Electrical control system based on computer vision

Abstract

The invention discloses an electrical control system based on computer vision, and belongs to the technical field of computer vision and electrical control. The system comprises an infrared camera module, an image processing module, a control module and an FPGA hardware module. The infrared camera module is used for collecting an infrared image of a target area; the image processing module is used for carrying out target detection and recognition on the infrared image based on a lightweight target detection model and outputting category and position information of a target object; the control module is used for generating a control instruction according to the category and position information; and the FPGA hardware module is used for receiving and executing the control instruction. According to the method, a lightweight target detection model is combined with FPGA hardware acceleration, so that the efficiency and the real-time performance of target detection are improved, the recognition capability under a complex illumination condition is enhanced by using infrared imaging characteristics, and the method is suitable for scenes needing real-time intelligent control, such as smart home, industrial automation and energy-saving management.

Description

Technical Field

[0001] This invention relates to the fields of computer vision and electrical control technology, and specifically to an electrical control system based on computer vision. Background Technology

[0002] With the rapid development of artificial intelligence (AI) and the Internet of Things (IoT) technologies, computer vision-based intelligent control systems have been widely applied in fields such as smart homes, industrial automation, and energy management. Computer vision technology captures images or videos through cameras and uses deep learning algorithms to detect, classify, and recognize targets in the images, thereby achieving environmental perception and intelligent control. However, existing systems of this kind still have many shortcomings in practical applications.

[0003] The intelligent upgrade of existing systems faces multiple constraints: First, there is a contradiction between the real-time performance of target detection and the consumption of computing resources. Traditional target detection algorithms (such as Faster R-CNN and SSD) typically have high computational complexity and a large number of parameters, making it difficult to meet real-time requirements in embedded devices or low-power scenarios, thus limiting their application scope. Second, the intelligence level of infrared cameras is insufficient. Most existing infrared camera systems are limited to basic image acquisition and monitoring, failing to deeply integrate with advanced target detection algorithms and thus unable to achieve intelligent decision-making and control based on image content. Third, the electrical control system lacks flexibility. Existing control systems mostly rely on preset fixed rules or simple sensor triggers, making it difficult to adaptively adjust to complex and dynamic scene information (such as target category, quantity, and location), resulting in energy waste and affecting user experience. Finally, the advantages of hardware performance are not fully utilized. Although hardware acceleration technologies such as FPGAs can improve processing speed, their integration with target detection algorithms is low in existing systems, and they have not been optimized for specific algorithm models. Summary of the Invention

[0004] Technical problems to be solved

[0005] To address the shortcomings of existing technologies, this invention provides an electrical control system based on computer vision. It solves the problems of high computational complexity in traditional target detection algorithms, making it difficult to achieve real-time performance on embedded devices; it addresses the limitations of infrared cameras in simple image acquisition and monitoring, failing to integrate with intelligent algorithms to achieve more advanced target detection and control functions; it resolves the lack of flexibility in electrical control systems, making it difficult to dynamically adjust device status based on visual information; and it addresses the low integration of hardware acceleration modules with target detection algorithms, failing to fully leverage hardware performance advantages.

[0006] Technical solution

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: an electrical control system based on computer vision, including an infrared camera module, an image processing module, a control module and an FPGA hardware module.

[0008] The infrared camera module is used to capture infrared images of the target area. The infrared camera module includes a lens, an infrared sensor, and an image processing chip. The image processing chip preprocesses the acquired images and transmits them to the image processing module via a data interface. An adjustable mounting slot is provided at the bottom of the camera housing, and an angle adjustment mechanism is installed inside the adjustable mounting slot. The angle adjustment mechanism is detachably connected to the infrared camera bracket, facilitating installation and adjustment of the shooting angle.

[0009] The image processing module is communicatively connected to the infrared camera module and is used to receive infrared images and perform target detection and recognition based on a target detection model, generating target object category and location information. The image processing module includes an image preprocessing unit, a feature extraction unit, and a target detection unit. The image preprocessing unit is used to denoise, convert to grayscale, and resize the infrared image. The feature extraction unit is used to extract image features. The target detection unit generates target bounding boxes and category information based on a lightweight target detection model. The lightweight target detection model is an optimized model using pruning, quantization, or distillation techniques to reduce computational complexity and improve the real-time performance of target detection.

[0010] The control module is communicatively connected to the image processing module and is used to generate control commands based on the category and location information of the target object. The control module includes a rule base and a decision unit. The rule base stores the mapping relationship between target categories and electrical equipment control commands. The decision unit queries the rule base based on the target category and location information and generates control commands.

[0011] The FPGA hardware module is communicatively connected to the control module and is used to receive control commands and execute corresponding electrical control operations. The FPGA hardware module includes a signal receiving unit, a signal processing unit, and a control signal generation unit. The signal receiving unit receives control commands, the signal processing unit parses the control commands, and the control signal generation unit generates control signals for the electrical equipment. The FPGA hardware module can process multiple control commands in parallel to achieve synchronous or independent control of multiple electrical devices.

[0012] The system's workflow is as follows: the infrared camera module captures images of the target area and transmits them to the image processing module; the image processing module preprocesses the images and performs target detection and recognition based on a lightweight target detection model; the control module generates control commands based on the detection results; the FPGA hardware module receives the control commands and executes corresponding electrical control operations to control lighting equipment, air conditioning equipment, or power equipment.

[0013] Furthermore, the system also includes a status feedback unit, which is used to feed back the operating status of the electrical equipment to the control module to achieve closed-loop control.

[0014] Beneficial effects

[0015] This invention provides an electrical control system based on computer vision. It has the following advantages:

[0016] By adopting a lightweight target detection model and combining it with the accelerated processing of FPGA hardware modules, the computational complexity and latency of target detection are reduced, and the real-time response capability of the system is improved, making it suitable for embedded devices and low-power scenarios.

[0017] By combining an infrared camera module with an intelligent target detection algorithm, the application functions of the infrared camera are expanded, enabling it to go beyond basic image acquisition and achieve target recognition and classification in complex environments, thereby improving the system's environmental adaptability and detection accuracy.

[0018] Through the built-in rule base and decision unit of the control module, the system can dynamically generate control commands based on the target detection results, thereby achieving on-demand and precise control of electrical equipment and reducing energy waste.

[0019] By setting up a status feedback unit, the operating status of electrical equipment is fed back to the control module, forming a closed-loop control circuit, which enhances the control accuracy and operational reliability of the system.

[0020] By deeply integrating FPGA hardware modules with target detection algorithms, the advantages of hardware performance are fully utilized, thereby improving the overall efficiency of the system.

[0021] The system has a clear structure and a high degree of modularity. By replacing the rule base strategy or adjusting the target detection model, it can be easily applied to various scenarios such as smart homes, industrial automation, and energy management, and has good versatility and scalability. Attached Figure Description

[0022] Figure 1 System architecture diagram;

[0023] Figure 2 Schematic diagram of infrared camera acquisition module;

[0024] Figure 3 Flowchart of lightweight target detection model;

[0025] Figure 4 Schematic diagram of FPGA hardware control module;

[0026] Figure 5 Electrical appliance shutdown logic flowchart;

[0027] Figure 6 System structure diagram.

[0028] The components include: 1. Main control chassis; 2. Display panel mounting slot; 3. FPGA processing board; 4. Power module; 5. Heat dissipation and ventilation slot; 6. Interface connection slot; 7. LCD display screen; 8. Status indicator lights; 9. Operation buttons; 10. Infrared camera bracket; 11. USB interface; 12. Power interface; 13. Infrared camera module; 14. Camera housing; 15. Infrared lens; 16. Infrared sensor; 17. Image processing chip; 18. Data transmission line; 19. Adjustable mounting slot; 20. Angle adjustment mechanism; 21. Image processing server. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. (See attached drawings.) Figure 1-6 As shown, this embodiment of the invention provides an electrical control system based on computer vision, such as... Figure 1 As shown, the computer vision-based electrical control system of the present invention includes the following modules: an infrared camera module for capturing infrared images of the target area, supporting multispectral imaging to improve image quality; an image processing module for detecting and recognizing targets in the infrared images based on a target detection model, generating category and location information of the target object; a control module for generating control commands based on the category and location information of the target object, and implementing dynamic control through a rule base and decision unit; and an FPGA hardware module for receiving control commands and executing corresponding electrical control operations, including switching control and parameter adjustment of lighting equipment, air conditioning equipment, and power equipment. Through the collaboration between modules, the system realizes the entire process from image acquisition to electrical equipment control.

[0030] like Figure 2As shown, the infrared camera module includes: a lens for capturing images of the target area; an infrared sensor for sensing infrared spectral information, suitable for low-light or no-light environments; and an image processing chip for preliminary processing of the acquired images and transmitting them to the image processing module via a data interface. The infrared camera module connects to the image processing module via a USB interface and is powered by a power interface.

[0031] like Figure 3 As shown, the lightweight object detection model's processing flow includes the following steps: Input Image: Receive image data transmitted from the infrared camera module. Image Preprocessing: Denoise, convert to grayscale, and resize the image. Feature Extraction: Extract feature information from the image using a convolutional neural network. Object Detection: Generate the bounding box and category information of the target based on the head structure of the object detection model. Output Results: Transmit the detection results to the control module. Through lightweight optimization, the lightweight object detection model can achieve efficient object detection on embedded devices.

[0032] like Figure 4 As shown, the FPGA hardware control module includes: a signal receiving unit (receiving control commands generated by the control module), a signal processing unit (parses the control commands and generates control signals for the electrical equipment), a control signal generation unit (generating specific control signals for the electrical equipment based on the parsing results), and a status feedback unit (feedback the operating status of the electrical equipment to the control module). Through efficient hardware acceleration technology, the FPGA hardware module achieves rapid response and precise control of the electrical equipment.

[0033] like Figure 5 As shown, the specific process of the electrical appliance shutdown logic includes: 1. Target state detection: The image processing module detects the state of the target object in real time. 2. Condition judgment: Based on the detection results, it is determined whether the electrical equipment needs to be shut down. If the target meets the shutdown conditions, a shutdown command is generated. If no target is detected or the conditions are not met, the current state is maintained. 3. Shutdown operation: The shutdown command is transmitted to the FPGA hardware module to control the shutdown of the electrical equipment. 4. System state update: The operating status of the equipment is recorded and fed back to the control module. This logic process ensures the safety and energy efficiency of the electrical equipment.

[0034] like Figure 6As shown, this embodiment of the invention provides an electrical control system based on computer vision, including a main control chassis (1). A display panel mounting slot (2) is provided at the upper part of the front end of the main control chassis (1). An FPGA processing board (3) is fixedly connected to the middle of the inner wall of the main control chassis (1), and power modules (4) are fixedly connected to both sides of the inner wall of the main control chassis (1). An LCD display screen (7) is fixedly connected inside the display panel mounting slot (2). Multiple status indicator lights (8) are connected to the middle of the lower surface of the LCD display screen (7), and the status indicator lights (8) are electrically connected to the FPGA processing board (3). Operation buttons (9) are fixedly connected to the right side of the LCD display screen (7) panel. The main control chassis (1) is connected to the infrared camera module (13) and the image processing server (21) via a network cable. The infrared camera module (13) includes a camera housing (14), an infrared lens (15) is fixedly connected to the front end inside the camera housing (14), an infrared sensor (16) is fixedly connected to the middle part inside the camera housing (14), and an image processing chip (17) is connected to the rear end of the infrared sensor (16) via a network cable. The image processing chip (17) is electrically connected to the image processing server (21) or the FPGA processing board (3) inside the main control chassis (1) via a data transmission line (18). The bottom of the camera housing (14) is provided with an adjustable mounting slot (19), and an angle adjustment mechanism (20) is provided inside the adjustable mounting slot (19). The angle adjustment mechanism (20) is detachably connected to the infrared camera bracket (10). The control signal output terminal of the FPGA processing board (3) is electrically connected to the relay (21) through a circuit. The relay (21) is connected in series to the power supply circuit of the lighting equipment, air conditioning equipment or power equipment to realize the switching control of the equipment.

[0035] Specific Implementation in a Garage Environment: In a garage environment, the computer vision-based electrical control system of the present invention can be implemented as follows: An infrared camera module is installed at the garage entrance and parking space area to collect infrared image information in the garage in real time. The infrared camera can work normally in low light or no light environments to ensure all-weather monitoring. The image processing module preprocesses the collected infrared images, including noise reduction, grayscale conversion, and size adjustment. Subsequently, a lightweight target detection model is used to detect and identify people and vehicles in the images, outputting target category and location information. The control module determines whether there is activity of people or vehicles in the garage based on the recognition results output by the image processing module. If no one is detected in the garage and no vehicle is parked, a control command to turn off the lighting and ventilation equipment is automatically generated; if people or vehicles are detected entering the garage, a control command to turn on the lighting and ventilation equipment is automatically generated. The FPGA hardware module receives the control commands generated by the control module and controls the switching on and off of the lighting and ventilation equipment in the garage. The FPGA hardware module can realize parallel control of multiple electrical devices, ensuring response speed and control accuracy. The system can be further integrated with a status feedback unit to provide real-time feedback on the operating status of equipment within the garage to the control module, enabling closed-loop control and remote monitoring. Through the above implementation methods, the system can achieve intelligent management and energy-saving control of the garage environment, improving safety and user experience.

[0036] Specific Implementation in an Intersection Scenarios: In an intersection scenario, the computer vision-based electrical control system of this invention can be implemented as follows: An infrared camera module is deployed at an appropriate location at the urban intersection to collect infrared image information of the intersection area in real time. The infrared camera can operate normally at night or in inclement weather conditions, ensuring all-weather monitoring of traffic flow. The image processing module preprocesses the collected infrared images, including noise reduction, grayscale conversion, and size adjustment. Subsequently, a lightweight target detection model is used to detect and classify pedestrians, non-motorized vehicles, and motorized vehicles in the image, outputting the quantity and location information of each type of traffic participant. Based on the recognition results output by the image processing module, combined with real-time traffic flow and target type, the control module automatically generates control commands to adjust the intersection lighting, traffic lights, and warning devices. For example, when an increase in pedestrian flow is detected at night, the intersection lighting brightness is automatically increased; when traffic congestion is detected, the traffic light cycle is automatically adjusted or the warning devices are activated. The FPGA hardware module receives the control commands generated by the control module and quickly responds to and controls the intersection's lighting, traffic lights, and warning devices. The FPGA hardware module enables parallel control of multiple devices, ensuring real-time performance and safety in traffic management. The system can integrate a status feedback unit to provide real-time feedback of the operational status of each device at the intersection to the control module, achieving closed-loop control and remote traffic management. Through these implementation methods, the system achieves intelligent traffic management at intersections, improving traffic safety and operational efficiency.

[0037] Specific Implementation in an Office Building Scenario: In an office building scenario, the computer vision-based electrical control system of this invention can be implemented as follows: Infrared camera modules are installed in key areas such as the office building entrance, corridors, and meeting rooms to collect infrared image information in real time. The infrared cameras can work stably under different lighting conditions, ensuring all-weather monitoring of personnel flow. The image processing module preprocesses the collected infrared images, including noise reduction, grayscale conversion, and size adjustment. Subsequently, a lightweight target detection model is used to detect and identify personnel in the images, outputting information such as the number, distribution, and activity status of personnel. Based on the recognition results output by the image processing module and combined with the actual usage needs of the office building, the control module automatically generates control commands to adjust lighting, air conditioning, and power equipment. For example, when no one is detected in the meeting room, the lighting and air conditioning equipment are automatically turned off; when increased personnel flow is detected in the corridor, the lighting brightness is automatically increased or the air conditioning operating parameters are adjusted. The FPGA hardware module receives the control commands generated by the control module and controls the switching and parameter adjustment of lighting, air conditioning, and power equipment in the office building. The FPGA hardware module can realize parallel control of multiple devices, ensuring the comfort and energy efficiency of the office environment. The system can integrate a status feedback unit to provide real-time feedback of the operating status of each device to the control module, enabling closed-loop control and remote building management. Through the above implementation methods, the system can achieve intelligent management and energy-saving operation of the office building environment, improving office comfort and management efficiency.

[0038] Specific Implementation of Smart Home Scenarios: In a smart home scenario, the computer vision-based electrical control system of this invention can be implemented as follows: Infrared camera modules are installed in key areas such as the front door, living room, and bedroom to collect infrared image information of family members' activities in real time. The infrared cameras can work stably under different lighting conditions, ensuring all-weather monitoring of family members' activities. The image processing module preprocesses the collected infrared images, including noise reduction, grayscale conversion, and size adjustment. Subsequently, a lightweight target detection model is used to detect and identify family members or visitors in the images, outputting information such as the identity, number, and activity status of the individuals. Based on the recognition results output by the image processing module, and combined with the activity status and personalized needs of family members, the control module automatically generates control commands for adjusting home appliances such as lighting and air conditioning. For example, when a family member is detected entering the living room, the lighting and air conditioning are automatically turned on; when no one is detected in the bedroom, the relevant equipment is automatically turned off to save energy. The FPGA hardware module receives the control commands generated by the control module and controls the switching and parameter adjustment of lighting, air conditioning, and other equipment in the home environment. The FPGA hardware module can realize parallel control of multiple devices, ensuring the comfort and intelligence level of the home environment. The system can integrate a status feedback unit to provide real-time feedback on the operating status of each home appliance to the control module, enabling closed-loop control and remote home management. Through the above implementation methods, the system can achieve personalized and intelligent management of the smart home environment, improving living comfort and energy efficiency.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An electrical control system based on computer vision, characterized in that, The system includes an infrared camera module (13), an image processing module, a control module, and an FPGA hardware module (3). The infrared camera module (13) is used to acquire infrared images of the target area. The image processing module is communicatively connected to the infrared camera module (13) and is used to perform target detection and recognition on the infrared images based on a target detection model. The control module is communicatively connected to the image processing module and is used to generate electrical equipment control commands based on the recognition results. The FPGA hardware module (3) is communicatively connected to the control module and is used to receive the control commands and execute the control operations of the electrical equipment.

2. The computer vision-based electrical control system according to claim 1, characterized in that, The target detection model is a lightweight target detection model.

3. The computer vision-based electrical control system according to claim 2, characterized in that, The lightweight target detection model is optimized using pruning, quantization, or distillation techniques.

4. The computer vision-based electrical control system according to claim 1, characterized in that, The infrared camera module (13) includes a lens (15), an infrared sensor (16), and an image processing chip (17). The image processing chip (17) is used to preprocess the acquired images and transmit them to the image processing module through a data interface.

5. An electrical control system based on computer vision according to claim 1, characterized in that, The image processing module is configured to include an image preprocessing unit, a feature extraction unit, and a target detection unit. The image preprocessing unit is used to denoise, convert to grayscale, and resize the infrared image. The feature extraction unit is used to extract image features. The target detection unit generates target bounding boxes and category information based on the target detection model.

6. The computer vision-based electrical control system according to claim 1, characterized in that, The control module is configured to include a rule base and a decision unit. The rule base is used to store the mapping relationship between target categories and electrical equipment control instructions. The decision unit is used to query the rule base and generate control instructions based on the target category and location information.

7. The computer vision-based electrical control system according to claim 1, characterized in that, The FPGA hardware module (3) includes a signal receiving unit, a signal processing unit, and a control signal generation unit. The signal receiving unit is used to receive control commands, the signal processing unit is used to parse control commands, and the control signal generation unit is used to generate control signals for electrical equipment.

8. An electrical control system based on computer vision according to claim 1, characterized in that, The system further includes a status feedback unit, which is used to feed back the operating status of the electrical equipment to the control module to achieve closed-loop control.

9. An electrical control system based on computer vision according to claim 1, characterized in that, The system is applied to a garage environment. The infrared camera module (13) collects images of the garage. The image processing module identifies people and vehicles in the garage. The control module automatically controls the lighting and ventilation equipment according to the identification results. The FPGA hardware module (3) implements the on / off control of the equipment.

10. An electrical control system based on computer vision according to claim 1, characterized in that, The system is applied to smart home or office building scenarios. The infrared camera module (13) collects images of people's activities. The image processing module identifies the identity and number of people. The control module automatically adjusts lighting, air conditioning and power equipment according to the activity status. The FPGA hardware module (3) realizes parallel control of multiple devices.

Images