Working site safety distance early warning method and device, medium and computer equipment

Abstract

The invention discloses a working site safety distance early warning method and device, a medium and computer equipment, and the method comprises the steps: obtaining an original depth image containing working personnel and working equipment in real time, carrying out the denoising processing, obtaining a processed depth image, and recognizing and segmenting a working personnel region and a working equipment region; according to parameters of the depth camera, pixel coordinates of all pixel points in the processed depth image are converted into three-dimensional space coordinates, and the distance between the operating personnel and the operating equipment is calculated in real time based on the three-dimensional space coordinates of all the pixel points corresponding to the operating personnel and the operating equipment; and the alarm bell is dynamically adjusted for warning according to the comparison condition of the distance calculated in real time and the preset safety distance standard. Analysis and early warning are carried out in combination with a safety distance rule, so that the safety distance problem between the operating personnel and the operating equipment can be found in time, namely, key safety distance evaluation and control capability can be provided for a safety prevention and control system through early warning of potential risks.

Description

Technical Field

[0001] The present application relates to the fields of safety monitoring technology and computer vision technology, and in particular to a method and device, medium, and computer equipment for early warning of safety distance at a work site. Background Art

[0002] In the monitoring scenario of dangerous operations in power systems, real-time perception of the spatial distance between operators and operating equipment is a core requirement for ensuring personnel safety, optimizing equipment scheduling, and preventing collision accidents.

[0003] Traditional monitoring methods (such as manual inspections and fixed sensors) have problems such as large response delays, limited coverage, and poor environmental adaptability, making it difficult to meet the safety needs in complex dynamic operation scenarios. Summary of the Invention

[0004] In view of this, the present application provides a work site safety distance warning method and device, medium, and computer equipment, which combines safety distance rules for analysis and warning, timely discovers safety distance problems between workers and work equipment, and warns of potential risks, which can provide key safety distance assessment and control capabilities for the entire safety prevention and control system.

[0005] According to one aspect of the present application, a method for early warning of safety distance at a work site is provided, the method comprising:

[0006] Acquire a raw depth image containing workers and equipment in real time, wherein the raw depth image is acquired by a depth camera installed at the work site, and the pixel value in the raw depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to depth camera parameters;

[0007] Denoising the original depth image to obtain a processed depth image, and identifying and segmenting the operator area and the operating equipment area in the processed depth image;

[0008] After converting the pixel coordinates of each pixel point in the processed depth image into three-dimensional spatial coordinates based on the depth camera parameters, the three-dimensional spatial coordinates of each pixel point corresponding to the operator are determined based on the operator area, and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment are determined based on the operating equipment area;

[0009] Based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, the distance between the operator and the operating equipment is calculated in real time, and the alarm is dynamically adjusted to issue a warning based on the comparison between the real-time calculated distance and the preset safety distance standard.

[0010] According to another aspect of the present application, a work site safety distance warning device is provided, the device comprising:

[0011] An image acquisition module is used to acquire, in real time, an original depth image containing the operator and the operating equipment, wherein the original depth image is acquired by a depth camera installed at the operating site, and the pixel value in the original depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to a depth camera parameter;

[0012] An image processing module is used to perform denoising on the original depth image to obtain a processed depth image, and to identify and segment the operator area and the operating equipment area in the processed depth image;

[0013] A distance calculation module is used to convert the pixel coordinates of each pixel point in the processed depth image into three-dimensional spatial coordinates based on the depth camera parameters, and then determine the three-dimensional spatial coordinates of each pixel point corresponding to the operator based on the operator area, and determine the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment based on the operating equipment area;

[0014] The safety warning module is used to calculate the distance between the operator and the operating equipment in real time based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, and dynamically adjust the alarm to issue a warning based on the comparison between the real-time calculated distance and the preset safety distance standard.

[0015] According to another aspect of the present application, a medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned work site safety distance warning method is implemented.

[0016] According to another aspect of the present application, a computer device is provided, including a medium, a processor, and a computer program stored on the medium and executable on the processor, wherein the processor implements the above-mentioned work site safety distance warning method when executing the program.

[0017] By means of the above technical solution, the present application provides a method and device, medium, and computer equipment for early warning of safety distance at a work site, which acquires the original depth image containing the workers and the work equipment in real time, performs denoising processing, obtains the processed depth image, identifies and segments the worker area and the work equipment area; according to the depth camera parameters, converts the pixel coordinates of each pixel point in the processed depth image into three-dimensional space coordinates, calculates the distance between the workers and the work equipment in real time based on the three-dimensional space coordinates of each pixel point corresponding to the workers and the three-dimensional space coordinates of each pixel point corresponding to the work equipment, and dynamically adjusts the alarm to issue a warning based on the comparison between the distance calculated in real time and the preset safety distance standard. Combined with the safety distance rules for analysis and early warning, the safety distance problem between the workers and the work equipment can be discovered in a timely manner, and potential risks can be warned, providing key safety distance assessment and control capabilities for the entire safety prevention and control system.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 A schematic diagram of a process flow of a work site safety distance warning method provided by an embodiment of the present application is shown;

[0021] Figure 2 A schematic diagram of a process for another work site safety distance warning method provided by an embodiment of the present application is shown;

[0022] Figure 3 A schematic structural diagram of a work site safety distance warning device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that, unless there is a conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0024] In this embodiment, a method for early warning of safety distance at a work site is provided. Figure 1 As shown, the method includes:

[0025] Step 101: Acquire an original depth image containing workers and work equipment in real time, wherein the original depth image is acquired by a depth camera installed at the work site, and the pixel value in the original depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to depth camera parameters.

[0026] In the above embodiments of the present application, at the power operation site (such as substation maintenance, transmission line maintenance, distribution room operation and other scenarios), real-time acquisition of three-dimensional spatial information of operators and operating equipment through depth cameras (such as ToF cameras, structured light cameras, binocular stereo vision cameras, etc.) is the core link to ensure personnel safety, optimize operation processes, and realize intelligent monitoring. The original depth image is obtained in real time by the depth camera installed at the operation site, and after the original depth image is analyzed and processed, the processed depth image can provide spatial distance information between the operator and the operating equipment, providing basic data for subsequent safety distance assessment and control.

[0027] In particular, a depth image is a two-dimensional image acquired actively or passively. Each pixel value in the depth image directly reflects the physical distance (in millimeters or meters) from the corresponding point in the scene to the depth camera. Compared with traditional RGB color images, depth images have the following advantages:

[0028] 1. Directly provide spatial information: Without complex 3D reconstruction algorithms, the spatial position of objects in the scene can be obtained, that is, the pixel value is the physical distance (for example, a pixel value of 2.3 meters means that the point is 2.3 meters away from the camera), and spatial relationships can be calculated without complex algorithms.

[0029] 2. Strong anti-interference ability: It is insensitive to lighting changes and color differences, suitable for complex industrial environments such as low light and high reflection. It is insensitive to strong electromagnetic interference and lighting changes (such as night operations) in power scenes, and its stability is better than RGB cameras.

[0030] In power operation sites, for example, live maintenance scenarios at substations require operators to operate near high-voltage equipment (such as circuit breakers and disconnectors). Real-time monitoring is required to ensure the distance between operators and live objects is ≥1 meter, a safety threshold. Examples of operating equipment include robotic arms, insulated boom trucks, and mobile maintenance platforms.

[0031] For the depth image acquisition process, first start the depth camera, set the resolution (such as 640×480), frame rate (30FPS), and exposure time (adjusted according to ambient light). Next, use the depth camera to acquire a depth image (16-bit grayscale image, pixel value unit is millimeter). In particular, the depth image is a single-channel 16-bit unsigned integer with a pixel value range of:

[0032] 0: Invalid point (such as obstruction, out of range).

[0033] 1-65535: Actual distance (unit: mm, such as 2300 means 2.3 meters).

[0034] The depth camera can also obtain the following parameters through calibration to achieve the conversion from pixel coordinates to three-dimensional space coordinates, for example:

[0035] Internal reference:

[0036] Focal length (f x ,f y ), measured in pixels, is used to describe the relationship between lens focal length and pixel size.

[0037] Main point (c x ,c y ): Image center coordinates (pixel units).

[0038] Calibration method: Use a checkerboard calibration plate and calculate through OpenCV's calibrateCamera function.

[0039] External reference:

[0040] Rotation matrix (R) and translation vector (T): used to describe the relative position of the camera coordinate system and the world coordinate system.

[0041] Calibration method: Place calibration objects with known spatial coordinates (such as reflective marker balls) at the work site and solve them using the PnP algorithm.

[0042] Step 102 , denoising the original depth image to obtain a processed depth image, and identifying and segmenting the operator area and the operating equipment area in the processed depth image.

[0043] Next, because the depth camera is susceptible to environmental interference (such as electromagnetic noise, glass / metal reflection, and multipath reflection), the original depth image will have isolated noise points (such as flying spots), depth holes (such as no value in the occluded area), and jagged edges. Through noise suppression, that is, through bilateral filtering, joint bilateral filtering, or depth-guided filtering, the noise can be smoothed while retaining edge details (such as correcting the flying spot depth value to the neighborhood mean), and the processed depth image can be obtained, which can improve the accuracy of subsequent distance calculations. In particular, "hole filling" can also be performed, that is, the missing areas are completed using temporal consistency or depth interpolation algorithms (such as the fast marching method FMM).

[0044] Alternatively, as Figure 2 As shown, in step 102, "denoising the original depth image to obtain a processed depth image, and identifying and segmenting the operator area and the operating equipment area in the processed depth image" specifically includes:

[0045] Step 1021 , performing denoising processing on the original depth image using a bilateral filtering method to obtain a processed depth image.

[0046] Step 1022 : Determine a personnel-equipment distinction threshold in the processed depth image using a maximum inter-class variance algorithm, wherein the personnel-equipment distinction threshold is used to distinguish between a working personnel area and a working equipment area.

[0047] Step 1023 : Binarize the processed depth image based on the determined personnel and equipment distinction threshold to obtain a depth image to be identified, and determine the operating personnel area and the operating equipment area in the depth image to be identified.

[0048] In the above embodiment of the present application, the original depth image can be denoised using a bilateral filtering method, which can improve the subsequent recognition accuracy. Bilateral filtering is a nonlinear filtering method that combines spatial proximity and pixel value similarity, and can well preserve edge information while denoising. The bilateral filter is based on Gaussian filtering, but on this basis, it adds consideration of pixel value similarity, and its filtering weight is jointly determined by the spatial proximity factor and the brightness similarity factor. The spatial proximity factor ensures that the closer the pixel is to the current pixel, the greater its contribution to the filtering result, while the brightness similarity factor ensures that the closer the pixel is to the current pixel brightness, the greater its contribution to the filtering result. This dual weight mechanism enables bilateral filtering to effectively retain the edge information of the image while denoising, specifically:

[0049] You can use an image processing library (such as OpenCV) to read the raw depth image, ensuring that the image data is loaded in a suitable format (such as 16-bit unsigned integers) to accurately represent the depth information.

[0050] Next, the parameters of the bilateral filter are defined, including the spatial proximity factor (σspace), the brightness similarity factor (σcolor), and the diameter of the filter (d). The choice of these parameters will directly affect the balance between denoising effect and edge preservation.

[0051] The spatial proximity factor (σspace) is used to control the effect of spatial proximity on the filter weight. A larger σspace value means that pixels farther away will also have a greater impact on the filter result, which may cause blurred edges.

[0052] The brightness similarity factor (σcolor) controls the influence of pixel value similarity on the filter weight. A larger σ_color value means that pixels with a larger brightness difference from the current pixel will also have a greater impact on the filter result, which may reduce the denoising effect.

[0053] The filter diameter (d) defines the size of the filter, that is, the range of neighboring pixels considered. Larger d values increase the computational effort but may improve denoising results.

[0054] Next, use the bilateral filter function in the image processing library (such as cv2.bilateralFilter in OpenCV) to filter the original depth image. Pass the defined parameters to the filter function and specify the data type and range of the output image.

[0055] Next, save the bilaterally filtered image data as a processed depth image file. Ensure that the data type and range of the output image are consistent with the input image for subsequent processing and analysis.

[0056] In particular, in power operation scenarios, depth images may contain specific noise patterns (such as noise caused by electromagnetic interference or sensor noise) or key edge information that needs to be preserved (such as equipment edges or personnel outlines). Therefore, when applying bilateral filtering, the filter parameters can be adjusted according to the specific scenario to optimize the balance between denoising effect and edge preservation.

[0057] Next, the maximum inter-class variance algorithm is used for binarization processing. Specifically, this can be achieved through the inter-class variance calculation formula. That is, the inter-class variance corresponding to each pixel in the processed depth image is calculated according to the inter-class variance calculation formula, and the pixel value corresponding to the maximum inter-class variance is determined as the threshold for distinguishing between people and equipment. The inter-class variance calculation formula is:

[0058] σ k 2 =P k,人员 (m 人员 -m G ) 2 +P k,设备 (m 设备 -m G ) 2 ,

[0059] σ k is the inter-class variance of the k-th pixel in the processed depth image, P k,人员 is the probability that pixel k belongs to the operator, m 人员 is the average gray level of the operator, m G is the average grayscale of the processed depth image, P k,设备 is the probability that pixel k belongs to the operating device, m 设备 is the average grayscale of the operating equipment;

[0060] Then, based on the determined personnel and equipment distinction threshold, the processed depth image is binarized to obtain a depth image to be identified, and the personnel pixel assignment value of the personnel and the equipment pixel assignment value of the equipment are determined;

[0061] In the depth image to be identified, for any pixel point, when the pixel value of the pixel point is greater than or equal to the person-device distinction threshold, the pixel value of the pixel point is assigned to the operator pixel assigned value; when the pixel value of the pixel point is less than the person-device distinction threshold, the pixel value of the pixel point is assigned to the operator pixel assigned value;

[0062] The worker is identified based on the display area of the worker pixel assigned value, and the work equipment is identified based on the display area of the work equipment pixel assigned value.

[0063] To this end, the operator area and the operating equipment area can be determined in the depth image to be identified for subsequent distance calculation.

[0064] In step 103, the pixel coordinates of each pixel point in the processed depth image are converted into three-dimensional space coordinates according to the depth camera parameters, and the three-dimensional space coordinates of each pixel point corresponding to the operator are determined based on the operator area, and the three-dimensional space coordinates of each pixel point corresponding to the operating equipment are determined based on the operating equipment area.

[0065] Next, converting the pixel coordinates in the depth image into three-dimensional space coordinates is a core task in computer vision and robotic perception. The core is to use camera parameters to establish the projection relationship between pixel coordinates and three-dimensional space points, thereby preparing for subsequent distance calculation.

[0066] Optionally, the depth camera parameters include the focal length of the depth camera and the coordinates of the principal point of the depth camera. For step 103, "converting the pixel coordinates of each pixel point in the processed depth image into three-dimensional space coordinates according to the depth camera parameters" specifically includes:

[0067] In step 1031, each pixel in the processed depth image is projected into a point in three-dimensional space using a pinhole camera model to obtain the three-dimensional spatial coordinates of each pixel in the processed depth image. The pinhole camera model is:

[0068]

[0069] (u, v) is the pixel coordinate, (X, Y, Z) is the three-dimensional space coordinate, d is the depth value in the image depth image, the unit of the depth value is consistent with the focal length of the depth camera, s is the depth value scaling factor, (f x ,f y ) is the focal length of the depth camera, (c x,cy) is the coordinate of the principal point of the depth camera.

[0070] In the above embodiment of the present application, a pinhole camera model can be used to implement the method. In the pinhole camera model, (u, v) is the pixel coordinate (row and column number, with the origin usually in the upper left corner), (X, Y, Z) is the three-dimensional space coordinate (camera coordinate system, with the origin being the optical center), d is the depth value in the image depth image, and the unit is consistent with the focal length, such as millimeters or meters. s is the depth value scaling factor. If the depth map has been physically normalized, s = 1. (f x ,f y ) is the focal length of the depth camera, in pixels, provided by the intrinsic parameter matrix K, (c x ,cy) is the coordinate of the principal point of the depth camera, in pixel units, the intersection of the optical axis and the image plane.

[0071] In the pinhole camera model, the camera coordinate system, with its origin at the optical center and the z-axis pointing forward, points to the front of the camera. The image coordinate system, with its origin at the top-left pixel, has the u-axis pointing right and the v-axis pointing downward. The pinhole camera model formula back-projects pixel coordinates into three-dimensional space using the principle of similar triangles.

[0072] In particular, the camera intrinsic parameter matrix K can be obtained by camera calibration in the form of:

[0073]

[0074] Next, preprocess the depth value. That is, the depth map unit is millimeters, while the focal length unit is pixels. Make sure the units are consistent. For example, if fx and fy are in meters, convert d to meters.

[0075] In particular, if the depth image is not in the same unit as the focal length, the depth value can be scaled as follows:

[0076]

[0077] For example, if f x =525px, the actual focal length is 525mm, but the depth map unit is meter, so Z = d·0.001.

[0078] In step 104, based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, the distance between the operator and the operating equipment is calculated in real time, and the alarm is dynamically adjusted to issue an alert based on the comparison between the real-time calculated distance and the preset safety distance standard.

[0079] Next, in the power operation scenario, the three-dimensional spatial coordinates of the operating personnel and the operating equipment are obtained through a depth camera, and the alarm is dynamically adjusted based on real-time distance calculation to issue a warning. This is the core technology for building an active safety protection system, which aims to solve the problems of "delayed early warning", "high false alarm rate" and "low efficiency of human-machine collaboration" in traditional safety management.

[0080] The Euclidean distance d between the operator and the operating equipment is calculated in real time using three-dimensional spatial coordinates. Combined with the preset safety distance standard (e.g., the safety distance for 10kV equipment is ≥ 0.7 meters), when d approaches the preset safety distance standard, the system automatically triggers an alarm. Specifically, a three-level early warning mechanism can be established, as shown in Table 1:

[0081] Table 1

[0082]

[0083] Specifically, a multi-channel speaker array can be used to deliver targeted alerts based on the relative position of workers and equipment (e.g., only to those near the danger zone). Alternatively, a combination of voice prompts (e.g., "There is a live device 3 meters to the left") and LED strobes can be used to improve information transmission efficiency. Furthermore, a whitelist of device operations can be set. This means that if safety tools such as insulating rods and electroscopes are detected in contact with equipment, the alarm will be automatically suppressed (to avoid false alarms). Furthermore, if a person is stationary more than 1 meter from the equipment and shows no signs of moving, the alarm frequency will be reduced (to reduce interference).

[0084] Optionally, in step 104, "calculating the distance between the operator and the operating equipment in real time based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment" specifically includes:

[0085] Step 1041 traverses all point pairs between the operator and the operating equipment, calculates the Euclidean distance of each point pair in real time using the Euclidean distance calculation formula, and uses the calculated minimum Euclidean distance as the final distance. The Euclidean distance calculation formula is:

[0086]

[0087] (x1, y1, z1) are the three-dimensional spatial coordinates of the operator, and (x2, y2, z2) are the three-dimensional spatial coordinates of the operating equipment.

[0088] In the above embodiment of the present application, the coordinate point set of the operating personnel and the operating equipment is obtained, and by traversing all the point pairs between the operating personnel and the operating equipment, the Euclidean distance of each point pair is calculated in real time using the Euclidean distance calculation formula, and the calculated minimum Euclidean distance is used as the final distance to prepare for subsequent safety distance matching.

[0089] Optionally, in step 104, "dynamically adjusting the alarm to issue an alert based on the comparison between the distance calculated in real time and the preset safety distance standard" specifically includes:

[0090] Step 1042: Determine at least one classification threshold according to the preset safety standard distance, and divide at least two alarm level intervals based on the preset safety standard distance and the determined classification threshold.

[0091] Step 1043: If the currently calculated distance is within the first alarm level range, trigger the alarm to sound an alarm in a first alarm mode, wherein the first alarm mode includes outputting a periodic alarm signal at a preset first frequency.

[0092] Step 1044: If the currently calculated distance is within the second alarm level range, trigger the alarm to sound an alarm in a second alarm mode, wherein the second alarm mode includes continuously outputting an alarm signal.

[0093] In the above embodiment of the present application, for example, for a 10kV live working scenario, the following classification thresholds can be set:

[0094] Level 3 threshold (safe working distance): 70cm (electric field strength > 4kV / m). When the distance between workers and energized equipment is less than this value, there is a direct risk of electric shock and workers must immediately stop working and evacuate.

[0095] Level 2 threshold (warning distance): 200 cm (electric field strength 528 V / m). When the distance is less than this value, although the safe operating distance has not been reached, the electric field strength is already high, requiring increased monitoring and protective measures.

[0096] Level 1 threshold (approach warning distance): 400cm (electric field strength 100V / m). When the distance is less than this value, the operator is approaching energized equipment and a warning signal is issued.

[0097] Then, based on the preset safety standard distance and the determined classification threshold, two alarm level intervals are divided:

[0098] First Alarm Level Range: 0cm < Distance ≤ 70cm (within the third level threshold). This range is extremely dangerous, as workers are too close to energized equipment and are at high risk of electric shock.

[0099] Second Alarm Level: 70cm < Distance ≤ 200cm (within the second threshold). This area is a dangerous zone with high electric field strength. Operators must operate with caution and take appropriate protective measures.

[0100] Different alarm modes are triggered according to different alarm level intervals, for example:

[0101] First Alarm Mode: When the currently calculated distance falls within the first alarm level range (0cm < distance ≤ 70cm), the alarm is triggered to output a periodic alarm signal at a preset first frequency (e.g., 3 times per second). This mode is designed to quickly attract the attention of operators through high-frequency, periodic alarm sounds, prompting them to take immediate evacuation measures.

[0102] Second Alarm Mode: When the currently calculated distance is within the second alarm level range (70cm < distance ≤ 200cm), the alarm is triggered and continuously outputs an alarm signal. This mode uses continuous alarm sounding to remind operators that they have entered the danger zone and need to strengthen monitoring and take protective measures.

[0103] To this end, the distance between workers and energized equipment (operating equipment) is monitored in real time at live working sites and compared with classification thresholds. When the distance enters the different alarm level ranges, the corresponding alarm mode is automatically triggered, improving the safety of workers during work.

[0104] Optionally, the classification threshold is based on a preset proportion of a preset safety standard distance, or is determined by a fixed preset threshold. Regarding step 1042, "dividing at least two alarm level intervals based on the preset safety standard distance and the determined classification threshold" specifically includes:

[0105] Step 10421: determine the interval less than or equal to the classification threshold as the first alarm level interval, and determine the interval from greater than the classification threshold to less than or equal to the preset safety standard distance as the second alarm level interval.

[0106] In the above embodiment of the present application, the classification threshold can be determined based on the preset proportion of the preset safety standard distance. For example, the minimum safety distance (preset safety standard distance) for 10kV live working is 0.7 meters.

[0107] Next, the classification threshold is calculated:

[0108] The first-level alarm threshold (d1) can be 80% of the safety distance, that is, 0.7m×80%=0.56m, which triggers a high-frequency sound and light alarm, prompting the operator to adjust his posture immediately.

[0109] The second level alarm threshold (d2) can be 60% of the safety distance, that is, 0.7m×60%=0.42m, which triggers a forced stop command and the system automatically locks the operating authority of the working tool.

[0110] Tiered thresholds can also be determined based on fixed thresholds of equipment characteristics. For example, for high-voltage switchgear maintenance, the first-level threshold could be 0.3m (busbar phase safety distance within the switchgear), and the second-level threshold could be 0.1m (prohibited ingress insulation partition area). For ultra-high voltage converter stations, the first-level threshold could be 5m (valve tower safety distance within the valve hall), and the second-level threshold could be 2m (prohibited ingress laser protection area).

[0111] In particular, you can also set up multi-level alarms:

[0112] For the first-level warning, if the calculated spatial distance is less than the safety distance threshold but greater than 80% of the safety distance threshold (such as 0.56 meters), the alarm will flash at a frequency of 1Hz and beep intermittently.

[0113] For the second-level alarm: if the calculated spatial distance is less than 80% of the safety distance threshold (such as 0.56 meters), the alarm will continue to beep and the red flashing light will be on. By adding a timely alarm triggering mechanism, the safety of the operators during operation can be improved in real time.

[0114] To this end, through the deep integration of graded thresholds and multimodal alarms, we can achieve the transition from passive response to active defense.

[0115] Furthermore, in a specific embodiment, for example, in a live substation operation safety monitoring scenario, an operator operating an insulating rod approaches a 10kV circuit breaker, and the distance between the operator and the live object needs to be monitored in real time. A ToF camera (depth camera) is deployed on the top of the maintenance platform, covering an area with a radius of 5 meters. The depth image is then used to segment the operator and equipment areas and calculate the closest distance. If the distance is less than 1 meter, an audible and visual alarm is triggered, and the operation is stopped.

[0116] By applying the technical solution of this embodiment, an early warning is provided when the distance deviates from the safe range so that safety measures can be taken. By utilizing computer vision and image processing technology, the safe distance between operators and equipment can be quickly measured and evaluated, providing strong support for the safety management and risk control of the power system, and improving the safety and efficiency of the work site.

[0117] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides a work site safety distance warning device, such as Figure 3 As shown, the device includes:

[0118] Image acquisition module 201, used to acquire in real time an original depth image containing workers and work equipment, wherein the original depth image is acquired by a depth camera installed at the work site, and the pixel value in the original depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to depth camera parameters;

[0119] The image processing module 202 is used to perform denoising on the original depth image to obtain a processed depth image, and to identify and segment the operator area and the operating equipment area in the processed depth image;

[0120] The distance calculation module 203 is configured to convert the pixel coordinates of each pixel in the processed depth image into three-dimensional spatial coordinates based on the depth camera parameters, and then determine the three-dimensional spatial coordinates of each pixel corresponding to the operator based on the operator area, and determine the three-dimensional spatial coordinates of each pixel corresponding to the operating equipment based on the operating equipment area;

[0121] The safety warning module 204 is used to calculate the distance between the operator and the operating equipment in real time based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, and dynamically adjust the alarm to issue a warning based on the comparison between the real-time calculated distance and the preset safety distance standard.

[0122] It should be noted that for other corresponding descriptions of the functional units involved in the work site safety distance warning device provided in the embodiment of the present application, please refer to Figures 1 to 2 The corresponding description in the method will not be repeated here.

[0123] Based on the above Figures 1 to 2 The method shown in FIG. 1 is a method for performing the above-mentioned operation. Accordingly, the embodiment of the present application further provides a medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned operation is performed. Figures 1 to 2 The safety distance warning method for the work site is shown.

[0124] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0125] Based on the above Figures 1 to 2 The method shown, and Figure 3 In order to achieve the above-mentioned purpose, the embodiment of the present application further provides a computer device, which can be a personal computer, a server, a network device, etc. The computer device includes a medium and a processor; the medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figures 1 to 2 The safety distance warning method for the work site is shown.

[0126] Optionally, the computer device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a Wi-Fi interface), etc.

[0127] Those skilled in the art will understand that the computer device structure provided in this embodiment does not constitute a limitation on the computer device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0128] The medium may also include an operating system and a network communication module. An operating system is a program that manages and stores the hardware and software resources of a computer device, supporting the execution of information processing programs and other software and / or programs. The network communication module facilitates communication between components within the medium, as well as with other hardware and software within the physical device.

[0129] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by hardware to obtain the original depth image containing the operator and the operating equipment in real time, perform denoising processing, obtain the processed depth image, identify and segment the operator area and the operating equipment area; according to the depth camera parameters, the pixel coordinates of each pixel point in the processed depth image are converted into three-dimensional space coordinates, and based on the three-dimensional space coordinates of each pixel point corresponding to the operator and the three-dimensional space coordinates of each pixel point corresponding to the operating equipment, the distance between the operator and the operating equipment is calculated in real time, and according to the comparison between the distance calculated in real time and the preset safety distance standard, the alarm is dynamically adjusted to issue a warning. Combined with the safety distance rules for analysis and early warning, the safety distance problem between the operator and the equipment is discovered in a timely manner, potential risks are warned, and key safety distance assessment and control capabilities are provided for the entire safety prevention and control system.

[0130] Those skilled in the art will understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required to implement the present application. Those skilled in the art will understand that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0131] The serial numbers of the above application are for descriptive purposes only and do not represent the advantages or disadvantages of the implementation scenarios. The above disclosures are only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any changes that can be made by those skilled in the art should fall within the scope of protection of the present application.

Claims

1. A method for early warning of safe distance at a work site, characterized in that: The method comprises: Acquire a raw depth image containing workers and equipment in real time, wherein the raw depth image is acquired by a depth camera installed at the work site, and the pixel value in the raw depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to depth camera parameters; Denoising the original depth image to obtain a processed depth image, and identifying and segmenting the operator area and the operating equipment area in the processed depth image; After converting the pixel coordinates of each pixel point in the processed depth image into three-dimensional spatial coordinates based on the depth camera parameters, the three-dimensional spatial coordinates of each pixel point corresponding to the operator are determined based on the operator area, and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment are determined based on the operating equipment area; Based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, the distance between the operator and the operating equipment is calculated in real time, and the alarm is dynamically adjusted to issue a warning based on the comparison between the real-time calculated distance and the preset safety distance standard.

2. The method according to claim 1, characterized in that The method of dynamically adjusting the alarm to issue an alert based on the comparison between the real-time calculated distance and the preset safety distance standard includes: Determining at least one classification threshold according to a preset safety standard distance, and dividing at least two alarm level intervals based on the preset safety standard distance and the determined classification threshold; If the currently calculated distance is within the first alarm level range, the alarm is triggered to sound an alarm through a first alarm mode, wherein the first alarm mode includes outputting a periodic alarm signal at a preset first frequency; If the currently calculated distance is within the second alarm level range, the alarm is triggered to sound an alarm in a second alarm mode, wherein the second alarm mode includes continuously outputting an alarm signal.

3. The method according to claim 2, characterized in that The classification threshold is based on a preset proportion of a preset safety standard distance, or is determined by a fixed preset threshold. The preset safety standard distance and the determined classification threshold are used to divide at least two alarm level intervals, including: An interval less than or equal to the classification threshold is determined as a first alarm level interval, and an interval greater than the classification threshold to less than or equal to a preset safety standard distance is determined as a second alarm level interval.

4. The method according to claim 1, wherein The identifying and segmenting of the operator area and the operating equipment area in the processed depth image includes: In the processed depth image, a personnel-equipment distinction threshold is determined by a maximum inter-class variance algorithm, wherein the personnel-equipment distinction threshold is used to distinguish between a working personnel area and a working equipment area; After binarization processing is performed on the processed depth image based on the determined personnel and equipment differentiation threshold, a depth image to be identified is obtained, and an operating personnel area and an operating equipment area are determined in the depth image to be identified.

5. The method according to claim 1, wherein The depth camera parameters include the focal length of the depth camera and the coordinates of the principal point of the depth camera. The pixel coordinates of each pixel in the processed depth image are converted into three-dimensional space coordinates according to the depth camera parameters, including: The pinhole camera model is used to project each pixel point in the processed depth image into each point in the three-dimensional space to obtain the three-dimensional space coordinates of each pixel point in the processed depth image. The pinhole camera model is: (u, v) is the pixel coordinate, (X, Y, Z) is the three-dimensional space coordinate, d is the depth value in the image depth image, the unit of the depth value is consistent with the focal length of the depth camera, s is the depth value scaling factor, (f x ,f y ) is the focal length of the depth camera, (c x ,cy) is the coordinate of the principal point of the depth camera.

6. The method according to any one of claims 1 to 5, characterized in that The denoising process is performed on the original depth image to obtain a processed depth image, including: The original depth image is denoised using the bilateral filtering method to obtain a processed depth image.

7. The method according to claim 6, characterized in that The real-time calculation of the distance between the operator and the operating equipment based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment includes: Traverse all point pairs between the operator and the operating equipment, calculate the Euclidean distance of each point pair in real time using the Euclidean distance calculation formula, and use the calculated minimum Euclidean distance as the final distance. The Euclidean distance calculation formula is: d is the Euclidean distance, (x1, y1, z1) is the three-dimensional spatial coordinate of the operator, and (x2, y2, z2) is the three-dimensional spatial coordinate of the operating equipment.

8. A safety distance warning device for a work site, characterized in that: The device comprises: An image acquisition module is configured to acquire, in real time, an original depth image of the operator and the operating equipment, wherein the original depth image is acquired by a depth camera installed at the operating site, and the pixel value in the original depth image represents the physical distance from the pixel to the depth camera, and the depth camera corresponds to depth camera parameters; An image processing module is used to perform denoising on the original depth image to obtain a processed depth image, and to identify and segment the operator area and the operating equipment area in the processed depth image; A distance calculation module is used to convert the pixel coordinates of each pixel point in the processed depth image into three-dimensional spatial coordinates based on the depth camera parameters, and then determine the three-dimensional spatial coordinates of each pixel point corresponding to the operator based on the operator area, and determine the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment based on the operating equipment area; The safety warning module is used to calculate the distance between the operator and the operating equipment in real time based on the three-dimensional spatial coordinates of each pixel point corresponding to the operator and the three-dimensional spatial coordinates of each pixel point corresponding to the operating equipment, and dynamically adjust the alarm to issue a warning based on the comparison between the real-time calculated distance and the preset safety distance standard.

9. A medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for early warning of safety distance at a work site as claimed in any one of claims 1 to 7 is implemented.

10. A computer device comprising a medium, a processor, and a computer program stored on the medium and executable on the processor, wherein: When the processor executes the computer program, the method for early warning of safety distance at a work site as described in any one of claims 1 to 7 is implemented.

Images