Adaptive substation meter auxiliary identification device and method based on ambient light sensor

Abstract

The invention relates to a self-adaptive substation meter auxiliary recognition device and method based on an ambient light sensor, and the device comprises a control module, and a camera, the ambient light sensor and an LED light supplement module which are connected with the control module. The ambient light sensor is used for sensing the light intensity and color temperature of the surrounding environment of a substation meter to be detected in real time; the control module is used for controlling the LED light supplementing module to carry out light supplementing according to the light intensity and the color temperature of the surrounding environment of the transformer substation meter to be measured, receiving the meter image shot by the camera and automatically identifying the meter reading; the LED light supplementing module is used for receiving the adjusting instruction of the control module and supplementing light to the to-be-tested transformer substation meter; and the camera is used for shooting the to-be-measured transformer substation meter to obtain a meter image and transmitting the meter image to the control module. Compared with the prior art, the method has the advantages that the requirements of manual operation are greatly reduced while the efficiency and the accuracy of meter identification are improved, and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of power system automation equipment, and in particular to an adaptive substation meter auxiliary identification device and method based on an ambient light sensor. Background Art

[0002] With the deep integration of new-generation information technology and power grid operation and inspection services, the use of AI cameras to read meters instead of traditional manual meter reading has been increasingly applied in substations. For meter recognition in dim environments such as switchgear action counters, a camera with integrated fill light is currently commonly used to increase the brightness of the dial display area to assist the AI ​​camera in obtaining a clear dial image.

[0003] However, the fill light strategy in dim environments has led to new problems. Due to the influence of weather, day and night, season, lighting and other factors, the intensity of the meter's external light source will change dynamically, resulting in the following limitations of fixed fill light parameters: fixed fill light intensity leads to insufficient fill light or overexposure; a single color temperature cannot adapt to the color temperature requirements in different environments; the slow response speed cannot quickly adapt to the rapid changes in ambient light; and the high energy consumption affects the battery life of the meter's auxiliary identification device.

[0004] Therefore, developing a device that can automatically adjust the fill light according to the ambient light to ensure the stability and robustness of the image quality collected by the meter has become the key to improving the accuracy of meter recognition. Summary of the invention

[0005] The purpose of the present invention is to provide an adaptive substation meter auxiliary identification device and method based on an ambient light sensor, which improves the meter identification accuracy by automatic fill light.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] An adaptive substation meter auxiliary identification device based on an ambient light sensor comprises a control module and a camera, an ambient light sensor and an LED fill light module connected thereto.

[0008] Ambient light sensor: installed around the substation meter to be tested, the ambient light sensor is used to sense the light intensity and color temperature of the surrounding environment of the substation meter to be tested in real time;

[0009] Control module: used to process the light intensity and color temperature of the surrounding environment of the substation meter to be tested, send adjustment instructions to the LED fill light module according to the processing results to control the LED fill light module to perform fill light, and receive the meter image taken by the camera to automatically identify the meter reading;

[0010] LED fill light module: used to receive the adjustment instruction of the LED fill light module to fill light for the substation meter to be tested;

[0011] Camera: used to take pictures of the substation meters to be tested, obtain meter images, and transmit them to the control module.

[0012] Furthermore, the control module comprises:

[0013] Ambient light analysis unit: connected to the camera, ambient light sensor and LED fill light module, the ambient light analysis unit is used to calculate the optimal fill light intensity and color temperature based on the meter image taken by the camera, the light intensity and color temperature of the surrounding environment of the substation meter to be tested, and send corresponding adjustment instructions to the LED fill light module, wherein the intelligent fill light algorithm is processed based on a pre-trained machine learning model, and is trained through meter images under different environmental conditions and their corresponding optimal fill light parameters during the training process;

[0014] Image processing unit: connected to the camera, used for receiving the meter image and automatically identifying the meter reading.

[0015] Furthermore, the ambient light analysis unit is provided with a feedback mechanism, and the feedback mechanism is used to monitor the fill light effect in real time after issuing an adjustment instruction, and to perform fine adjustment according to the meter image taken by the camera.

[0016] Furthermore, the image processing unit comprises:

[0017] A preprocessing subunit: used for preprocessing the meter image, wherein the preprocessing operations include denoising, contrast enhancement, and geometric correction, wherein a non-local mean filter or a deep learning denoising model is used to remove random noise in the meter image while retaining the edge and texture details of the meter image. The contrast enhancement is a local contrast enhancement method, including one of a Retinex algorithm and a Gamma correction. The geometric correction includes one of an affine transformation and a perspective transformation, which is used to correct the angle and position of the meter image to ensure that the meter remains horizontal or vertical in the meter image.

[0018] Identification subunit: used to automatically identify meter readings based on the pre-processed meter image using image recognition algorithms.

[0019] Furthermore, the image recognition algorithm includes a feature extraction model, a convolutional neural network recognition model and a fusion module. The feature extraction model obtains the meter type, pointer position, and scale position of the meter through Hough circle detection, Hough line detection, and template matching methods. The convolutional neural network recognition model extracts the meter type, pointer position, and scale position, and the results of the feature extraction model and the convolutional neural network recognition model are input into the fusion module for multimodal fusion, and a comprehensive recognition result is output, including the readings corresponding to the meter pointer position and scale position.

[0020] Furthermore, the LED fill light module includes a plurality of LED lamps with different color temperatures and / or brightness levels to provide fill lights at different levels.

[0021] Furthermore, it also includes a storage module, which is respectively connected to the camera, the ambient light sensor and the LED fill light module, and is used to store the meter image collected by the camera, the light intensity and color temperature of the surrounding environment of the substation meter to be tested perceived by the ambient light sensor, and the fill light parameters set by the LED fill light module.

[0022] Furthermore, it also includes a communication module, which is connected to the storage module and is used to remotely transmit the data stored in the storage module and receive remote control instructions.

[0023] Furthermore, the communication module uses one or more wireless communication technologies including Wi-Fi, BLE, Zigbee, and 4G / 5G for remote communication.

[0024] The present invention also provides an auxiliary identification method according to the above-mentioned adaptive substation meter auxiliary identification device based on ambient light sensor, characterized in that it includes the following steps:

[0025] The ambient light sensor senses the light intensity and color temperature of the surrounding environment of the substation meter to be tested in real time, and the meter image before the camera fill light is transmitted to the ambient light analysis unit;

[0026] The ambient light analysis unit receives the meter image before fill light and the light intensity and color temperature of the surrounding environment of the substation meter to be tested, calculates the optimal fill light intensity and color temperature using an intelligent fill light algorithm, and sends corresponding adjustment instructions to the LED fill light module;

[0027] The LED fill light module receives the adjustment instruction and adjusts the LED light of the corresponding color temperature and / or brightness level according to the adjustment instruction;

[0028] The camera takes a picture of the substation meter to be tested after supplementary lighting, obtains the meter image, and transmits it to the image processing unit;

[0029] The preprocessing subunit in the image processing unit preprocesses the meter image first, and then transmits it to the recognition subunit. The recognition subunit processes the preprocessed meter image using an image recognition algorithm to recognize the meter reading.

[0030] The storage module receives and stores the meter image collected by the camera, the light intensity and color temperature of the surrounding environment of the substation meter to be tested perceived by the ambient light sensor, and the fill light parameters set by the LED fill light module;

[0031] The communication module remotely transmits the data stored in the storage module and receives remote control instructions.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] (1) The device of the present invention senses the light conditions of the surrounding environment of the substation meter in real time through the integrated ambient light sensor, and controls the LED fill light module to automatically adjust the fill light intensity and color temperature through the control module accordingly, thereby ensuring that the camera can obtain high-quality meter images under any light conditions, thereby improving the accuracy of meter recognition.

[0034] (2) The LED fill light module of the present invention is provided with LED lamps of different color temperatures and / or brightness levels, which can be graded to adjust the fill light according to the actual light conditions, and can flexibly select the most appropriate fill light scheme to ensure the best fill light effect.

[0035] (3) The communication module of the device of the present invention can transmit the collected data to the remote server in real time, support remote monitoring and data analysis, and realize intelligent management.

[0036] (4) The device of the present invention is not only suitable for indoor environments, but can also be used for equipment monitoring in complex outdoor environments, ensuring that high-quality images can be obtained in various application scenarios. In addition, in substations, the device can automatically identify the readings of various meters, thereby reducing the inspection pressure of operation and maintenance personnel and improving operation and maintenance efficiency and data real-time performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the device structure of the present invention. DETAILED DESCRIPTION

[0038] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0039] Example 1

[0040] This embodiment provides an adaptive substation meter auxiliary identification device based on an ambient light sensor, such as Figure 1 As shown, the device includes a camera 1, an ambient light sensor 2, an LED fill light module 3, a control module 4, a storage module 5 and a communication module 6, wherein the camera 1, the ambient light sensor 2, and the LED fill light module 3 are all connected to the control module 4 and the storage module 5, and the communication module 6 is connected to the storage module 5.

[0041] Ambient light sensor 2: This embodiment uses a high-precision ambient light sensor 2, which is installed around the substation meter to be tested and is used to sense the light intensity and color temperature of the environment around the substation meter to be tested in real time and transmit it to the control module 4.

[0042] Control module 4: includes two units, namely, ambient light analysis unit 41 and image processing unit 42. The ambient light analysis unit 41 is connected to the ambient light sensor 2 and the LED fill light module 3. The unit performs environmental analysis on the light intensity and color temperature data of the surrounding environment of the substation meter to be tested collected by the ambient light sensor 2 and the substation meter to be tested photographed by the camera 1, and uses the intelligent fill light algorithm to calculate the optimal fill light intensity and color temperature, and forms an adjustment instruction, which is sent to the LED fill light module 3 to control the state of the LED. The image processing unit 42 is connected to the camera 1. The unit mainly performs image recognition processing on the meter image collected by the camera 1 and identifies the meter reading in the image. The unit includes a preprocessing subunit 421 and a recognition subunit 422. The preprocessing subunit 421 is used to perform preprocessing operations such as denoising and contrast enhancement on the meter image. The recognition subunit 422 uses an image recognition algorithm to automatically recognize the meter reading, wherein the image recognition algorithm can use a feature extraction algorithm, etc. The input parameters received by the intelligent fill light algorithm include the meter image taken by the camera 1, the ambient color temperature and the ambient light intensity measured by the ambient light sensor 2. The intelligent fill light algorithm is input into the pre-trained machine learning model according to the above input parameters. The model is trained based on a large number of meter images under different environmental conditions and their corresponding optimal fill light settings, and can predict the most suitable fill light intensity and color temperature under the current conditions. According to the prediction results of the machine learning model, the ambient light analysis unit 41 outputs the fill light parameters. For example, if shadows or reflections are detected on the surface of the meter, the ambient light analysis unit 41 will adjust the fill light angle or intensity accordingly to eliminate these effects; if the ambient light color temperature is warm, the fill light color temperature will be appropriately increased to maintain the natural color of the image.

[0043] The ambient light analysis unit 41 also has a feedback mechanism, that is, after issuing an adjustment instruction, it monitors the fill light effect in real time and makes fine adjustments according to the actual image quality to ensure the best recognition effect.

[0044] Specifically, the preprocessing subunit (421) is used to preprocess the meter image, and the preprocessing operation includes denoising, contrast enhancement, and geometric correction. The preprocessing operation specifically includes denoising, contrast enhancement, and geometric correction. Among them, non-local mean filtering or deep learning denoising model is used to remove random noise in the image, while retaining the edge and texture details of the image to ensure that the meter scale and pointer are clearly visible. Local contrast enhancement technology, such as Retinex algorithm or Gamma correction, is used to further improve the visual quality of the image, so that the text, numbers and scales on the meter are more obvious, which is convenient for subsequent feature extraction and recognition. For images with incorrect shooting angles or perspective distortion, geometric transformations (such as affine transformations and perspective transformations) are used to correct the angle and position of the image to ensure that the meter remains horizontal or vertical in the image, which is convenient for subsequent pointer and scale analysis.

[0045] Identification subunit (422): used to automatically identify the meter reading and identify the meter based on the pre-processed meter image using an image recognition algorithm. The image recognition algorithm includes a feature extraction model, a convolutional neural network recognition model and a fusion module. The feature extraction algorithm obtains the meter type, pointer position and scale position of the meter through methods such as Hough circle detection, Hough line detection and template matching. The convolutional neural network recognition algorithm uses a convolutional neural network pre-trained with a large number of meter images under different environmental conditions and different scale values ​​to extract the meter type, pointer position and scale position through the convolutional neural network. The results obtained by the above two methods are fused through multi-modal fusion to output a comprehensive recognition result. Finally, the corresponding reading rules are selected according to the meter type to calculate the readings corresponding to the meter pointer position and scale position.

[0046] LED fill light module 3: includes multiple LED lights with different color temperatures and / or brightness levels, and controls the LED lights of different levels to turn on or off according to the adjustment instructions of the control module 4, so as to perform different levels of fill light, so that the substation meters to be tested are always under the optimal lighting conditions.

[0047] Camera 1: used to capture the meter image after the LED fill light module 3 adjusts the lighting conditions, and then transmit it to the control module 4 to realize automatic recognition of the meter reading.

[0048] Storage module 5: used to store the meter image collected by the camera 1, the light intensity and color temperature of the surrounding environment of the substation meter to be tested perceived by the ambient light sensor 2, and the fill light parameters set by the LED fill light module 3 for subsequent processing.

[0049] Communication module 6: used to transmit all data stored in storage module 5 to the remote server in real time, support remote monitoring and data analysis, realize intelligent management, and receive remote control instructions. Real-time transmission can adopt wireless communication technology such as Wi-Fi, BLE, Zigbee, 4G / 5G, etc.

[0050] The device is also designed with low power consumption, fault self-detection and alarm functions. The low power consumption design is mainly achieved through intelligent working mode control strategy. The device automatically enters sleep mode when idle, and supports low power communication protocol, which significantly reduces energy consumption and prolongs service life. The fault self-detection and alarm functions enable the device to have fault self-detection and alarm functions, and can send abnormal alarms to the remote server in time under abnormal circumstances to ensure the normal operation of the device.

[0051] The execution steps of the above auxiliary identification device are:

[0052] The ambient light sensor 2 senses the light intensity and color temperature of the surrounding environment of the substation meter to be tested in real time, and the camera 1 captures the meter image before fill-in light, and transmits it to the ambient light analysis unit 41;

[0053] The ambient light analysis unit 41 receives the meter image before fill light and the light intensity and color temperature of the surrounding environment of the substation meter to be tested, calculates the optimal fill light intensity and color temperature using an intelligent fill light algorithm, and sends corresponding adjustment instructions to the LED fill light module 3;

[0054] The LED fill light module 3 receives the adjustment instruction, and adjusts the LED light of the corresponding color temperature and / or brightness level according to the adjustment instruction;

[0055] The camera 1 takes a picture of the substation meter to be tested after supplementary lighting, obtains a meter image, and transmits it to the image processing unit 42;

[0056] The preprocessing subunit 421 in the image processing unit 42 preprocesses the meter image first, and then transmits it to the recognition subunit 422. The recognition subunit 422 processes the preprocessed meter image using an image recognition algorithm to automatically recognize the meter reading.

[0057] The storage module 5 receives and stores the meter image collected by the camera 1, the light intensity and color temperature of the surrounding environment of the substation meter to be tested sensed by the ambient light sensor 2, and the fill light parameters set by the LED fill light module 3;

[0058] The communication module 6 remotely transmits the data stored in the storage module 5 and receives remote control instructions.

[0059] Example 2

[0060] This embodiment provides an adaptive substation meter auxiliary identification device based on an ambient light sensor. Different from Embodiment 1, the storage module 5 in this embodiment is an external storage device.

[0061] The rest is the same as in Example 1.

[0062] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0063] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An adaptive substation meter auxiliary identification device based on ambient light sensor, characterized in that: It comprises a control module (4) and a camera (1), an ambient light sensor (2), and an LED fill light module (3) all connected thereto. Ambient light sensor (2): installed around the substation meter to be tested, the ambient light sensor (2) is used to sense the light intensity and color temperature of the surrounding environment of the substation meter to be tested in real time; A control module (4) is used to process the light intensity and color temperature of the surrounding environment of the substation meter to be tested, send an adjustment instruction to the LED fill light module (3) according to the processing result to control the LED fill light module (3) to perform fill light, and receive the meter image taken by the camera (1) to automatically identify the meter reading; LED fill light module (3): used for receiving adjustment instructions of the LED fill light module (3) to fill light for the substation meter to be tested; Camera (1): used to photograph the substation meter to be tested, obtain the meter image, and transmit it to the control module (4).

2. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 1 is characterized in that: The control module (4) comprises: An ambient light analysis unit (41) is connected to the camera (1), the ambient light sensor (2) and the LED fill light module (3), and is used to calculate the optimal fill light intensity and color temperature using an intelligent fill light algorithm based on the meter image captured by the camera (1) and the light intensity and color temperature of the environment surrounding the substation meter to be tested, and to send corresponding adjustment instructions to the LED fill light module (3), wherein the intelligent fill light algorithm is processed based on a pre-trained machine learning model, and is trained during the training process using meter images under different environmental conditions and their corresponding optimal fill light parameters; An image processing unit (42) is connected to the camera (1) and is used to receive the meter image and automatically identify the meter reading.

3. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 2 is characterized in that: The ambient light analysis unit (41) is provided with a feedback mechanism, which is used to monitor the fill light effect in real time after issuing an adjustment instruction, and to perform fine adjustments according to the meter image captured by the camera (1).

4. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 2 is characterized in that: The image processing unit (42) comprises: A preprocessing subunit (421) is used to preprocess the meter image, wherein the preprocessing operation includes denoising, contrast enhancement, and geometric correction, wherein a non-local mean filter or a deep learning denoising model is used to remove random noise in the meter image while retaining the edge and texture details of the meter image, the contrast enhancement is a local contrast enhancement method, including one of a Retinex algorithm and a Gamma correction, and the geometric correction includes one of an affine transformation and a perspective transformation, which is used to correct the angle and position of the meter image to ensure that the meter remains horizontal or vertical in the meter image; The identification subunit (422) is used to automatically identify the meter reading based on the pre-processed meter image by using an image recognition algorithm.

5. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 4 is characterized in that: The image recognition algorithm includes a feature extraction model, a convolutional neural network recognition model and a fusion module. The feature extraction model obtains the meter type, pointer position and scale position of the meter through Hough circle detection, Hough line detection and template matching methods. The convolutional neural network recognition model extracts the meter type, pointer position and scale position, and inputs the results of the feature extraction model and the convolutional neural network recognition model into the fusion module for multimodal fusion, and outputs a comprehensive recognition result, including the readings corresponding to the meter pointer position and the scale position.

6. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 1 is characterized in that: The LED fill light module (3) comprises a plurality of LED lamps with different color temperatures and / or brightness levels, so as to provide fill lights of different levels.

7. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 1 is characterized in that: It also comprises a storage module (5), which is respectively connected to the camera (1), the ambient light sensor (2) and the LED fill light module (3), and is used to store the meter image collected by the camera (1), the light intensity and color temperature of the surrounding environment of the substation meter to be tested perceived by the ambient light sensor (2), and the fill light parameters set by the LED fill light module (3).

8. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 7 is characterized in that: It also comprises a communication module (6), the communication module (6) being connected to the storage module (5), the communication module (6) being used for remotely transmitting the data stored in the storage module (5) and receiving remote control instructions.

9. The adaptive substation meter auxiliary identification device based on ambient light sensor according to claim 8 is characterized in that: The communication module (6) uses one or more wireless communication technologies including Wi-Fi, BLE, Zigbee, and 4G / 5G for remote communication.

10. An auxiliary identification method of the adaptive substation meter auxiliary identification device based on ambient light sensor according to any one of claims 1 to 9, characterized in that: The following steps are involved: The ambient light sensor (2) senses the light intensity and color temperature of the surrounding environment of the substation meter to be tested in real time, and the meter image before the camera (1) fills the light, and both are transmitted to the ambient light analysis unit (41); The ambient light analysis unit (41) receives the meter image before fill light and the light intensity and color temperature of the surrounding environment of the substation meter to be tested, calculates the optimal fill light intensity and color temperature using an intelligent fill light algorithm, and sends corresponding adjustment instructions to the LED fill light module (3); The LED fill light module (3) receives an adjustment instruction and adjusts the LED light of the corresponding color temperature and / or brightness level according to the adjustment instruction; The camera (1) photographs the substation meter to be tested after supplementary lighting, obtains a meter image, and transmits the image to an image processing unit (42); The preprocessing subunit (421) in the image processing unit (42) preprocesses the meter image first, and then transmits it to the recognition subunit (422). The recognition subunit (422) processes the preprocessed meter image using an image recognition algorithm to recognize the meter reading. The storage module (5) receives and stores the meter image collected by the camera (1), the light intensity and color temperature of the surrounding environment of the substation meter to be tested sensed by the ambient light sensor (2), and the fill light parameters set by the LED fill light module (3); The communication module (6) remotely transmits the data stored in the storage module (5) and receives remote control instructions.

Images