Intelligent visual early warning system and method for preventing cable take-up stand from being wound by personnel

Abstract

The invention discloses an intelligent visual early warning system and method for preventing personnel from being wound on a cable take-up stand, and relates to the technical field of active protection in industrial safety production, and the system comprises a video collection module which employs a wide-angle high-definition camera disposed in a working area of a take-up machine to collect images of the working area of the take-up machine in real time, a data basis is provided for subsequent analysis; the intelligent analysis module integrates YOLOv5 personnel real-time detection, PP-LiteSeg wire coil segmentation and thickness calculation functions, dynamically demarcates dangerous and warning areas according to preset parameters through a dynamic area calibration algorithm, and provides accurate intelligent analysis support for safety protection of the take-up machine; and the alarm execution module is used for carrying out graded warning by utilizing red, yellow and green industrial voice lamps based on an analysis result of the intelligent analysis module, and is linked with the PLC to control the take-up and arrangement machine to carry out emergency braking when detecting that a person invades a dangerous area. The method is suitable for real-time monitoring of the working area of the take-up and arrangement machine in the cable production workshop, and the occurrence probability of safety accidents can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of active protection in industrial safety production, and specifically relates to an intelligent vision early warning system and method for preventing personnel from being wound around a wire and cable take-up rack. Background Art

[0002] The current protective measures adopted by wire and cable production workshop winding and arranging machines have significant deficiencies, which are specifically manifested as follows:

[0003] Gravity induction blanket: In terms of induction performance, its trigger delay exceeds 1.5 seconds, making it difficult to respond in a timely manner at the moment when danger occurs; the false trigger rate is as high as 35%, and frequent false alarms not only interfere with the normal production order, but also lead to a decrease in the trust of operators in the protection device. In terms of the installation structure, the gravity induction blanket is laid on the ground, and the raised structure forms a risk of tripping on the ground area. When workers walk quickly or carry materials in the workshop, they are easily tripped, which instead increases the potential danger of personnel injury.

[0004] Intercepting rope: Although the setting of the intercepting rope can play a protective role to a certain extent, it has a negative impact on production efficiency, reducing the production efficiency by 18%. More seriously, when workers deliberately bypass it in order to improve work efficiency, the intercepting rope becomes ineffective and cannot play a protective function, and the potential safety hazard still exists.

[0005] Infrared grating: The protection mechanism of the infrared grating depends on physical contact to trigger a response, and this characteristic makes it unable to cope when facing high-speed rotating wire and cable. When the wire and cable linear speed reaches or exceeds 5 m / s, there is a time difference between the occurrence of danger and the triggering of the infrared grating to respond, and it is impossible to predict danger and make a reaction in advance, resulting in protection lag and making it difficult to ensure personnel safety.

[0006] General vision system: In practical applications, the general vision system cannot dynamically adapt to the change of the wire reel size. Whenever the wire reel model is changed, manual calibration is required, which increases the complexity of operation and labor costs. In addition, the system has poor anti-interference ability, and the false alarm rate exceeds 40% under light interference. A large number of false alarm messages will make operators exhausted to deal with, reducing the attention to real danger information, and thus weakening the protection effect.

[0007] All of the above protection technologies have their own limitations, and they cannot predict dangerous behaviors in real time and dynamically adjust protection strategies according to actual situations, making it difficult to meet the increasing safety protection requirements of wire and cable production workshops. Summary of the Invention

[0008] In view of the current needs and deficiencies in the development of technology, the present invention provides an intelligent vision early warning system and method for preventing personnel from being wound around a wire and cable take-up rack.

[0009] In the first aspect, the present invention provides an intelligent visual warning system for preventing personnel from getting entangled in a cable take-up rack. The technical solution adopted to solve the above technical problems is as follows:

[0010] An intelligent visual warning system for preventing personnel from getting entangled in a cable take-up rack, comprising:

[0011] The video acquisition module uses a wide-angle high-definition camera deployed in the working area of the wire take-up machine to collect images of the working area of the wire take-up machine in real time, providing a data basis for subsequent analysis;

[0012] The intelligent analysis module integrates YOLOv5 real-time personnel detection, PP-LiteSeg wire drum segmentation and thickness calculation functions, and uses a dynamic area calibration algorithm to dynamically define danger and warning areas based on preset parameters, providing accurate intelligent analysis support for the safety protection of the wire take-up machine;

[0013] The alarm execution module, based on the analysis results of the intelligent analysis module, uses red, yellow and green industrial voice lights to provide graded warnings. When a person is detected intruding into the dangerous area, the PLC is linked to control the emergency braking of the wire take-up and traversing machine.

[0014] Optionally, the intelligent analysis modules involved specifically include:

[0015] The personnel behavior detection unit is used to receive the collected images from the video acquisition module and use the advanced YOLOv5 algorithm to perform real-time and accurate detection of personnel in the working area of the wire take-up machine, and timely capture various behaviors of personnel;

[0016] The cable drum segmentation unit is used to accurately segment the cable drum based on the detection results of the personnel behavior detection unit using the lightweight PP-LiteSeg model with high performance. It also accurately calculates the thickness of the cable to provide reliable data for subsequent production analysis;

[0017] The dynamic area calibration algorithm unit is used to dynamically and accurately divide the danger area and the warning area based on a preset algorithm.

[0018] Preferably, the preset algorithm includes:

[0019] The calculation formula of the edge width of the dangerous area A is: A = R + ΔL + 0.33D;

[0020] The calculation formula of the edge width of the warning area B is: B = A + 0.25D + K*v;

[0021] Among them, R represents the reference radius of the take-up reel, D is the dynamic detection value of the cable layer thickness, ΔL is the mechanical tolerance compensation, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of the personnel movement.

[0022] Optionally, the alarm execution module involved constructs a hierarchical early warning and emergency response system with a three-color industrial voice light (red, yellow, and green) as the core carrier;

[0023] When the intelligent analysis module detects that a person enters the warning area and has not yet reached the boundary of the dangerous area, the alarm execution module automatically triggers the yellow warning light to flash continuously, accompanied by a low-frequency prompt sound, to remind the person of potential risks with acoustic and optical signals. At this time, the wire coiling and discharging machine still maintains a normal operating state;

[0024] When the intelligent analysis module detects that a person's torso crosses the boundary of the dangerous area and triggers the dangerous alarm condition, the alarm execution module immediately switches to the red warning mode. The red light flashes at a high frequency and emits a high-decibel sharp warning sound. At the same time, an emergency stop instruction is sent to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the wire coiling and discharging machine and starts the braking device to stop the equipment from operating within a very short time.

[0025] In a second aspect, the present invention provides an intelligent vision early warning method for preventing personnel from being wound around a wire coiling rack. The technical solution adopted to solve the above technical problems is as follows:

[0026] An intelligent vision early warning method for preventing personnel from being wound around a wire coiling rack includes the following steps:

[0027] S1. Deploy a wide-angle high-definition camera in the working area of the wire coiling machine, and collect the image of the working area of the wire coiling machine in real time through the wide-angle high-definition camera to provide a data basis for subsequent analysis;

[0028] S2. Based on the image collected by the wide-angle high-definition camera, perform real-time and accurate detection of personnel in the working area of the wire coiling machine. Based on the detection results, perform accurate segmentation of the wire coil, and at the same time accurately calculate the thickness of the wire cable to dynamically and accurately divide the dangerous area and the warning area;

[0029] S3. Based on the division results, use a three-color industrial voice light (red, yellow, and green) for hierarchical warning. When it is detected that a person invades the dangerous area, link the PLC to control the wire coiling and discharging machine to perform an emergency brake.

[0030] Optionally, the involved step S2 specifically includes:

[0031] S2.1. Based on the image collected by the wide-angle high-definition camera, use the advanced YOLOv5 algorithm to perform real-time and accurate detection of personnel in the working area of the wire coiling machine, and timely capture various behavioral actions of the personnel;

[0032] S2.2. Based on the detection results, use the lightweight PP-LiteSeg model with high performance to achieve accurate segmentation of the wire coil. At the same time, accurately calculate the thickness of the wire cable to provide reliable data for subsequent production analysis;

[0033] S2.3. Dynamically and accurately divide the danger area and warning area based on a preset algorithm.

[0034] Preferably, the preset algorithm includes:

[0035] The formula for the edge width of the danger area A is: A = R + ΔL + 0.33D;

[0036] The formula for the edge width of the warning area B is: B = A + 0.25D + K*v;

[0037] Wherein, R represents the reference radius of the take-up reel, D is the dynamically detected value of the cable layer thickness, ΔL is the mechanical tolerance compensation amount, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of personnel movement.

[0038] Optionally, the specific steps of step S3 include:

[0039] Pre-build a hierarchical early warning and emergency response system with industrial voice lights of red, yellow, and green as the core carriers;

[0040] When it is detected that a person enters the warning area and has not yet touched the boundary of the danger area, the yellow warning light is automatically triggered to flash continuously, accompanied by a low-frequency prompt sound, to remind the person of potential risks with sound and light signals. At this time, the pay-off and take-up machine still maintains a normal operating state;

[0041] When it is detected that a person's torso crosses the boundary of the danger area and triggers the danger alarm condition, immediately switch to the red warning mode. The red light flashes at a high frequency and emits a high-decibel sharp warning sound. At the same time, send an emergency stop instruction to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the pay-off and take-up machine and starts the braking device to stop the equipment from running within a very short time.

[0042] The beneficial effects of an intelligent vision early warning system and method for preventing personnel entanglement in a cable take-up rack of the present invention compared with the prior art are:

[0043] 1. The present invention is applicable to the collaborative work of three links: video acquisition, intelligent analysis, and alarm execution, realizing real-time monitoring, accurate analysis, and intelligent protection of the working area of the pay-off and take-up machine in the cable production workshop, effectively ensuring personnel safety and the smooth progress of production;

[0044] 2. By deploying a wide-angle high-definition camera in the working area of the take-up machine, the present invention can collect images of this area in real time. The characteristics of wide-angle high-definition ensure the comprehensiveness of the collection range and the clarity of the images, providing a rich and accurate data basis for subsequent analysis. This enables the system to obtain various information during the operation of the take-up machine, including personnel activities, the status of the cable reel, etc., providing strong support for subsequent accurate analysis and decision-making;

[0045] 3. The present invention conducts multi-faceted intelligent analysis based on the collected images; conducts real-time and accurate detection of personnel, and can promptly discover abnormal behavior of personnel, such as whether they are close to dangerous areas, etc.; accurately divides the cable drum and accurately calculates the cable thickness, which helps to understand the production progress and equipment operation status; dynamically and accurately divides the dangerous area and the warning area, and adjusts the safety range in real time according to the actual situation, thereby improving the pertinence and effectiveness of safety protection;

[0046] 4. The present invention uses red, yellow and green industrial voice lights to give graded warnings based on the division results of dangerous areas and warning areas; different colors of lights and voice prompts can make personnel intuitively understand the current safety status and serve as timely reminders; when it is detected that a person has intruded into a dangerous area, the PLC is linked to control the emergency braking of the wire take-up and arrangement machine, which can quickly stop the operation of the equipment to avoid harm to personnel and ensure the life safety of personnel to the greatest extent;

[0047] 5. On the one hand, the present invention reduces the workload and errors of manual monitoring, reduces the probability of safety accidents, and improves production safety; on the other hand, it avoids the problem of reduced production efficiency caused by personnel deliberately circumventing protective devices or false triggering of protective devices, ensures the smooth progress of production, and improves overall production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Attached Figure 1 is a module connection diagram of the first embodiment of the present invention;

[0049] Attached Figure 2 It is a method flow chart of embodiment 2 of the present invention. DETAILED DESCRIPTION

[0050] In order to make the technical solution, the technical problem solved and the technical effect of the present invention more clearly understood, the technical solution of the present invention is clearly and completely described below in conjunction with specific embodiments.

[0051] Embodiment 1:

[0052] Reference Figure 1 This embodiment provides an intelligent visual warning system for preventing personnel from getting entangled in a cable take-up rack, which includes:

[0053] The video acquisition module uses a wide-angle high-definition camera deployed in the working area of the wire take-up machine to collect images of the working area of the wire take-up machine in real time, providing a data basis for subsequent analysis;

[0054] The intelligent analysis module integrates YOLOv5 real-time personnel detection, PP-LiteSeg wire drum segmentation and thickness calculation functions, and uses a dynamic area calibration algorithm to dynamically define danger and warning areas based on preset parameters, providing accurate intelligent analysis support for the safety protection of the wire take-up machine;

[0055] The alarm execution module, based on the analysis results of the intelligent analysis module, uses red, yellow and green industrial voice lights to provide graded warnings. When a person is detected intruding into the dangerous area, the PLC is linked to control the emergency braking of the wire take-up and traversing machine.

[0056] In this embodiment, the intelligent analysis module involved specifically includes:

[0057] The personnel behavior detection unit is used to receive the collected images from the video acquisition module and use the advanced YOLOv5 algorithm to perform real-time and accurate detection of personnel in the working area of the wire take-up machine, and timely capture various behaviors of personnel;

[0058] The cable drum segmentation unit is used to accurately segment the cable drum based on the detection results of the personnel behavior detection unit using the lightweight PP-LiteSeg model with high performance. It also accurately calculates the thickness of the cable to provide reliable data for subsequent production analysis;

[0059] The dynamic area calibration algorithm unit is used to dynamically and accurately divide the danger area and the warning area based on a preset algorithm.

[0060] The preset algorithms include:

[0061] The calculation formula of the edge width of the dangerous area A is: A = R + ΔL + 0.33D;

[0062] The calculation formula of the edge width of the warning area B is: B = A + 0.25D + K*v;

[0063] Among them, R represents the reference radius of the take-up reel, D is the dynamic detection value of the cable layer thickness, ΔL is the mechanical tolerance compensation, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of the personnel movement.

[0064] It should be added that the YOLOv5 algorithm plays a key role. Before being applied in this embodiment, it has gone through a large number of pre-training processes. The R & D personnel have collected image data of personnel in the cable production workshop covering diverse scenarios such as different lighting conditions, personnel postures, and styles of work clothing, and constructed a dedicated training dataset. By performing iterative training on this dataset and adjusting network parameters, the YOLOv5 algorithm can accurately identify the positions, actions, and behavioral states of personnel within the working area of the take-up machine. During actual operation, the personnel behavior detection unit receives the high-definition image data transmitted by the video acquisition module in real time. With the powerful object detection ability of the YOLOv5 algorithm, it can quickly lock on to personnel targets. Whether it is the normal equipment operation of workers or abnormal behaviors such as suddenly breaking into the dangerous area, they can all be captured in a timely manner, and the detection results are output in real time to provide basic data for subsequent processing.

[0065] The spool segmentation unit works based on the output results of the personnel behavior detection unit. The PP-LiteSeg model has also been pre-trained. The researchers have collected a large number of images of spools of different models and specifications in the cable production workshop, and marked key information such as spool contours and cable distributions to form training samples. During the training process, by optimizing the model parameters, the PP-LiteSeg model can accurately perform pixel-level segmentation on spools of different sizes and states while maintaining light weight and efficient operation. In actual application, the PP-LiteSeg model quickly identifies the spool boundaries based on the input image data, accurately segments the spool from the background, and then precisely calculates the thickness of the cable. Through the dynamic monitoring of the cable thickness on the spool, it can not only provide important data for production analysis such as cable winding uniformity and production progress, but also assist in judging whether the spool is approaching full load, providing a basis for the automatic control of the pay-off and take-up machine.

[0066] The dynamic area calibration algorithm unit, based on the data provided by the personnel behavior detection unit and the spool segmentation unit, combines preset algorithms to dynamically and accurately divide the dangerous area and the warning area. The dynamic area calibration algorithm unit will comprehensively consider parameters such as the reference radius of the take-up spool, the dynamic detection value of the cable layer thickness, the mechanical tolerance compensation amount, the personnel movement speed compensation coefficient, and the instantaneous speed of personnel movement, and calculate and delimit the corresponding area ranges in real time according to established mathematical formulas, such as the width of side A of the dangerous area: A = R + ΔL + 0.33D, and the width of side B of the warning area: B = A + 0.25D + K * v. As the cable thickness on the spool changes and personnel move within the workshop, the dangerous area and the warning area can be automatically adjusted dynamically to ensure that the safety protection area always fits the actual production scenario, providing accurate risk judgment basis for the alarm execution module, thereby realizing the dynamic and accurate management of the safety protection of the take-up machine.

[0067] In this embodiment, the alarm execution module takes an industrial voice lamp with three colors of red, yellow, and green as the core carrier to build a hierarchical early warning and emergency response system;

[0068] When the intelligent analysis module detects that a person enters the warning area and has not yet reached the boundary of the dangerous area, the alarm execution module automatically triggers the yellow warning light to flash continuously, accompanied by a low-frequency prompt sound, to remind the person of potential risks with acoustic and optical signals. At this time, the wire winding and unwinding machine still maintains a normal operating state;

[0069] When the intelligent analysis module detects that a person's torso crosses the boundary of the dangerous area and triggers the dangerous alarm condition, the alarm execution module immediately switches to the red warning mode, the red light flashes at a high frequency and emits a high-decibel sharp warning sound. At the same time, an emergency stop instruction is sent to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the wire winding and unwinding machine and starts the braking device to stop the equipment from operating within a very short time.

[0070] Embodiment Two:

[0071] Refer to Appendix Figure 2 , this embodiment proposes an intelligent vision early warning method for preventing personnel from being wound by a wire reel, which includes the following steps:

[0072] S1. Deploy a wide-angle high-definition camera in the working area of the wire winding machine, and collect the images of the working area of the wire winding machine in real time through the wide-angle high-definition camera to provide a data basis for subsequent analysis.

[0073] S2. Based on the images collected by the wide-angle high-definition camera, perform real-time and accurate detection of the personnel in the working area of the wire winding machine. Based on the detection results, perform accurate segmentation of the wire reel, and at the same time accurately calculate the thickness of the wire cable to dynamically and accurately divide the dangerous area and the warning area.

[0074] The specific steps involved in S2 include:

[0075] S2.1. Based on the images collected by the wide-angle high-definition camera, use the advanced YOLOv5 algorithm to perform real-time and accurate detection of the personnel in the working area of the wire winding machine, and capture various behavioral actions of the personnel in a timely manner;

[0076] S2.2. Based on the detection results, use the lightweight PP-LiteSeg model with high performance to achieve accurate segmentation of the wire reel. At the same time, accurately calculate the thickness of the wire cable to provide reliable data for subsequent production analysis;

[0077] S2.3. Dynamically and accurately divide the dangerous area and the warning area based on the preset algorithm.

[0078] The preset algorithm includes:

[0079] The formula for calculating the edge width of the dangerous area A is: A = R + ΔL + 0.33D;

[0080] The formula for calculating the edge width of the warning area B is: B = A + 0.25D + K * v;

[0081] Among them, R represents the reference radius of the take-up reel, D is the dynamic detection value of the cable layer thickness, ΔL is the mechanical tolerance compensation amount, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of personnel movement.

[0082] It should be added that the YOLOv5 algorithm plays a key role. Before being applied in this embodiment, it has gone through a large number of pre-training processes. The R & D personnel have collected image data of personnel in the cable production workshop covering diverse scenarios such as different lighting conditions, personnel postures, and work clothing styles, and constructed a dedicated training dataset. By performing iterative training on this dataset and adjusting the network parameters, the YOLOv5 algorithm can accurately identify the positions, actions, and behavior states of personnel within the working area of the take-up machine. During actual operation, after receiving the high-definition image data transmitted by the video acquisition module in real time, with the help of the powerful object detection ability of the YOLOv5 algorithm, it can quickly lock the personnel target. Whether it is the normal equipment operation of workers or the abnormal behavior of suddenly breaking into the dangerous area, it can be captured in time, and the detection results are output in real time, providing basic data for subsequent processing.

[0083] The PP-LiteSeg model has also been pre-trained. The researchers have collected a large number of images of different models and specifications of wire reels in the cable production workshop, and marked key information such as the wire reel contour and cable distribution to form training samples. During the training process, by optimizing the model parameters, the PP-LiteSeg model can accurately perform pixel-level segmentation on wire reels of different sizes and states while maintaining lightweight and efficient operation. In actual application, based on the input image data, the PP-LiteSeg model quickly identifies the wire reel boundary, accurately segments the wire reel from the background, and then precisely calculates the thickness of the cable. Through the dynamic monitoring of the cable thickness on the wire reel, it can not only provide important data such as cable winding uniformity and production progress for production analysis, but also assist in judging whether the wire reel is approaching full load, providing a basis for the automatic control of the wire take-up and pay-off machine.

[0084] Based on the foregoing data, the dangerous area and the warning area are dynamically and accurately divided in combination with a preset algorithm. By integrating parameters such as the reference radius of the take-up reel, the dynamically detected value of the cable layer thickness, the mechanical tolerance compensation amount, the personnel movement speed compensation coefficient, and the instantaneous speed of personnel movement, and according to a fixed mathematical formula, such as the width of side A of the dangerous area: A = R + ΔL + 0.33D, and the width of side B of the warning area: B = A + 0.25D + K * v, the corresponding area ranges are calculated and delimited in real time. With the change of the cable thickness on the reel and the movement of personnel in the workshop, the dangerous area and the warning area can be automatically adjusted dynamically to ensure that the safety protection area always fits the actual production scenario, providing an accurate risk judgment basis for the alarm execution module, thereby realizing the dynamic and accurate management of the safety protection of the take-up machine.

[0085] In this embodiment, the alarm execution module involved constructs a hierarchical early warning and emergency response system with an industrial voice lamp in three colors: red, yellow, and green as the core carrier;

[0086] When the intelligent analysis module detects that a person enters the warning area and has not reached the boundary of the dangerous area, the alarm execution module automatically triggers the yellow warning light to flash continuously and is accompanied by a low-frequency prompt sound, reminding the person of potential risks through the sound and light signals. At this time, the take-up and pay-off machine still maintains a normal operating state;

[0087] When the intelligent analysis module detects that a person's torso crosses the boundary of the dangerous area and triggers the dangerous alarm condition, the alarm execution module immediately switches to the red warning mode. The red light flashes frequently and emits a high-decibel sharp warning sound. At the same time, an emergency stop instruction is sent to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the take-up and pay-off machine and starts the braking device to stop the equipment from running within a very short time.

[0088] S3. Based on the division result, use an industrial voice lamp in three colors: red, yellow, and green to give hierarchical warnings. When it is detected that a person invades the dangerous area, link the PLC to control the take-up and pay-off machine to brake urgently.

[0089] The specific steps involved in step S3 include:

[0090] Pre-construct a hierarchical early warning and emergency response system with an industrial voice lamp in three colors: red, yellow, and green as the core carrier;

[0091] When it is detected that a person enters the warning area and has not reached the boundary of the dangerous area, automatically trigger the yellow warning light to flash continuously and be accompanied by a low-frequency prompt sound, reminding the person of potential risks through the sound and light signals. At this time, the take-up and pay-off machine still maintains a normal operating state;

[0092] When it is detected that the human torso crosses the boundary of the dangerous area and triggers the dangerous alarm condition, immediately switch to the red warning mode, with the red light flashing at a high frequency and emitting a high-decibel sharp warning sound. At the same time, send an emergency stop instruction to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the coiling and uncoiling machine and starts the braking device to stop the equipment from running within an extremely short time.

[0093] Taking the embodiment of a wire reel with a diameter of 1m as an example:

[0094] During initialization, set the radius R = 500mm and the safety margin ΔL = 25mm;

[0095] When the detected cable layer thickness D = 150mm, calculate the dangerous area A as:

[0096] A = 500 + 25 + 0.33 * 150 = 574.5mm;

[0097] If a person approaches at a speed of 1.5m / s, the warning area B expands to:

[0098] B = 574.5 + 0.25 * 150 + 0.6 * 1.5 = 619.5mm;

[0099] After continuous intrusion for more than 1 second, an audible and visual alarm is triggered, and the PLC control system completes the shutdown operation within 200ms.

[0100] In summary, it can be seen that by adopting the intelligent vision warning system and method for preventing personnel entanglement of a cable take-up rack of the present invention, the protection response time can be shortened to within 200ms, the false alarm rate can be reduced by 78%, and the adaptive time of the rewinding parameters can be less than 1 minute, which can significantly improve the active safety protection ability in industrial scenarios.

[0101] The above uses specific individual examples to elaborate in detail the principle and implementation manner of the present invention. These embodiments are only used to help understand the core technical content of the present invention. Based on the above specific embodiments of the present invention, those skilled in the art of the present technology, without departing from the principle of the present invention, any improvements and modifications made to the present invention shall fall within the scope of patent protection of the present invention.

Claims

1. An intelligent vision warning system for preventing personnel from being wound by a cable take-up rack, characterized in that, It includes: The video acquisition module uses a wide-angle high-definition camera deployed in the working area of the wire take-up machine to collect images of the working area of the wire take-up machine in real time, providing a data basis for subsequent analysis; The intelligent analysis module integrates YOLOv5 real-time personnel detection, PP-LiteSeg wire drum segmentation and thickness calculation functions, and uses a dynamic area calibration algorithm to dynamically define danger and warning areas based on preset parameters, providing accurate intelligent analysis support for the safety protection of the wire take-up machine; The alarm execution module, based on the analysis results of the intelligent analysis module, uses red, yellow and green industrial voice lights to provide graded warnings. When a person is detected intruding into the dangerous area, the PLC is linked to control the emergency braking of the wire take-up and traversing machine.

2. The intelligent vision warning system for preventing personnel entanglement of the cable take-up rack according to claim 1, wherein The intelligent analysis module specifically includes: The personnel behavior detection unit is used to receive the collected images from the video acquisition module and use the advanced YOLOv5 algorithm to perform real-time and accurate detection of personnel in the working area of the wire take-up machine, and timely capture various behaviors of personnel; The cable drum segmentation unit is used to accurately segment the cable drum based on the detection results of the personnel behavior detection unit using the lightweight PP-LiteSeg model with high performance. It also accurately calculates the thickness of the cable to provide reliable data for subsequent production analysis; The dynamic area calibration algorithm unit is used to dynamically and accurately divide the danger area and the warning area based on a preset algorithm.

3. The intelligent vision warning system for preventing personnel from being wound by the cable take-up rack according to claim 2, characterized in that The preset algorithm includes: The calculation formula of the edge width of the dangerous area A is: A = R + ΔL + 0.33D; The calculation formula of the edge width of the warning area B is: B = A + 0.25D + K*v; Among them, R represents the reference radius of the take-up reel, D is the dynamic detection value of the cable layer thickness, ΔL is the mechanical tolerance compensation, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of the personnel movement.

4. The intelligent vision warning system for preventing personnel from being wound by the cable take-up rack according to claim 1, wherein The alarm execution module uses the red, yellow and green industrial voice lights as the core carrier to build a hierarchical early warning and emergency response system; When the intelligent analysis module detects that a person has entered the warning area but has not yet touched the boundary of the danger zone, the alarm execution module automatically triggers the yellow warning light to flash continuously, accompanied by a low-frequency prompt sound, to remind the person of the potential risk with sound and light signals. At this time, the wire take-up and traversing machine remains in normal operation; When the intelligent analysis module detects that a person's torso has crossed the boundary of the danger zone and triggered a danger alarm condition, the alarm execution module immediately switches to the red warning mode, the red light flashes at a high frequency and emits a high-decibel sharp warning sound. At the same time, an emergency stop command is sent to the PLC control system through the control bus. After receiving the command, the PLC control system quickly cuts off the power output of the wire take-up and traversing machine and activates the braking device to stop the equipment in a very short time.

5. An intelligent vision warning method for preventing personnel from being wound by a cable take-up rack, characterized in that, The steps include: S1. Deploy a wide-angle HD camera in the working area of the wire take-up machine to collect images of the working area of the wire take-up machine in real time, providing a data basis for subsequent analysis; S2. Based on the wide-angle high-definition camera to collect images, the personnel in the working area of the wire take-up machine are detected in real time and accurately. Based on the detection results, the wire reel is accurately segmented, and the thickness of the cable is accurately calculated, and the danger area and warning area are dynamically and accurately divided; S3. Based on the partitioning result, use industrial voice lights in three colors of red, yellow, and green for hierarchical warning. When a person is detected to enter the dangerous area, link the PLC to control the wire winding and unwinding machine to make an emergency brake.

6. The intelligent vision warning method for preventing personnel from being wound by the cable take-up rack according to claim 5, characterized in that, The specific steps of step S2 include: S2.

1. Based on the images collected by the wide-angle high-definition camera, use the advanced YOLOv5 algorithm to perform real-time and accurate detection of the personnel in the working area of the wire winding machine, and capture various behavioral actions of the personnel in a timely manner; S2.

2. Based on the detection result, use the lightweight PP-LiteSeg model with high performance to achieve accurate segmentation of the wire reel. At the same time, accurately calculate the thickness of the cable to provide reliable data for subsequent production analysis; S2.

3. Dynamically and accurately divide the dangerous area and the warning area based on the preset algorithm.

7. The intelligent visual warning method for preventing personnel from being wound by the cable take-up rack according to claim 6, wherein The preset algorithm includes: The formula for the edge width of the dangerous area A is: A = R + ΔL + 0.33D; The formula for the edge width of the warning area B is: B = A + 0.25D + K * v; Among them, R represents the reference radius of the wire reel, D is the dynamically detected value of the cable layer thickness, ΔL is the mechanical tolerance compensation amount, K is the personnel movement speed compensation coefficient, and v is the instantaneous speed of the personnel movement.

8. The intelligent vision warning method for preventing personnel from being wound by the cable take-up rack according to claim 5, characterized in that, The specific steps of step S3 include: Pre-build a hierarchical early warning and emergency response system with industrial voice lights in three colors of red, yellow, and green as the core carriers; When it is detected that a person enters the warning area and has not yet touched the boundary of the dangerous area, automatically trigger the yellow warning light to flash continuously and be accompanied by a low-frequency reminder sound to remind the person of potential risks with light and sound signals. At this time, the wire winding and unwinding machine still maintains a normal operating state; When it is detected that a person's torso crosses the boundary of the dangerous area and triggers the danger alarm condition, immediately switch to the red warning mode. The red light flashes at a high frequency and emits a high-decibel sharp warning sound. At the same time, send an emergency stop instruction to the PLC control system through the control bus. After receiving the instruction, the PLC control system quickly cuts off the power output of the wire winding and unwinding machine, starts the braking device, and makes the equipment stop running within a very short time.

Images