each frame image 201 in the second image queue into M×N grid-like image data blocks (e.g., 32×32 grid). Each data block corresponds to a fixed segmentation position in the image.

[0049] Constructing a data block array: Image data blocks at the same segmentation position in all image frames are combined in time order to form a data block array 202. Each data block array 202 corresponds to a fixed spatial position in the image, and the elements in the array are the pixel data of the image block at that position at different times.

[0050] Trajectory trace determination: Its core principle is that when a moving target passes through a certain grid area, the pixel value of that area will change continuously and significantly 203, and the difference between adjacent frames will show obvious abrupt changes; while in areas where no target has passed through, the difference between adjacent frames is mainly composed of random noise, and the value is stable and close to zero.

[0051] Perform the following analysis on each data block array: Sort the image data blocks in the array according to the time node to obtain the image block arrangement; Subtract the corresponding pixels of every two adjacent image data blocks in the arrangement to generate multiple difference blocks; Calculate the sum of the absolute values ​​of all pixel differences within each difference block, which is taken as the block difference value of that difference block. The magnitude of the block difference value reflects the degree of pixel change in that region at that moment. Calculate the mean difference of all block differences as the baseline level of background noise; Enumerate all block differences one by one. If the deviation between a block difference and the mean difference is greater than a preset deviation threshold, it is determined that there is a trajectory trace in the data block array and it is marked as a trajectory array.

[0052] The deviation threshold can be adaptively set to 3 times the mean difference based on the 3σ criterion, which can filter out 99.7% of random noise and significantly reduce the probability of false detection.

[0053] Trajectory generation: Extract the time nodes corresponding to the differences in abnormal blocks in the trajectory array as trajectory time nodes; use the grid center coordinates of the corresponding segmentation position in the trajectory array as the trajectory image coordinates of the trajectory point; arrange all trajectory image coordinates in sequence according to the order of trajectory time nodes to obtain the continuous movement trajectory of the known work unit in the image coordinate system.

[0054] For example, a construction vehicle moves at a constant speed from the lower left corner to the upper right corner of the image, passing through grid blocks (5,2), (6,3), and (7,4) in sequence. The data block arrays of the corresponding three grids all detect block differences exceeding the threshold, with the corresponding time nodes being t1, t2, and t3, respectively. By sorting the center image coordinates of the three grids by time, the vehicle's movement trajectory consisting of three trajectory points can be obtained.

[0055] Construction of the first coordinate mapping table: The trajectory points in the image coordinate system are paired one by one with the entity coordinates of the same time node in the first coordinate queue to obtain multiple sets of corresponding point pairs of "image coordinates - entity coordinates". These corresponding point pairs constitute the first coordinate mapping table.

[0056] Furthermore, the perspective transformation matrix (homography matrix) from the image coordinate system to the entity coordinate system can be solved based on this set of point pairs. The image coordinate intervals corresponding to the construction site area are then discretized, and the entity coordinates corresponding to each discrete image coordinate are pre-calculated and stored as a lookup table in the form of a first coordinate mapping table. Compared to real-time matrix operations, the mapping table lookup method has lower computational latency and can meet the high-performance requirements of real-time early warning.

[0057] In step 104, non-known work units are identified from the video images, first entity coordinates are obtained based on the image coordinates of the non-known work units and the first coordinate mapping table, and warnings are issued to the work units based on the first entity coordinates.

[0058] In some implementations, identifying unknown work units from video images and obtaining the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table includes: Identify multiple first moving targets from video images; Based on the number of the plurality of first moving targets and the plurality of second moving targets, it is determined whether there is an unknown work unit, wherein the second moving targets are targets determined based on the positioning system, the targets determined based on the positioning system are known work units, and each second moving target corresponds to a second entity coordinate; If there are unknown work units, perform the following steps: For each first moving target, extract the coordinates in the image as the first coordinates; For each first coordinate, the third entity coordinate is obtained by mapping and transformation through the first coordinate mapping table; For each second entity coordinate, find the nearest coordinate from multiple third entity coordinates and use it as the matching coordinate; Use the coordinates of the third entity that has not yet been matched as the coordinates of the first entity.

[0059] In some implementations, the step of issuing a warning to the work unit based on the first entity coordinates includes: Based on the coordinates of the first entity and the speed and direction of the unknown work unit, plan a predicted trajectory line; The coordinates of a fourth entity that is less than a first distance threshold from the predicted trajectory line are used as warning targets, wherein the coordinates of the fourth entity are the coordinates of a known work unit, and the entity coordinates of the known work unit are determined based on the positioning system. A warning message is sent to the warning target indicating that an unknown work unit is approaching.

[0060] For example, the core early warning step of the method of the present invention uses the logic of "full visual recognition + matching of known targets" to filter out unregistered and unknown work units, and realizes proactive early warning based on motion trend prediction.

[0061] Identification of Unknown Work Units and Coordinate Calculation: All known work units are equipped with positioning terminals, allowing them to obtain their physical coordinates in real time via the positioning system; video images can capture all moving targets within the site area. When the number of visually identified targets exceeds the number reported by positioning, the excess targets are unknown work units without positioning terminals, such as private vehicles and unregistered temporary construction workers, posing a significant safety hazard to the construction site. The specific implementation process is as follows: Visual target detection: Target detection algorithms (such as the YOLO series) are used to infer real-time video images to identify all personnel, construction vehicles, and social vehicles as the first moving targets in the site area. The center point coordinates of each target detection box are extracted as its first coordinates in the image coordinate system.

[0062] Existence determination: Count the total number of first moving targets, and simultaneously obtain the number of all second moving targets (i.e., known work units) reported by the positioning system, along with their corresponding coordinates. If the number of first moving targets is greater than the number of second moving targets, then it is determined that there are unknown work units within the site.

[0063] Coordinate transformation and matching filtering: The first coordinates of all first moving targets are transformed by looking up the first coordinate mapping table to obtain the corresponding third entity coordinates; Using the coordinates of each second entity (the known actual positioning coordinates of the work unit) as a reference, the spatial distance between it and all third entity coordinates is calculated using Euclidean distance. The third entity coordinate with the smallest distance is selected as the matching coordinate of the known work unit. To avoid mismatches, a maximum matching distance threshold can be set. If the distance exceeds the threshold, it is determined to be abnormal positioning data and is not included in the matching process.

[0064] After matching all known work units, the coordinates of the remaining unmatched third entity are the coordinates of the first entity corresponding to the unknown work unit.

[0065] Security alert execution: For identified unknown work units, the system performs risk assessment based on movement trend prediction and sends alerts to threatened known work units. The specific process is as follows: Motion trajectory prediction: Based on the coordinates of the first entity in multiple consecutive frames of an unknown work unit, the Kalman filter algorithm is used to estimate its current instantaneous velocity and direction of motion. Based on this, a predicted trajectory line for a future preset duration (typically 10 seconds) is generated by extrapolation, representing the short-term motion trend of the target.

[0066] Warning target selection: Traverse the fourth entity coordinates (real-time positioning coordinates) of all known work units, calculate the vertical distance from each coordinate to the predicted trajectory line; mark known work units whose distance is less than the first distance threshold as warning targets. The first distance threshold can be dynamically configured according to the risk level of the construction area. For example, it can be set to 8 meters for high-risk areas such as the perimeter of deep foundation pits and hoisting operation areas, and 5 meters for ordinary operation areas, to achieve graded warning.

[0067] Warning message push: Send warning messages to the terminal devices corresponding to the warning target (such as vehicle-mounted warning terminals, smart safety helmets for workers, and handheld terminals for on-site safety officers). The message content includes key information such as the current location, relative distance, estimated arrival time, and direction of movement of the unknown work unit, prompting on-site workers to take timely avoidance or control measures.

[0068] The present invention provides an implementation method for identifying traffic construction safety risks. First, a first coordinate queue and a first image queue are acquired. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted by time nodes. Each first coordinate corresponds to a first image in the first image queue. Then, the multiple first images in the first image queue are standardized to obtain a second image queue. Next, the movement trajectory of the known work unit in the image coordinate system is extracted from the second image queue, and a first coordinate mapping table is constructed based on the movement trajectory and the first coordinate queue to transform the image coordinate system into the entity coordinate system. Finally, non-known work units are identified from video images, and first entity coordinates are obtained based on the image coordinates of the non-known work units and the first coordinate mapping table. An early warning is then issued to the work units based on the first entity coordinates.

[0069] This invention is designed for open-air construction scenarios in the transportation industry. It addresses the characteristics of wide site areas, high mobility of personnel and vehicles, complex and variable working conditions, and numerous blind spots in traditional manual management, and has multiple significant beneficial effects.

[0070] This invention relies on known work units equipped with positioning terminals to simultaneously collect entity coordinates and visual trajectories to automatically construct a coordinate mapping table. Camera spatial calibration can be completed without the need for manual deployment of calibration targets, significantly reducing on-site deployment and maintenance costs. It can also adapt to dynamic scenarios such as changes in monitoring perspective and site movement during construction.

[0071] This invention effectively counteracts image illumination fluctuations caused by daylight, dust obstruction, and changes in weather by using grayscale inter-frame difference verification and adaptive standardization processing. This significantly improves the robustness of target detection and trajectory extraction under complex working conditions and reduces false detections and missed detections caused by environmental interference.

[0072] This invention employs a segmented difference trajectory trace analysis method, which transforms global computation into local segmented processing, thereby reducing computational overhead while ensuring detection accuracy and adapting to real-time operation of edge devices.

[0073] This invention integrates full visual recognition with location registration data to accurately identify unknown work units without location terminals, filling the regulatory blind spots of the traditional management and control system; combined with motion trajectory prediction, it enables early warning of risks, shifting safety management from post-event handling to pre-event prevention, effectively reducing the risk of site collisions and unauthorized entry accidents, and comprehensively improving the efficiency of intelligent management and control of construction safety.

[0074] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0075] The following are embodiments of the apparatus of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0076] Figure 3 This is a functional block diagram of the traffic construction safety risk identification device provided in the embodiments of the present invention, with reference to... Figure 3 The traffic construction safety risk identification device includes: a work unit coordinate acquisition module 301, an image standardization processing module 302, a coordinate mapping module 303, and a risk warning module 304, wherein: The work unit coordinate acquisition module 301 is used to acquire a first coordinate queue and a first image queue. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted according to time nodes. Each first coordinate corresponds to a first image in the first image queue. Image standardization processing module 302 is used to standardize multiple first images in the first image queue to obtain a second image queue; The coordinate mapping module 303 is used to extract the movement trajectory of the known work unit in the image coordinate system from the second image queue, and construct a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system based on the movement trajectory and the first coordinate queue. The risk warning module 304 is used to identify unknown work units from video images, obtain the first entity coordinates based on the image coordinates of the unknown work units and the first coordinate mapping table, and issue a warning to the work units based on the first entity coordinates.

[0077] Figure 4 This is a functional block diagram of the electronic device provided in an embodiment of the present invention. For example... Figure 4 As shown, the electronic device 4 in this embodiment includes a processor 400 and a memory 401, wherein the memory 401 stores a computer program 402 that can run on the processor 400. When the processor 400 executes the computer program 402, it implements the steps in the various traffic construction safety risk identification methods and embodiments described above, for example... Figure 1 Steps 101 to 104 are shown.

[0078] For example, the computer program 402 may be divided into one or more modules / units, which are stored in the memory 401 and executed by the processor 400 to complete the present invention.

[0079] The electronic device 4 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The electronic device 4 may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 4 and does not constitute a limitation on electronic device 4. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 4 may also include input / output devices, network access devices, buses, etc.

[0080] The processor 400 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0081] The memory 401 can be an internal storage unit of the electronic device 4, such as a hard disk or memory. The memory 401 can also be an external storage device of the electronic device 4, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 401 can include both internal and external storage units of the electronic device 4. The memory 401 is used to store the computer program 402 and other programs and data required by the electronic device 4. The memory 401 can also be used to temporarily store data that has been output or will be output.

[0082] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0083] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0084] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0085] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0086] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0087] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0088] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various methods and apparatus embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0089] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for identifying safety risks during traffic construction, characterized in that, include: Obtain a first coordinate queue and a first image queue, wherein the first coordinate queue corresponds to a known job unit, the first coordinate queue includes multiple first coordinates sorted according to time nodes, and each first coordinate corresponds to a first image in the first image queue; The first images in the first image queue are standardized to obtain the second image queue. Extract the movement trajectory of the known work unit in the image coordinate system from the second image queue, and construct a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system based on the movement trajectory and the first coordinate queue; Identify unknown work units from video images, obtain the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table, and issue an early warning to the work units based on the first entity coordinates.

2. The method for identifying traffic construction safety risks according to claim 1, characterized in that, The step of standardizing multiple first images in the first image queue to obtain a second image queue includes: Multiple first images in the first image queue are converted to grayscale to obtain multiple first grayscale images; Extract the mean and standard deviation of each first grayscale image and construct a grayscale array; Based on multiple grayscale arrays, the inter-frame differences of the multiple first grayscale images are verified to obtain the inter-frame difference index; If the inter-frame difference index is higher than the index threshold, then each first grayscale image is adjusted according to the plurality of grayscale arrays; Otherwise, the plurality of first grayscale images are arranged in order of time nodes to obtain the second image queue.

3. The method for identifying traffic construction safety risks according to claim 2, characterized in that, The step of adjusting each first grayscale image according to the plurality of grayscale arrays includes: Extract multiple grayscale means and multiple standard deviations from the multiple grayscale arrays; Based on the multiple grayscale means and the multiple standard deviations, determine the target mean and the target standard deviation; For each first grayscale image, perform the following steps: Based on the standard deviation of the first grayscale image and the target standard deviation, the pixel values ​​of the first grayscale image are scaled to obtain the first intermediate grayscale image. Pixel value compensation is performed based on the mean of the first intermediate grayscale image and the target mean to obtain the adjusted first grayscale image.

4. The method for identifying traffic construction safety risks according to claim 1, characterized in that, Extracting the movement trajectory of the known work unit in the image coordinate system from the second image queue includes: Each image in the second image queue is segmented according to a preset segmentation template; Image data blocks corresponding to the same segmentation position are constructed into a data block array; Perform trajectory trace analysis on each data block array, and use the data block array containing trajectory traces as the trajectory array; Extract the time nodes corresponding to the image blocks containing the trajectory from the trajectory array, and use them as trajectory time nodes; Determine the coordinates of the trajectory image based on the segmentation positions corresponding to the trajectory array; The trajectory image coordinates are arranged according to the order of the trajectory time nodes to obtain the movement trajectory.

5. The method for identifying traffic construction safety risks according to claim 4, characterized in that, The step of performing trajectory trace analysis on each data block array, using the data block array containing trajectory traces as the trajectory array, includes: Perform the following steps for each data block array: Arrange the image data blocks in the data block array according to the corresponding time nodes to obtain the image block arrangement; Subtract two adjacent image data blocks to obtain multiple difference blocks; Calculate the sum of the absolute values ​​of multiple values ​​within each difference block, and use this as the block difference; Calculate the mean difference between multiple block differences; If there is a block difference value that deviates from the mean difference value by more than the deviation threshold, then the data block array is used as the trajectory array.

6. The method for identifying traffic construction safety risks according to any one of claims 1-5, characterized in that, The step of identifying unknown work units from video images and obtaining the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table includes: Identify multiple first moving targets from video images; Based on the number of the plurality of first moving targets and the plurality of second moving targets, it is determined whether there is an unknown work unit, wherein the second moving targets are targets determined based on the positioning system, the targets determined based on the positioning system are known work units, and each second moving target corresponds to a second entity coordinate; If there are unknown work units, perform the following steps: For each first moving target, extract the coordinates in the image as the first coordinates; For each first coordinate, the third entity coordinate is obtained by mapping and transformation through the first coordinate mapping table; For each second entity coordinate, find the nearest coordinate from multiple third entity coordinates and use it as the matching coordinate; Use the coordinates of the third entity that has not yet been matched as the coordinates of the first entity.

7. The method for identifying traffic construction safety risks according to claim 6, characterized in that, The step of issuing an early warning to the work unit based on the first entity coordinates includes: Based on the coordinates of the first entity and the speed and direction of the unknown work unit, plan a predicted trajectory line; The coordinates of a fourth entity that is less than a first distance threshold from the predicted trajectory line are used as warning targets, wherein the coordinates of the fourth entity are the coordinates of a known work unit, and the entity coordinates of the known work unit are determined based on the positioning system. A warning message is sent to the warning target indicating that an unknown work unit is approaching.

8. A traffic construction safety risk identification device, characterized in that, For implementing the traffic construction safety risk identification method as described in any one of claims 1-7, the traffic construction safety risk identification device comprises: The work unit coordinate acquisition module is used to acquire a first coordinate queue and a first image queue. The first coordinate queue corresponds to a known work unit and includes multiple first coordinates sorted by time nodes. Each first coordinate corresponds to a first image in the first image queue. An image normalization processing module is used to normalize multiple first images in the first image queue to obtain a second image queue. The coordinate mapping module is used to extract the movement trajectory of the known work unit in the image coordinate system from the second image queue, and construct a first coordinate mapping table for transforming the image coordinate system to the entity coordinate system based on the movement trajectory and the first coordinate queue. as well as, The risk warning module is used to identify unknown work units from video images, obtain the coordinates of a first entity based on the image coordinates of the unknown work units and the first coordinate mapping table, and issue a warning to the work units based on the first entity coordinates.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7 above.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7 above.

Images