Weak current engineering safety monitoring system

Abstract

The invention relates to the technical field of image recognition, in particular to a weak current engineering safety monitoring system, which remarkably improves the accuracy of face recognition of a worker in a weak current engineering environment by introducing a dynamic brightness threshold and a skin color compensation mechanism. The system can monitor and adjust the illumination condition in real time, and ensures that people with different skin colors can be accurately identified in various complex illumination environments. In addition, the system also has an intelligent warning function, so that unauthorized personnel are effectively prevented from entering a specific area, and on-site safety management is enhanced. Meanwhile, the integrated sound analysis module further improves the response speed and accuracy of the system, so that the whole monitoring process is more efficient and reliable. The technical defects of a traditional monitoring system are overcome, and a safer and more efficient solution is provided for weak current engineering construction.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and specifically to a security monitoring system for weak current engineering. Background Art

[0002] In the construction environment of weak current engineering, the lighting conditions are complex and changeable. There are often typical problems in the construction area, such as uneven light distribution (such as the mixed light difference between the equipment room, corridor and operation area), dynamic light source interference (such as the alternating action of temporary construction lamps and natural light), and local shadows caused by obstacles (such as the dark areas formed by pipeline erection). The traditional monitoring system uses a fixed brightness threshold for face recognition and supplementary light control, and there are multiple technical defects: First, the fixed threshold cannot adapt to the instantaneous brightness fluctuations caused by equipment movement or material stacking in the construction scene. When the face of the operator is in the dark area due to occlusion, the existing system fails to establish a real-time feedback mechanism, and the supplementary light delay increases the failure rate of face feature capture (experimental data shows that the misrecognition rate of people with dark skin color is as high as 38% in an environment below 150 Lux); Second, special working conditions such as metal material reflection and dust diffusion in the construction environment are likely to cause local overexposure, and the traditional brightness analysis module does not integrate an optical environment filtering algorithm, resulting in the loss of key biometric features due to high light overflow on the faces of people with light skin color; The construction team often includes multi-skin-color personnel working together, but the traditional system does not construct a skin color-illuminance compensation correlation model, resulting in a significant difference in the face brightness compliance rate of different skin color personnel under the same supplementary light intensity. Therefore, there is an urgent need for a security monitoring system that can accurately identify the faces of personnel in the construction environment of weak current engineering. Summary of the Invention

[0003] The technical problem solved by the present invention is to provide a security monitoring system for weak current engineering to solve the technical problem of low face recognition rate of traditional monitoring systems in the construction environment of weak current engineering.

[0004] The basic solution provided by the present invention: A security monitoring system for weak current engineering, including a server and a number of lighting terminals and monitoring terminals wirelessly connected to the server;

[0005] The monitoring terminal includes a video monitoring module, an image recognition module, and an information transmission module;

[0006] The video surveillance module is used to collect surveillance images of the target area; the image recognition module includes a face recognition module and a brightness analysis module; the brightness analysis module is provided with a brightness threshold. When the brightness of the face area of the target person in the surveillance image does not reach the brightness threshold, the brightness analysis module sends a brightness insufficient signal to the server through the information transmission module. After receiving the brightness insufficient signal, the server sends a brightness compensation signal to the lighting terminal in the target area, and the lighting terminal adjusts the light brightness according to the brightness compensation signal until the brightness of the face area of the target person reaches the brightness threshold; the face recognition module is used to perform face feature recognition on the target person in the surveillance image and upload the recognized face features to the server through the information transmission module;

[0007] The server includes a region division module, an identity recognition module, a personnel information database, and a warning module; the region division module is used to divide each region according to the regional work content and allocate the work permissions of the corresponding staff; the personnel information database includes personnel information including face features and work permissions; the identity recognition module is used to compare the face features with the personnel information database and identify the personnel information; the warning module is used to send a warning signal to the lighting terminal in the target area when a person without the work permission of the target area enters the area; after receiving the warning signal, the lighting terminal emits a red warning light signal for reminder.

[0008] Further, the brightness analysis module includes a skin color classification model, a threshold dynamic adjustment module, and a brightness determination module;

[0009] The threshold dynamic adjustment module identifies the skin color type of the face area through the skin color classification model, and performs brightness compensation analysis on the face area according to the identified skin color:

[0010]

[0011] where L ′ is the compensated brightness threshold, L is the initial brightness threshold, g i is the brightness correction factor of skin color type i, k is the compensation coefficient, M skin_i is the brightness reference value corresponding to the skin color type, M current is the average brightness value of the face area detected in the current surveillance image;

[0012] The brightness determination module makes the calculation of the face area brightness more accurate by compensating the face area brightness through the skin color.

[0013] Further, the construction of the skin color classification model includes the steps:

[0014] Data collection and preprocessing: Collect facial graphics of different skin colors under different lighting conditions and backgrounds, remove low-quality or irrelevant image data through data cleaning, and perform preprocessing operations on the image data;

[0015] Skin color region extraction: Extract the skin color region in the image through the skin color detection algorithm;

[0016] Feature extraction: Calculate the statistical features of the skin color region and select the features that have the most influence on skin color classification;

[0017] Model training: Label the skin color type for each image, divide the data set into a training set, a validation set and a test set; Select a neural network to train the model using the training set data and adjust the parameters on the validation set;

[0018] Model optimization: Evaluate the performance of the model on the test set and adjust the model parameters according to the evaluation results.

[0019] Furthermore, the lighting terminal includes a lighting lamp, a lighting adjustment module, a light sensor and a communication module; The light sensor is used to collect the ambient light intensity; The lighting adjustment module is provided with a light intensity threshold, and the lighting adjustment module is used to adjust the brightness of the lighting lamp when receiving a brightness compensation signal or when the light intensity does not reach the light intensity threshold; The communication module is used for wireless connection with the server.

[0020] Furthermore, the lighting terminal further includes a self-check module; The self-check module is set with a time duration threshold. When the time duration for the lighting adjustment module to adjust the brightness reaches the time duration threshold and the ambient light intensity of the target lighting lamp still does not reach the light intensity threshold, the self-check module sends a maintenance message to the server through the communication module. After receiving the maintenance message, the server issues a lighting abnormality warning through the warning module.

[0021] Furthermore, the lighting terminal further includes a pedestrian detection module; The pedestrian detection module includes a sound sensor, an analog-to-digital conversion module and a sound analysis module. The sound sensor is used to capture the ambient sound signal; The analog-to-digital conversion module is used to perform analog-to-digital conversion on the collected analog sound signal; The sound analysis module is used to judge whether there is a pedestrian passing by according to the converted digital sound signal. When it is judged that there is a pedestrian passing by in the target area, the sound analysis module sends a pedestrian signal to the server through the communication module; The server sends a face detection signal to the monitoring terminal in the target area, and the monitoring terminal performs face feature recognition on the target person in the monitoring image.

[0022] Furthermore, the monitoring terminal further includes an energy-saving control module; The server is further used to receive the ambient light intensity information of the lighting terminal and transmit it to the monitoring terminal; The energy-saving control module adjusts the resolution and frame rate of the monitoring image collected by the video monitoring module according to the ambient light intensity information received by the monitoring terminal; The lower the ambient light intensity, the higher the resolution and frame rate of the collected monitoring image.

[0023] The principle and advantages of the present invention are as follows: By introducing a dynamic brightness threshold and a skin color compensation mechanism, the present invention significantly improves the accuracy of facial recognition of staff in the weak current engineering environment. The system can monitor and adjust the lighting conditions in real time to ensure that personnel with different skin colors can be accurately recognized in various complex lighting environments. In addition, the system also has an intelligent warning function to effectively prevent unauthorized personnel from entering specific areas and enhance on-site safety management. At the same time, the integrated sound analysis module further improves the response speed and accuracy of the system, making the entire monitoring process more efficient and reliable. In summary, the present invention not only solves the technical defects of traditional monitoring systems but also provides a safer and more efficient solution for weak current engineering construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a logic block diagram of an embodiment of a safety monitoring system for weak current engineering according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The following is a further detailed description through specific embodiments:

[0026] The specific implementation process is as follows:

[0027] Embodiment 1

[0028] Embodiment 1 is basically as shown in the Figure 1 drawing. A safety monitoring system for weak current engineering includes a server and a number of lighting terminals and monitoring terminals wirelessly connected to the server. The monitoring terminal includes a video monitoring module, an image recognition module, an energy-saving control module, and an information transmission module; the video monitoring module is used to collect monitoring images of the target area; the image recognition module includes a face recognition module and a brightness analysis module; the brightness analysis module is provided with a brightness threshold. When the brightness of the face area of the target person in the monitoring image does not reach the brightness threshold, the brightness analysis module sends a brightness deficiency signal to the server through the information transmission module. After receiving the brightness deficiency signal, the server sends a brightness compensation signal to the lighting terminal in the target area, and the lighting terminal adjusts the lighting brightness according to the brightness compensation signal until the brightness of the face area of the target person reaches the brightness threshold; the face recognition module is used to perform face feature recognition on the target person in the monitoring image and upload the recognized face features to the server through the information transmission module.

[0029] The server includes a region division module, an identity recognition module, a personnel information database, and a warning module; the region division module is used to divide each region according to the regional work content and assign the work permissions of the corresponding staff; the personnel information database includes personnel information including face features and work permissions; the identity recognition module is used to compare the personnel information database according to the face features and identify the personnel information; the warning module is used to send a warning signal to the lighting terminal in the target region when a person without the work permission of the target region enters the region; after receiving the warning signal, the lighting terminal emits a red warning light signal for reminder.

[0030] Specifically, a skin color classification model is provided in the brightness analysis module, and the establishment steps of the skin color classification model include:

[0031] 1. Data collection and preprocessing: Collect facial images of different skin colors under different lighting conditions and backgrounds, remove low-quality or irrelevant image data through data cleaning, and perform preprocessing operations such as image size normalization, denoising, and color correction on the image data.

[0032] 2. Skin color region extraction: Extract the skin color region in the image through a skin color detection algorithm.

[0033] The steps of the skin color detection algorithm in this solution include:

[0034] 1) Color space selection: Convert the image from the RGB color space to a color space more suitable for skin color detection, such as HSV, YCbCr, or normalized RGB.

[0035] 2) Preliminary detection of skin color region: Use the skin color to define the range of skin color in the selected color space. For example, in the HSV space, define the value range of hue to include most skin colors. Then, extract the possible skin color regions through threshold segmentation technology.

[0036] 3) Refined processing of skin color region: Apply morphological operations (such as dilation and erosion) to remove noise and clarify the skin color region, and then perform connectivity analysis to eliminate non-skin color regions to ensure that only continuous skin color regions are retained.

[0037] 4) Skin color model establishment: Select a single Gaussian model as the skin color model, and use the labeled skin color and non-skin color data to train the skin color model.

[0038] 5) Use the prior knowledge of the face shape and the texture features of the skin color region to verify whether the detected skin color region may be a face to exclude non-face skin color regions.

[0039] 6) Extract features from the skin color region, such as color histogram, texture features, etc., and use a pre-trained classifier to finally confirm the skin color region.

[0040] 3. Feature extraction: Calculate the statistical features of the skin color region, including mean, variance, and histogram, and then select the features (color, texture, or shape) that have the most influence on skin color classification.

[0041] 4. Model training: Label the skin color type for each image, such as light color, natural color, dark color, etc., and divide the dataset into a training set, a validation set, and a test set. Select a suitable machine learning model, such as a neural network, use the training set data to train the model, and adjust the parameters on the validation set.

[0042] 5. Model optimization: Evaluate the performance of the model on the test set, use metrics such as accuracy, recall, and F1-score, perform cross-validation to ensure the stability and generalization ability of the model, and finally adjust the model parameters, such as the learning rate, kernel function, etc., according to the evaluation results.

[0043] When the brightness analysis module performs specific analysis, it also has a dynamic parameter adjustment step to ensure that the brightness analysis module can dynamically adjust the brightness calculation parameters according to the identified skin color type, so that the face regions of different skin colors can be accurately evaluated. The dynamic parameter adjustment first uses the skin color classification model to identify the skin color type of the face region and classify it into different skin color categories. Then, an initial brightness reference value is set for each skin color type, and this reference value represents the brightness level of the average face under that skin color type. According to the skin color type, different brightness correction factors are set, and these brightness correction factors are used to adjust the brightness calculation parameters to compensate for the differences in light reflection of different skin colors. Finally, the corresponding brightness correction factor is applied according to the identified skin color type to adjust the brightness calculation parameters. For example, for dark skin color, the gain factor in the brightness calculation parameters needs to be increased to accurately reflect its actual brightness. After the dynamic parameter adjustment, calculate the brightness of the face region and compensate for the brightness analysis of the face region according to the skin color. The brightness compensation algorithm involved can be expressed as:

[0044] L ′ = L + g i × k × (M skin_i - M current )

[0045] In the formula, L ′ is the compensated brightness threshold, L is the initial brightness threshold, g i is the brightness correction factor for skin color type i, k is the compensation coefficient, M skin_i is the brightness reference value corresponding to the skin color type, and M current is the average brightness value of the face region detected in the current monitored image. Compensating the brightness of the face region according to the skin color can make the calculation of the brightness of the face region more accurate. In this embodiment, after identifying the skin color type of the face, the corresponding brightness reference value M skin_i is loaded according to skin color type i. Among them, for light skin color, Mskin_1 = 120, normal skin color M skin_2 = 100, dark skin color M skin_3 = 80. Then, set the brightness correction factor g for skin color type i i , and combine it with the dynamic parameter M skin_i and g i to calculate the final brightness threshold L ′ . Among them, for light skin color, g1 = 0.8, for normal skin color, g2 = 1, and for dark skin color, g3 = 1.5. In this embodiment, if a person with dark skin color is detected, at this time, L = 100, k = 0.5, M skin_3 = 80, M current = 60, g3 = 1.5, then:

[0046] L ′ = 100 + 1.5×0.5×(80 - 60) = 100 + 15 = 115

[0047] Not only through the dual adjustment of M skin_i and g i , it adapts to the reflection differences of different skin colors to light, but also aims at the characteristic of low reflectivity of dark skin color, increases g i to improve the compensation sensitivity and avoid misjudgment of underexposure.

[0048] The main function of the energy-saving control module is to automatically control the working state of the video surveillance module according to the preset environmental conditions.

[0049] Specifically, the energy-saving control module in this solution can adjust the resolution and frame rate of the video images collected by the camera of the monitoring terminal. The monitoring lighting lamps in the monitoring terminal area are also used to upload the collected environmental light intensity information. The server forwards the environmental light intensity information to the corresponding monitoring terminal. The energy-saving control module of the monitoring terminal divides the environmental conditions into low light (night), medium light (dusk or cloudy day), and high light (sunny day) according to the environmental light intensity. Under low light conditions, the camera may need to increase the resolution to capture more details and reduce the frame rate to reduce noise and improve image quality. The energy-saving control module can set the collected video images to a resolution of 1080p and a frame rate of 30fps (frames per second). Under medium light, the resolution can be appropriately reduced, while the frame rate remains at a medium level to balance image quality and resource consumption. The energy-saving control module can set the collected video images to a resolution of 720p and a frame rate of 60fps. Under good lighting conditions, the resolution can be further reduced and the frame rate increased to maximize resource efficiency and maintain high monitoring fluency. The energy-saving control module can set the collected video images to a resolution of 480p and a frame rate of 120fps. Through the above structure, the video surveillance module can combine the information collected by the monitoring lighting lamps and adjust the resolution and frame rate in real time according to the continuous change of the lighting conditions.

[0050] The lighting terminal in this solution includes a lighting lamp, a lighting adjustment module, a light sensor, a self-check module, a pedestrian detection module, and a communication module; the light sensor is used to collect the ambient light intensity; the lighting adjustment module is set with a light intensity threshold, and the lighting adjustment module is used to adjust the brightness of the lighting lamp when receiving a brightness compensation signal or when the light intensity does not reach the light intensity threshold; the communication module is used to wirelessly connect to the server.

[0051] Specifically, the self-check module is set with a time duration threshold. When the time duration for the lighting adjustment module to adjust the brightness reaches the time duration threshold and the ambient light intensity of the target lighting lamp still does not reach the light intensity threshold, the self-check module sends a maintenance message to the server through the communication module. After receiving the maintenance message, the server issues a lighting abnormality warning through the warning module.

[0052] The pedestrian detection module includes a sound sensor, an analog-to-digital conversion module, and a sound analysis module. The sound sensor uses low-cost sound sensors such as LM386 to capture ambient sound signals. The analog-to-digital conversion module is used to perform analog-to-digital conversion on the collected analog sound signals. The sound analysis module is used to determine whether there are pedestrians passing by based on the converted digital sound signals. The sound analysis module in this solution uses an STM32 single-chip microcomputer to realize the monitoring and recognition of pedestrian sound characteristics by virtue of its characteristics of low power consumption, low cost, and stable performance.

[0053] The above are only embodiments of the present invention. Common knowledge such as specific structures and characteristics known in the solution is not described in detail here. Those of ordinary skill in the art know all the common technical knowledge in the technical field to which the invention belongs before the application date or the priority date, can know all the existing technologies in this field, and have the ability to apply the conventional experimental means before this date. Those of ordinary skill in the art can, under the inspiration given in this application, combine their own abilities to improve and implement this solution. Some typical well-known structures or well-known methods should not become an obstacle for those of ordinary skill in the art to implement this application. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope required by this application should be based on the content of its claims, and the specific implementation manners described in the specification can be used to explain the content of the claims.

Claims

1. A safety monitoring system for weak current engineering, characterized in that: It includes a server, and several lighting terminals and monitoring terminals wirelessly connected to the server; The monitoring terminal includes a video monitoring module, an image recognition module, and an information transmission module; The video monitoring module is used to collect monitoring images of the target area; the image recognition module includes a face recognition module and a brightness analysis module; The brightness analysis module is provided with a brightness threshold. When the brightness of the face area of the target person in the monitoring image does not reach the brightness threshold, the brightness analysis module sends a brightness insufficient signal to the server through the information transmission module. After receiving the brightness insufficient signal, the server sends a brightness compensation signal to the lighting terminal in the target area, and the lighting terminal adjusts the lighting brightness according to the brightness compensation signal until the brightness of the face area of the target person reaches the brightness threshold; the face recognition module is used to perform face feature recognition on the target person in the monitoring image and upload the recognized face features to the server through the information transmission module; The server includes a region division module, an identity recognition module, a personnel information database, and a warning module; the region division module is used to divide each region according to the regional work content and allocate the work permissions of the corresponding staff; the personnel information database includes personnel information including face features and work permissions; the identity recognition module is used to compare the face features with the personnel information database and identify the personnel information; the warning module is used to send a warning signal to the lighting terminal in the target area when a person without the work permission of the target area enters the area; after receiving the warning signal, the lighting terminal emits a red warning light signal for reminder.

2. The safety monitoring system for weak current engineering according to claim 1, characterized in that: The brightness analysis module includes a skin color classification model, a threshold dynamic adjustment module, and a brightness determination module; The threshold dynamic adjustment module identifies the skin color type of the face area through the skin color classification model, and performs brightness compensation analysis on the face area according to the identified skin color: L ′ = L + g i × k × (M skin_i - M current ) Wherein, L ′ is the compensated brightness threshold, L is the initial brightness threshold, g i is the brightness correction factor for skin color type i, k is the compensation coefficient, M skin_i is the brightness reference value corresponding to the skin color type, M current is the average brightness value of the face region detected in the current monitored image; The brightness determination module makes the calculation of the brightness of the face area more accurate by compensating the brightness of the face area through the skin color.

3. The safety monitoring system for weak current engineering according to claim 2, characterized in that: The construction of the skin color classification model includes the following steps: Data collection and preprocessing: Collect facial graphics of different skin colors under different lighting conditions and backgrounds, remove low-quality or irrelevant image data through data cleaning, and perform preprocessing operations on the image data; Skin color area extraction: Extract the skin color area in the image through the skin color detection algorithm; Feature extraction: Calculate the statistical features of the skin color area, and select the features that have the greatest impact on skin color classification; Model training: Label the skin color type for each image, divide the data set into a training set, a validation set, and a test set; select a neural network to train the model using the training set data and adjust the parameters on the validation set; Model optimization: Evaluate the performance of the model on the test set, and adjust the model parameters through the evaluation results.

4. The safety monitoring system for a weak current project according to claim 3, wherein: The lighting terminal includes a lighting lamp, a lighting adjustment module, a light sensor, and a communication module; the light sensor is used to collect the ambient light intensity; the lighting adjustment module is provided with a light intensity threshold, and the lighting adjustment module is used to adjust the brightness of the lighting lamp when receiving the brightness compensation signal or when the light intensity does not reach the light intensity threshold; the communication module is used to wirelessly connect to the server.

5. The safety monitoring system for a weak current project according to claim 4, characterized in that: The lighting terminal further includes a self-check module; the self-check module is set with a duration threshold. When the duration of the lighting adjustment module adjusting the brightness reaches the duration threshold and the ambient light intensity of the target lighting fixture still does not reach the light intensity threshold, the self-check module sends a maintenance message to the server through the communication module. After receiving the maintenance message, the server issues a lighting anomaly warning through the warning module.

6. The safety monitoring system for a weak current project according to claim 5, characterized in that: The lighting terminal further includes a pedestrian detection module; the pedestrian detection module includes a sound sensor, an analog-to-digital conversion module, and a sound analysis module; the sound sensor is used to capture ambient sound signals; the analog-to-digital conversion module is used to perform analog-to-digital conversion on the collected analog sound signals. The sound analysis module is used to determine whether there is a pedestrian passing by according to the converted digital sound signal. When it is determined that a pedestrian passes through the target area, the sound analysis module sends a pedestrian signal to the server through the communication module. The server sends a face detection signal to the monitoring terminal in the target area, and the monitoring terminal performs face feature recognition on the target person in the monitoring image.

7. An intelligent and weak current engineering safety monitoring system according to claim 6, characterized in that: The monitoring terminal further includes an energy-saving control module; the server is further used to receive the ambient light intensity information of the lighting terminal and transmit it to the monitoring terminal; the energy-saving control module adjusts the resolution and frame rate of the monitoring image collected by the video monitoring module according to the ambient light intensity information received by the monitoring terminal. The lower the ambient light intensity, the higher the resolution and frame rate of the collected monitoring image.

Images