Construction safety monitoring method and system based on artificial intelligence

Abstract

The invention provides a construction safety monitoring method and system based on artificial intelligence. The method comprises the following steps: smoothing a construction image to obtain a smoothed construction image; performing brightness calibration on the smoothed construction image to obtain a calibrated construction image; marking the calibrated construction image to form a training sample; inputting the training sample into a YOLO network for training to obtain a construction safety prediction model; inputting the construction image of the target engineering site into the construction safety prediction model to obtain a construction state; when the construction state is abnormal, an alarm is given out. According to the invention, the construction image is smoothed and calibrated, the image is used as a training sample and is input into the YOLO network for training to obtain the construction safety prediction model, the working state of each construction worker can be accurately identified based on the trained model, and when the construction state is abnormal, the system can give an alarm in time, so that the construction safety is improved. Management personnel are helped to take corresponding measures, and the probability of accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of construction risk monitoring, and in particular to an artificial intelligence-based construction safety monitoring method and system. Background Art

[0002] In recent years, with the development of economy and society, the status of infrastructure industry in my country's national economy has been increasing. However, as the centralized implementation place of construction production factors, the working environment of infrastructure construction site is constantly changing, and there are great safety hazards. With the rapid development of the construction industry, construction safety issues have become increasingly prominent. Frequent accidents have not only caused serious injuries to construction workers, but also led to huge economic losses and social impacts. Traditional construction safety monitoring methods mainly rely on manual inspections, but manual inspections are limited by human resources and often cannot effectively monitor the construction site in real time and comprehensively. Summary of the invention

[0003] To solve the above problems, the purpose of the present invention is to provide a construction safety monitoring method and system based on artificial intelligence.

[0004] A construction safety monitoring method based on artificial intelligence, comprising:

[0005] Step 1: Obtain construction images of construction workers at the construction site;

[0006] Step 2: Smoothing the construction image to obtain a smoothed construction image;

[0007] Step 3: Perform brightness calibration on the smoothed construction image to obtain a calibrated construction image;

[0008] Step 4: Annotate the calibrated construction images to form training samples;

[0009] Step 5: Input the training samples into the YOLO network for training to obtain the construction safety prediction model;

[0010] Step 6: Input the construction image of the target project site into the construction safety prediction model to obtain the construction status;

[0011] Step 7: When the construction status is abnormal, an alarm is issued.

[0012] Preferably, the step 2: smoothing the construction image to obtain a smoothed construction image comprises:

[0013] Step 2.1: evenly divide the construction image into a number of square smoothing windows;

[0014] Step 2.2: Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window;

[0015] Step 2.3: Construct a smoothing function based on spatial similarity and distance similarity;

[0016] Step 2.4: Use the smoothing function to smooth the square smoothing window to obtain a smoothed construction image.

[0017] Preferably, the step 2.2: calculating the spatial similarity and distance similarity between each pixel point in each square smoothing window includes:

[0018] Using the formula:

[0019]

[0020] Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window; where w s (i,j) represents the spatial similarity, w v (i, j) represents the distance similarity, S(i, j) represents the pixel value of the square smoothing window at the position (i, j), S(x, y) represents the pixel value of the square smoothing window at the position (x, y), δ s Represents the smoothed value.

[0021] Preferably, the step 2.3: constructing a smoothing function based on spatial similarity and distance similarity comprises:

[0022] Using the formula:

[0023]

[0024] Construct a smoothing function; where B(x,y) represents the smoothing value, w p Represents the normalized weight coefficient.

[0025] Preferably, the step 3: performing brightness calibration on the smoothed construction image to obtain a calibrated construction image comprises:

[0026] Step 3.1: Calculate the brightness value of each pixel of the smoothed construction image;

[0027] Step 3.2: Perform gamma correction on the brightness value of each pixel to obtain a gamma-corrected brightness value;

[0028] Step 3.3: Calculate the brightness gain matrix based on the brightness value after gamma correction;

[0029] Step 3.4: Use the gain matrix to calibrate the smoothed construction image to obtain a calibrated construction image.

[0030] Preferably, the step 3.2: performing gamma correction on the brightness value of each pixel to obtain a gamma-corrected brightness value includes:

[0031] Using the formula:

[0032]

[0033] Gamma correction is performed on the brightness value of each pixel to obtain the brightness value after gamma correction; where I g (x,y) represents the brightness value after gamma correction, I(x,y) represents the brightness value of the pixel at the position (x,y), R(x,y) represents the value of the red color channel at the position (x,y), G(x,y) represents the value of the green color channel at the position (x,y), and B(x,y) represents the value of the blue color channel at the position (x,y).

[0034] Preferably, the step 3.4: calibrating the smoothed construction image using the gain matrix to obtain a calibrated construction image includes:

[0035] Step 3.4.1: Determine a gain matrix based on the ratio between the gamma-corrected brightness value and the original brightness value;

[0036] Step 3.4.2: Use the gain matrix to calibrate the pixel value of each channel on the smoothed construction image to obtain a calibrated construction image; wherein the calibration formula is:

[0037]

[0038] Among them, λ(x,y) represents the gain matrix, R new (x, y) represents the value of the red color channel at the position (x, y), G new (x, y) represents the value of the green channel at the position (x, y), B new (x,y) represents the value of the blue color channel at the position (x,y).

[0039] Preferably, the step 5: inputting the training samples into the YOLO network for training to obtain a construction safety prediction model includes:

[0040] The Adam optimizer is used to adjust the learning rate of the YOLO network, and the sum of the positioning loss function, confidence loss function and classification loss function is used as the overall loss function to optimize the YOLO network to obtain the construction safety prediction model.

[0041] The present invention also provides a construction safety monitoring system based on artificial intelligence, comprising:

[0042] A construction image acquisition module is used to acquire construction images of construction workers at the construction site;

[0043] A smoothing processing module, used for performing smoothing processing on the construction image to obtain a smoothed construction image;

[0044] A calibration module, used for performing brightness calibration on the smoothed construction image to obtain a calibrated construction image;

[0045] An annotation module, used to annotate the calibrated construction images to form training samples;

[0046] The training module is used to input the training samples into the YOLO network for training to obtain a construction safety prediction model;

[0047] A construction status judgment module, used to input the construction image of the target project site into the construction safety prediction model to obtain the construction status;

[0048] The alarm module is used to issue an alarm when the construction status is abnormal.

[0049] The present invention also provides an electronic device, comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, and wherein the computer program, when executed by the processor, implements the steps in the above-mentioned construction safety monitoring method based on artificial intelligence.

[0050] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the above-mentioned construction safety monitoring method based on artificial intelligence are implemented.

[0051] The beneficial effect of the artificial intelligence-based construction safety monitoring method and system provided by the present invention is that: compared with the prior art, the present invention smoothes and calibrates the construction images of the construction site, and uses the smoothed and calibrated images as training samples, which are input into the YOLO network for training to obtain a construction safety prediction model. The model can then be used to accurately identify the working status of each construction worker, and when the construction status is abnormal, the system will issue an alarm in time to help managers take corresponding measures and reduce the probability of accidents.

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 A flowchart of a construction safety monitoring method based on artificial intelligence provided by an embodiment of the present invention is shown;

[0055] Figure 2 A schematic diagram of an electronic device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0056] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0057] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0058] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0059] See also Figure 1 , a construction safety monitoring method based on artificial intelligence, comprising:

[0060] Step 1: Obtain construction images of construction workers at the construction site;

[0061] Step 2: Smoothing the construction image to obtain a smoothed construction image;

[0062] Furthermore, step 2: performing smoothing processing on the construction image to obtain a smoothed construction image includes:

[0063] Step 2.1: evenly divide the construction image into a number of square smoothing windows;

[0064] Step 2.2: Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window;

[0065] In step 2.2, the spatial similarity and distance similarity between each pixel under each square smoothing window are calculated, including:

[0066] Using the formula:

[0067]

[0068] Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window; where w s (i,j) represents the spatial similarity, w v (i, j) represents the distance similarity, S(i, j) represents the pixel value of the square smoothing window at the position (i, j), S(x, y) represents the pixel value of the square smoothing window at the position (x, y), δ s Represents the smoothed value.

[0069] Step 2.3: Construct a smoothing function based on spatial similarity and distance similarity; wherein the smoothing function is:

[0070]

[0071] Construct a smoothing function; where B(x,y) represents the smoothing value, w p Represents the normalized weight coefficient.

[0072] Step 2.4: Use the smoothing function to smooth the square smoothing window to obtain a smoothed construction image.

[0073] The present invention combines the two factors of spatial similarity and distance similarity of pixels and performs weighted averaging of pixel values ​​in a square smoothing window to calculate a smoothed construction image, which can not only maintain the edge while smoothing the image, but also remove image noise.

[0074] Step 3: Perform brightness calibration on the smoothed construction image to obtain a calibrated construction image;

[0075] Lighting conditions at construction sites often vary significantly due to time, weather, or equipment location. Brightness calibration can eliminate the impact of these changes on image content, ensuring that the features learned by the model during training are more representative and robust.

[0076] Furthermore, step 3 includes:

[0077] Step 3.1: Calculate the brightness value of each pixel of the smoothed construction image;

[0078] Step 3.2: Perform gamma correction on the brightness value of each pixel to obtain a gamma-corrected brightness value;

[0079] Wherein, in the step 3.2, the present invention may adopt the formula:

[0080]

[0081] Gamma correction is performed on the brightness value of each pixel to obtain the brightness value after gamma correction; where I g (x,y) represents the brightness value after gamma correction, I(x,y) represents the brightness value of the pixel at the position (x,y), R(x,y) represents the value of the red color channel at the position (x,y), G(x,y) represents the value of the green color channel at the position (x,y), and B(x,y) represents the value of the blue color channel at the position (x,y).

[0082] Step 3.3: Calculate the brightness gain matrix based on the brightness value after gamma correction;

[0083] Step 3.4: Use the gain matrix to calibrate the smoothed construction image to obtain a calibrated construction image.

[0084] Furthermore, step 3.4 includes:

[0085] Step 3.4.1: Determine a gain matrix based on the ratio between the gamma-corrected brightness value and the original brightness value;

[0086] Step 3.4.2: Use the gain matrix to calibrate the pixel value of each channel on the smoothed construction image to obtain a calibrated construction image; wherein the calibration formula is:

[0087]

[0088] Among them, λ(x,y) represents the gain matrix, R new (x, y) represents the value of the red color channel at the position (x, y), G new (x, y) represents the value of the green channel at the position (x, y), B new(x,y) represents the value of the blue color channel at the position (x,y).

[0089] The calibrated images can be better mapped to the expected input range of the model, reducing the training error caused by uneven lighting, thereby accelerating the convergence of the model and improving training efficiency.

[0090] Step 4: Annotate the calibrated construction images to form training samples;

[0091] First, select a suitable image annotation tool, such as LabelImg and RectLabel, and then determine the annotation strategy based on the type of object to be identified. For example, if a construction worker is not wearing a helmet or is too close to construction equipment, the image is marked as an abnormal working state of the construction worker. Finally, integrate the annotated images and the corresponding annotation files into a training dataset.

[0092] Step 5: Input the training samples into the YOLO network for training to obtain the construction safety prediction model;

[0093] In an embodiment of the present invention, the Adam optimizer can be used to adjust the learning rate of the YOLO network, and the sum of the positioning loss function, the confidence loss function and the classification loss function is used as the overall loss function to optimize the YOLO network to obtain a construction safety prediction model.

[0094] The Adam optimizer can adaptively adjust the learning rate of each parameter, avoiding the trouble of manually setting the learning rate. For different learning stages, Adam will adjust the learning rate according to the first-order and second-order moment estimates of the gradient, improving the efficiency and stability of training. In addition, the present invention combines positioning loss, confidence loss and classification loss to evaluate the performance of the model from multiple dimensions, so that the model not only focuses on the location of the object, but also on the accuracy of object recognition and the existence of the target. This comprehensive consideration helps to improve the prediction accuracy of the model in complex construction scenarios.

[0095] Step 6: Input the construction image of the target project site into the construction safety prediction model to obtain the construction status;

[0096] Step 7: When the construction status is abnormal, an alarm is issued.

[0097] The present invention smoothes and calibrates the construction images of the construction site, and uses the smoothed and calibrated images as training samples, which are input into the YOLO network for training to obtain a construction safety prediction model. The model can then be used to accurately identify the working status of each construction worker, and when the construction status is abnormal, the system will issue an alarm in time to help managers take corresponding measures and reduce the probability of accidents.

[0098] The present invention also provides a construction safety monitoring system based on artificial intelligence, comprising:

[0099] A construction image acquisition module is used to acquire construction images of construction workers at the construction site;

[0100] A smoothing processing module, used for performing smoothing processing on the construction image to obtain a smoothed construction image;

[0101] A calibration module, used for performing brightness calibration on the smoothed construction image to obtain a calibrated construction image;

[0102] An annotation module, used to annotate the calibrated construction images to form training samples;

[0103] The training module is used to input the training samples into the YOLO network for training to obtain a construction safety prediction model;

[0104] A construction status judgment module, used to input the construction image of the target project site into the construction safety prediction model to obtain the construction status;

[0105] The alarm module is used to issue an alarm when the construction status is abnormal.

[0106] For details, see Figure 2 As shown, an embodiment of the present invention further provides an electronic device, which includes a bus 1110 , a processor 1120 , a transceiver 1130 , a bus interface 1140 , a memory 1150 and a user interface 1160 .

[0107] In an embodiment of the present invention, the electronic device also includes: a computer program stored in the memory 1150 and executable on the processor 1120, and when the computer program is executed by the processor 1120, the above-mentioned artificial intelligence-based construction safety monitoring method is implemented.

[0108] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the above-mentioned construction safety monitoring method based on artificial intelligence are implemented. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present invention are the same as the beneficial effects of the construction safety monitoring method based on artificial intelligence described in the above technical solution, which will not be repeated here.

[0109] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technical solution that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A construction safety monitoring method based on artificial intelligence, characterized in that: include: Step 1: Obtain construction images of construction workers at the construction site; Step 2: Smoothing the construction image to obtain a smoothed construction image; Step 3: Perform brightness calibration on the smoothed construction image to obtain a calibrated construction image; Step 4: Annotate the calibrated construction images to form training samples; Step 5: Input the training samples into the YOLO network for training to obtain the construction safety prediction model; Step 6: Input the construction image of the target project site into the construction safety prediction model to obtain the construction status; Step 7: When the construction status is abnormal, an alarm is issued.

2. The construction safety monitoring method based on artificial intelligence according to claim 1 is characterized in that: The step 2: performing smoothing processing on the construction image to obtain a smoothed construction image comprises: Step 2.1: evenly divide the construction image into a number of square smoothing windows; Step 2.2: Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window; Step 2.3: Construct a smoothing function based on spatial similarity and distance similarity; Step 2.4: Use the smoothing function to smooth the square smoothing window to obtain a smoothed construction image.

3. The construction safety monitoring method based on artificial intelligence according to claim 2 is characterized in that: The step 2.2: calculating the spatial similarity and distance similarity between each pixel point under each square smoothing window includes: Using the formula: Calculate the spatial similarity and distance similarity between each pixel under each square smoothing window; where w s (i,j) represents the spatial similarity, w v (i, j) represents the distance similarity, S(i, j) represents the pixel value of the square smoothing window at the position (i, j), S(x, y) represents the pixel value of the square smoothing window at the position (x, y), δ s Represents the smoothed value.

4. The construction safety monitoring method based on artificial intelligence according to claim 3 is characterized in that: The step 2.3: constructing a smoothing function based on spatial similarity and distance similarity, comprises: Using the formula: Construct a smoothing function; where B(x,y) represents the smoothing value, w p Represents the normalized weight coefficient.

5. The construction safety monitoring method based on artificial intelligence according to claim 4 is characterized in that: The step 3: performing brightness calibration on the smoothed construction image to obtain a calibrated construction image, comprises: Step 3.1: Calculate the brightness value of each pixel of the smoothed construction image; Step 3.2: Perform gamma correction on the brightness value of each pixel to obtain a gamma-corrected brightness value; Step 3.3: Calculate the brightness gain matrix based on the brightness value after gamma correction; Step 3.4: Use the gain matrix to calibrate the smoothed construction image to obtain a calibrated construction image.

6. The construction safety monitoring method based on artificial intelligence according to claim 5 is characterized in that: The step 3.2: performing gamma correction on the brightness value of each pixel to obtain a gamma-corrected brightness value, includes: Using the formula: Gamma correction is performed on the brightness value of each pixel to obtain the brightness value after gamma correction; where I g (x,y) represents the brightness value after gamma correction, I(x,y) represents the brightness value of the pixel at the position (x,y), R(x,y) represents the value of the red color channel at the position (x,y), G(x,y) represents the value of the green color channel at the position (x,y), and B(x,y) represents the value of the blue color channel at the position (x,y).

7. The construction safety monitoring method based on artificial intelligence according to claim 6 is characterized in that: The step 3.4: using the gain matrix to calibrate the smoothed construction image to obtain a calibrated construction image, includes: Step 3.4.1: Determine a gain matrix based on the ratio between the gamma-corrected brightness value and the original brightness value; Step 3.4.2: Use the gain matrix to calibrate the pixel value of each channel on the smoothed construction image to obtain a calibrated construction image; wherein the calibration formula is: Among them, λ(x,y) represents the gain matrix, R new (x, y) represents the value of the red color channel at the position (x, y), G new (x, y) represents the value of the green color channel at the position (x, y), B new (x,y) represents the value of the blue color channel at the position (x,y).

8. The construction safety monitoring method based on artificial intelligence according to claim 1 is characterized in that: The step 5: inputting the training samples into the YOLO network for training to obtain a construction safety prediction model, including: The Adam optimizer is used to adjust the learning rate of the YOLO network, and the sum of the positioning loss function, confidence loss function and classification loss function is used as the overall loss function to optimize the YOLO network to obtain the construction safety prediction model.

9. A construction safety monitoring system based on artificial intelligence, characterized in that: include: A construction image acquisition module is used to acquire construction images of construction workers at the construction site; A smoothing processing module, used for performing smoothing processing on the construction image to obtain a smoothed construction image; A calibration module, used for performing brightness calibration on the smoothed construction image to obtain a calibrated construction image; An annotation module, used to annotate the calibrated construction images to form training samples; The training module is used to input the training samples into the YOLO network for training to obtain a construction safety prediction model; A construction status judgment module, used to input the construction image of the target project site into the construction safety prediction model to obtain the construction status; The alarm module is used to issue an alarm when the construction status is abnormal.

10. An electronic device, comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, wherein: When the computer program is executed by the processor, the steps of the construction safety monitoring method based on artificial intelligence as described in any one of claims 1 to 8 are implemented.

Images