Fault detection method of light equipment and light control system
The fault detection method for lighting equipment in private venues uses image analysis and machine learning to automatically identify and adjust lighting issues, enhancing maintenance efficiency and user experience by ensuring consistent lighting effects.
Patent Information
- Application Number
- CN202410966457.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-18
- Publication Date
- 2024-12-06
AI Technical Summary
Private entertainment venues face challenges in efficiently detecting and maintaining lighting equipment due to complex layouts and diverse brands, leading to issues like damage, reduced brightness, and uneven color, which affect lighting quality and user experience.
A fault detection method using cameras to capture lighting equipment images, analyze brightness and chromaticity features, and employ image processing techniques like histogram comparison and machine learning models to identify and locate faults, enabling automatic detection and adjustment of lighting effects.
Enhances maintenance efficiency by quickly identifying faults, reducing manual labor, and ensuring consistent lighting effects, thereby improving user experience and extending equipment life while lowering maintenance costs.
Smart Images

Figure CN119095241A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of lighting equipment, and in particular to a fault detection method for lighting equipment and a lighting control system. Background Art
[0002] With the continuous growth of the economy and the diversification of consumer demand, the entertainment industry has ushered in an unprecedented prosperity. In particular, private entertainment venues such as audio-visual rooms, mini KTVs, and mini theaters have been warmly welcomed by the majority of users because they provide a private and comfortable environment where people can freely express themselves and enjoy music and movies.
[0003] In these entertainment venues, in addition to sound effects, lighting effects are another key factor in creating atmosphere and enhancing user experience. Through carefully designed lighting effects, not only can a romantic, tranquil, exciting or other emotional atmosphere be created for users, but it can also be synchronized with the rhythm of music to bring users an immersive experience.
[0004] However, these private entertainment venues are compact and relatively closed. To achieve high-quality lighting effects, the on-site lighting layout is usually complicated, with a large number of lighting equipment. Usually, a large number of lighting units such as lamps or LED lights need to be installed in a very small space. These lighting equipment are distributed in multiple locations and are very complex. Under long-term, high-intensity use, it is easy to have problems such as damage to the lighting unit, reduced brightness or uneven color. Once these failures occur, it will not only affect the overall lighting effect, but also reduce the user experience. On-site maintenance and debugging work will become very difficult and time-consuming. In addition, lighting equipment of different brands and models may have different characteristics and service lives, which also brings additional challenges to the maintenance and replacement of lighting equipment.
[0005] Based on this, how to effectively detect and maintain the lighting equipment in private entertainment venues to ensure that they continuously and stably provide high-quality lighting effects has become an issue that the industry urgently needs to solve. Summary of the invention
[0006] In order to solve one of the above-mentioned defects, the present application provides a fault detection method and a lighting control system for lighting equipment to improve the fault detection efficiency of the lighting equipment.
[0007] A method for detecting a fault of a lighting device, comprising:
[0008] When the lighting device is in a lighting state, using a camera to capture a lighting device image including the lighting device;
[0009] Analyze the lighting equipment image and calculate the corresponding real-time brightness feature, and determine the real-time brightness feature according to a preset brightness threshold to obtain a brightness detection result;
[0010] Using a color histogram comparison method to detect chromaticity features of the lighting equipment image to obtain a chromaticity detection result;
[0011] The fault state of the lighting equipment is determined according to the brightness detection result and the chromaticity detection result.
[0012] In one embodiment, analyzing the lighting equipment image and calculating the corresponding real-time brightness feature, and determining the real-time brightness feature according to a preset brightness threshold to obtain a brightness detection result includes:
[0013] Performing color space conversion on the lighting equipment image;
[0014] Performing local contrast enhancement processing on the converted lighting equipment image;
[0015] Extract brightness features of lighting equipment images;
[0016] The extracted brightness features are equalized, and pixels whose brightness exceeds a preset brightness threshold are marked as brightness abnormal points.
[0017] In one embodiment, extracting brightness features of a lighting device image includes:
[0018]
[0019] Where L represents the average brightness of the lighting device image, I ij is the pixel value of the lighting device image at position (x, j), W and H are the width and height of the lighting device image respectively, w ij is the weight at position (x, j);
[0020] The equalization processing of the extracted brightness features includes:
[0021]
[0022] I enhanced =CLAHE (Igray)
[0023] Wherein, Igray(x,y) is the grayscale image of the lighting equipment image, CLAHE is the application of contrast-limited adaptive histogram equalization algorithm to enhance image contrast, and x and y represent the horizontal and vertical coordinates of the image, respectively.
[0024] In one embodiment, the color histogram comparison method is used to detect the chromaticity feature of the lighting device image to obtain the chromaticity detection result, including:
[0025] Obtaining a normalized histogram of the lighting device image;
[0026] Calculating the cross entropy of the normalized histogram of the lighting device image according to a preset standard color histogram;
[0027] The chromaticity detection result is obtained by comparing the cross entropy with the preset threshold.
[0028] In one embodiment, calculating the cross entropy of the normalized histogram of the lighting device image according to a preset standard color histogram includes:
[0029]
[0030] P={p1,p2,...,p3}
[0031] Q={q1,q2,...,q3}
[0032] Among them, H(P,Q) is the cross entropy, P is the preset standard color histogram, and Q is the normalized histogram of the lighting equipment image.
[0033] In one embodiment, the lighting equipment fault detection method further includes:
[0034] Divide the lighting equipment image into a plurality of image blocks, and extract identification feature information of each image block; wherein the identification feature information includes the brightness mean and brightness standard deviation and the chromaticity mean and chromaticity standard deviation of the image block;
[0035] Identifying a faulty image block from each image block according to the identification feature information;
[0036] The fault area of the lighting device is located according to the fault image block.
[0037] In one embodiment, identifying a faulty image block from each image block according to the identification feature information includes:
[0038] Selecting a matching lighting fault detection model according to the recognition accuracy, inputting the recognition feature information of the image block into the lighting fault detection model to perform fault classification and recognition to obtain a fault image block;
[0039] Wherein, the lighting fault detection model is a pre-trained classification model for identifying image brightness and chromaticity faults;
[0040] In one embodiment, the lighting fault detection model includes a binary classification SVM model of a support vector machine, and the SVM model training process includes:
[0041]
[0042] subject.to.y i (w·x i +b)≥1-ζ i ,ζ i ≥0
[0043] Among them, w and b are the parameters of the SVM model, C is the penalty term coefficient, and f(·) represents the SVM classification function.
[0044] In one embodiment, the lighting fault detection model includes a CNN classifier using a convolutional neural network, and the training process of the CNN classifier includes:
[0045]
[0046] in, Represents the predicted category probability, N represents the number of training samples, and y represents the output category probability.
[0047] In one embodiment, the lighting fault detection model includes a semantic segmentation network based on deep learning, and the melody process of the semantic segmentation network includes:
[0048]
[0049] Among them, L(·) represents the loss function, Represents the category of each pixel predicted by the semantic segmentation network, y i represents the true label, and N represents the number of training samples.
[0050] In one embodiment, identifying a faulty image block from each image block according to the identification feature information includes:
[0051] Detect the brightness and chromaticity of each image block and calculate the corresponding brightness error and chromaticity error;
[0052]
[0053] Among them, L ij Indicates brightness and C ij Indicates chromaticity; ΔL ij =|L ij -L 标准 | represents the brightness error of the image block, ΔC ij =|C ij -C 标准 | represents the chrominance error of the image block;
[0054] The brightness error and the chromaticity error are judged by using a preset recognition threshold to determine the faulty image block.
[0055] In one embodiment, the lighting equipment fault detection method further includes:
[0056] Obtaining standard state parameters of a lamp in the lighting device, wherein the standard state parameters represent the brightness and chromaticity corresponding to the normal operation of the lamp;
[0057] Acquire actual state parameters corresponding to the lamp tube in the lighting device; wherein the actual state parameters are the brightness and chromaticity corresponding to the lamp tube in the lighting device image;
[0058] The faulty lamp tube in the fault state is determined by using the preset state parameter threshold to judge the state parameter deviation value of the lamp tube.
[0059] In one embodiment, determining a faulty lamp in a faulty state by using a preset state parameter threshold to determine a state parameter deviation value of the lamp includes:
[0060] ΔC=|C R -C S
[0061] ΔL=|L R -L S
[0062] Among them, ΔL represents the brightness deviation value, ΔC represents the chromaticity deviation value, L R Indicates the brightness of the standard state parameter, L S Indicates the brightness of the actual state parameter; C R Indicates the chromaticity of the standard state parameter, C S Indicates the chromaticity of the actual state parameter.
[0063] In one embodiment, the lighting equipment fault detection method is characterized by further comprising:
[0064] Using at least two cameras to respectively capture a lighting equipment image including the lighting equipment to identify the faulty lamp tube and its position information in the lighting equipment image;
[0065] Determining the camera parameters of the camera according to the coordinate data of the preset calibration point image in the lighting equipment image;
[0066] The actual position of the faulty lamp is obtained by performing triangulation based on the position information of the faulty lamp identified by the two lighting equipment images and the corresponding camera parameters.
[0067] In one embodiment, the lighting equipment fault detection method further includes:
[0068] According to the lighting equipment image captured by the camera, the lighting effect of the scene within a certain range of the faulty lamp is obtained in real time;
[0069] According to the actual position of the faulty lamp tube, several lamp tubes around it are selected as target lamp tubes; the brightness and chromaticity of the target lamp tubes are adjusted, and the adjusted lighting effect is observed in real time according to the lighting equipment image taken by the camera until the adjusted lighting effect reaches a preset consistency range.
[0070] A lighting control system, characterized in that it comprises: a camera and a control device; wherein the control device is connected to the lighting device;
[0071] The camera is used to capture images of lighting equipment;
[0072] The control device is used to receive the lighting device image and execute the lighting device fault detection method to perform fault detection on the lighting device.
[0073] The technical solution of the above-mentioned embodiment is to use a camera to capture an image of the lighting equipment including the lighting equipment when the lighting equipment is in a lighting state, obtain corresponding real-time brightness characteristics and brightness detection results and real-time chromaticity characteristics and chromaticity detection results through the lighting equipment image, and determine the fault state of the lighting equipment according to the brightness detection results and chromaticity detection results. This technical solution automatically detects faults by using the image of the lighting equipment captured by the camera, quickly detects the fault state of the lighting equipment, reduces the time and labor intensity of manual inspection, and thus improves the maintenance efficiency of the lighting equipment.
[0074] Furthermore, by dividing the lighting equipment image into multiple image blocks for detection, the fault area can be located quickly and accurately, which provides important technical support for quickly determining the fault location and troubleshooting, and further improves the maintenance efficiency of the lighting equipment.
[0075] Furthermore, by detecting the status of the lamp tube, a faulty lamp tube can be quickly found and automatically located, thereby reducing the need for frequent manual maintenance and lowering the maintenance labor cost.
[0076] Furthermore, by performing fault analysis on at least two of the lighting equipment images, the faulty lamp tube and its position information in the lighting equipment image are identified, and combined with the solved camera parameters, the actual position of the faulty lamp tube is determined using a triangulation positioning method; the actual position of the lamp tube can be accurately found, providing important technical support for troubleshooting and facilitating rapid troubleshooting.
[0077] Furthermore, by adjusting the lamps around the faulty lamp, the consistency of lighting effects can be achieved, ensuring the stability and consistency of the lighting effects of the lighting equipment, enhancing the user experience, and through automatic detection and adjustment, it can ensure that the lighting effects on site are always in the best state, providing users with a continuous, stable, high-quality entertainment experience.
[0078] Additional aspects and advantages of the present application will be partially given in the following description, which will become apparent from the following description, or will be understood through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0079] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0080] Figure 1 This is a schematic diagram of the hardware environment for fault detection of a lighting device;
[0081] Figure 2 is a flow chart of a method for detecting a fault of a lighting device according to an embodiment;
[0082] Figure 3 is an electrical structure diagram of a lighting control system of an embodiment;
[0083] Figure 4 It is an example of an algorithm module diagram for controlling the operation of a device. DETAILED DESCRIPTION
[0084] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.
[0085] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, but does not exclude the presence or addition of one or more other features, integers, steps, operations.
[0086] refer to Figure 1 As shown, Figure 1 This is a schematic diagram of the hardware environment for fault detection of an example lighting device. At least one camera is used to capture a lighting device image of the lighting effect of the lighting device. The control system recognizes and analyzes the lighting device image and its lighting parameters, automatically detects faults, and quickly detects the fault status of the lighting device, thereby reducing the time and labor intensity of manual inspection, thereby improving the maintenance efficiency of the lighting equipment.
[0087] The technical solution of the present application can be applied to various entertainment venues where lighting equipment is installed, such as audio-visual rooms, mini KTVs, mini theaters, etc., to solve the common problems of damage to lighting equipment, uneven brightness or inconsistent chromaticity. The following embodiments will be described using an audio-visual room as an example; Figure 2 As shown, Figure 2 The present invention is a flowchart of a method for detecting a fault of a lighting device according to an embodiment of the present invention, comprising:
[0088] Step S10: when the lighting device is in a lighting state, a camera is used to capture a lighting device image including the lighting device.
[0089] In this step, at least one camera is set at a set position in the audio-visual room to photograph the lighting equipment when the lighting equipment is in a lighting state to obtain an image of the lighting equipment. For example, a high-definition camera can be used to capture the details of the lighting effect.
[0090] Step S20, analyzing the lighting equipment image and calculating the corresponding real-time brightness feature, and determining the real-time brightness feature according to a preset brightness threshold to obtain a brightness detection result.
[0091] In this step, the lighting equipment image is analyzed and the corresponding real-time brightness features are calculated, and then the real-time brightness features are judged using a preset brightness threshold to finally obtain the brightness detection result.
[0092] For example, in brightness detection, the brightness contribution of each pixel in the lighting device image can be obtained, and the formula is as follows:
[0093]
[0094] Taking the weight into account, the brightness calculation formula is as follows:
[0095]
[0096] Where L represents the average brightness of the lighting device image, I ij is the pixel value of the lighting device image at position (x, j), W and H are the width and height of the lighting device image respectively, w ij is the weight at position (x, j);
[0097] In one embodiment, the method of calculating the real-time brightness feature and obtaining the brightness detection result in step S10 may include the following:
[0098] S201, performing color space conversion on the lighting equipment image.
[0099] Specifically, the color lighting device image can be converted into a grayscale image or other color space suitable for brightness detection, such as YUV; for example, the lighting image is converted into a grayscale image Igray(x,y), where x and y represent the horizontal and vertical coordinates of the lighting device image, respectively.
[0100] S202, performing local contrast enhancement processing on the converted lighting equipment image.
[0101] Exemplarily, the CLAHE (Contrast Limited Adaptive Histogram Equalization) algorithm is applied to enhance the local contrast of the image so that the brightness feature is more prominent; the local enhancement process may include the following:
[0102] I enhanced =CLAHE (Igray)
[0103] Where Igray(x,y) is the grayscale image of the lighting equipment image, CLAHE is the application of contrast-limited adaptive histogram equalization algorithm to enhance image contrast, and I enhanced represents the enhanced image, x and y represent the horizontal and vertical coordinates of the image respectively.
[0104] S203, extracting brightness features of the lighting equipment image.
[0105] Specifically, the brightness feature of the enhanced lighting device image is extracted, for example, the pixel mean μ can be used as the brightness feature.
[0106] S204, performing equalization processing on the extracted brightness features, and marking the pixel points whose brightness exceeds a preset brightness threshold as brightness abnormal points.
[0107] In this step, the brightness characteristics of the lighting equipment image are globally or locally equalized to ensure that the image brightness distribution is within an appropriate range.
[0108] Exemplarily, it can be achieved by simple linear stretching, and the equalization process may include the following:
[0109]
[0110] Among them, I equalized (x, y) represents the enhanced image of the equalized lighting device image, and x and y represent the horizontal and vertical coordinates of the image respectively.
[0111] When analyzing brightness anomalies, the image of the lighting equipment after brightness equalization can be analyzed. In addition, it can be determined whether further processing or extraction of other information is required based on application requirements; for example, a preset brightness threshold T can be set and pixels whose brightness exceeds the threshold T can be marked as brightness anomalies.
[0112] The technology of the above-mentioned embodiment can realize accurate extraction of brightness features of lighting equipment images, and can accurately obtain the lighting effect details of the lighting equipment, which can be used to judge abnormal lighting brightness and accurately identify brightness fault states in lighting equipment.
[0113] Step S30: Using a color histogram comparison method to detect chromaticity features of the lighting equipment image to obtain a chromaticity detection result.
[0114] In this step, the color histogram comparison method is used to calculate the cross entropy of the normalized histogram to evaluate the chromaticity consistency of the light image.
[0115] In one embodiment, the method of detecting chromaticity features to obtain chromaticity detection results in step S30 may include the following steps:
[0116] S301, obtaining a normalized histogram of the lighting equipment image.
[0117] For example, the preset standard color histogram is P = {p1, p2, ..., p3}, and the normalized histogram of the actual lighting equipment image is Q = {q1, q2, ..., q3}.
[0118] S302, calculating the cross entropy of the normalized histogram of the lighting device image according to a preset standard color histogram. The cross entropy calculation formula is as follows:
[0119]
[0120] S303, obtaining a chromaticity detection result by comparing the cross entropy with a preset threshold.
[0121] Whether the chromaticity of the lighting equipment image is consistent can be determined by comparing the cross entropy with a preset threshold, thereby obtaining a chromaticity detection result.
[0122] As in the technical solution of the above-mentioned embodiment, a color histogram comparison method is used to analyze the chromaticity details of the light image, and based on the judgment of the chromaticity features, accurate detection of the chromaticity feature fault state of the lighting equipment is achieved.
[0123] Step S40: determining the fault state of the lighting device according to the brightness detection result and the chromaticity detection result.
[0124] In this step, a comprehensive judgment is made through the brightness detection results and the chromaticity detection results, the real-time lighting effects of the lighting equipment are accurately analyzed, and finally the fault status of the lighting equipment is determined.
[0125] As in the technical solution of the above-mentioned embodiment, when the lighting equipment is in a lighting state, a camera is used to capture an image of the lighting equipment including the lighting equipment, and corresponding real-time brightness characteristics and brightness detection results and real-time chromaticity characteristics and chromaticity detection results are obtained through the lighting equipment image, and the fault state of the lighting equipment is determined according to the brightness detection results and chromaticity detection results. By using the image of the lighting equipment captured by the camera to automatically detect faults, the fault state of the lighting equipment is quickly detected, which reduces the time and labor intensity of manual inspection, thereby improving the maintenance efficiency of the lighting equipment.
[0126] In order to make the technical solution of the present application more clear, more embodiments are described below.
[0127] In one embodiment, in order to achieve more accurate detection of fault status of lighting equipment, when analyzing the lighting equipment image, the lighting equipment image can be divided into image blocks and then image block detection can be performed to locate the fault area, thereby accurately locating the fault.
[0128] Accordingly, the lighting equipment fault detection method of the present application may further include:
[0129] A1: Divide the lighting equipment image into multiple image blocks, and extract identification feature information of each image block; wherein the identification feature information may include the brightness mean and brightness standard deviation and the chromaticity mean and chromaticity standard deviation of the image block.
[0130] Specifically, the lighting equipment image can be divided into n×m small blocks, and the brightness L of each block can be calculated. ij and chromaticity C ij Detection; extract brightness and chromaticity features for each block, including statistical features such as mean and standard deviation, where the brightness mean of the i-th image block is u i , the brightness standard deviation is σ i ,Similarly, chromaticity features can also be extracted by this method.
[0131] A2: Identify the faulty image blocks from each image block according to the identification feature information.
[0132] In one embodiment, the method for identifying a faulty image block may include the following:
[0133] First, detect the brightness and chromaticity of each image block, and calculate the corresponding brightness error and chromaticity error. The formula is as follows:
[0134]
[0135] Among them, L ij Indicates brightness and C ij Indicates chromaticity; ΔL ij =|Lij -L 标准 | represents the brightness error of the image block, ΔC ij =|C ij -C 标准 | represents the chrominance error of the image block.
[0136] Then, the brightness error and the chromaticity error are judged by using a preset recognition threshold to determine the faulty image block; specifically, by setting an appropriate recognition threshold, it is possible to determine which image blocks have faults.
[0137] In another embodiment, the above-mentioned fault image block can also be identified by pre-training a lighting fault detection model, which specifically includes the following:
[0138] A matching lighting fault detection model is selected according to the recognition accuracy, and the recognition feature information of the image block is input into the lighting fault detection model for fault classification and recognition to obtain a faulty image block; wherein the lighting fault detection model is a pre-trained classification model for identifying image brightness and chromaticity faults.
[0139] A3: Locate the faulty area of the lighting device according to the faulty image block.
[0140] Specifically, after positioning and calculating all fault image blocks, the fault areas of all lighting equipment in the audio-visual room can be obtained.
[0141] As in the technical solution of the above-mentioned embodiment, by dividing the lighting equipment image into multiple image blocks and extracting their brightness and chromaticity feature information respectively to detect the fault status, the fault area of the lighting equipment can be automatically detected quickly and accurately, thereby providing important technical support for quickly determining the fault location and troubleshooting, and further improving the maintenance efficiency of the lighting equipment.
[0142] In one embodiment, in order to more accurately find out the fault state of the lighting device, a fault detection may be performed on the lamp tubes in the lighting device, thereby accurately finding the lamp tubes in the fault state.
[0143] Accordingly, the technical solution of the present application may also include the following:
[0144] B1: Obtain standard state parameters of a lamp in a lighting device, wherein the standard state parameters represent the brightness and chromaticity corresponding to the lamp when it works normally.
[0145] For example, a standard state S can be set, where S represents the brightness and chromaticity of a normally working lamp. The same method as the image block is used to calculate the standard state S using the formula ΔL ij =|L ij -L 标准| represents the brightness error of the lamp, the formula is ΔC ij =|C ij -C 标准 |Indicates the chromaticity error of the lamp.
[0146] B2: Acquire actual state parameters corresponding to the lamp tube in the lighting equipment; wherein the actual state parameters are the brightness and chromaticity corresponding to the lamp tube in the lighting equipment image.
[0147] Exemplarily, the lighting equipment image acquired by the camera can generate an actual state parameter R, which represents the brightness and chromaticity corresponding to the light tube in the lighting equipment image.
[0148] B3: Using the preset state parameter threshold to judge the state parameter deviation value of the lamp tube to determine the faulty lamp tube in the fault state.
[0149] As an embodiment, the method for determining a faulty lamp in a faulty state may include the following steps:
[0150] ΔC=|C R -C S
[0151] ΔL=|L R -L S
[0152] Among them, ΔL represents the brightness deviation value, ΔC represents the chromaticity deviation value, L R Indicates the brightness of the standard state parameter, L S Indicates the brightness of the actual state parameter; C R Indicates the chromaticity of the standard state parameter, C S Indicates the chromaticity of the actual status parameter. If ΔL or ΔC exceeds the preset status parameter threshold, it is determined that the lamp status is inconsistent with the setting and the lamp is determined to be in a fault state.
[0153] As in the technical solution of the above-mentioned embodiment, by extracting the brightness and chromaticity characteristic information of the lamp from the image of the lighting equipment, the brightness and chromaticity of the lamp are detected for the fault status, so that the fault status of the lamp of the lighting equipment can be accurately detected, so that the faulty lamp in the audio-visual room can be accurately and automatically identified, thereby reducing the frequent manual maintenance needs and reducing the labor cost of maintenance.
[0154] In one embodiment, in order to perform fault detection on lighting equipment images, when using a pre-trained recognition model to perform fault detection, this embodiment also provides multiple model training methods, which can be used in the fault status detection of the above-mentioned image blocks and lamps.
[0155] (1) The lighting fault detection model includes a binary classification SVM model of a support vector machine. The SVM model training process includes:
[0156]
[0157] subject.to.y i (w·x i +b)≥1-ζ i ,ζ i ≥0
[0158] Among them, w and b are the parameters of the SVM model, C is the penalty term coefficient, and f(·) represents the SVM classification function.
[0159] (2) The lighting fault detection model includes a CNN classifier using a convolutional neural network, and the training process of the CNN classifier includes:
[0160]
[0161] in, Represents the predicted category probability, N represents the number of training samples, and y represents the output category probability; CNN can be used to learn the feature representation of images and perform accurate classification at the image level or image block level, which has higher applicability.
[0162] (3) The lighting fault detection model includes a semantic segmentation network based on deep learning, and the melody process of the semantic segmentation network includes:
[0163]
[0164] Among them, L(·) represents the loss function, Represents the category of each pixel predicted by the semantic segmentation network, y i Represents the true label, and N represents the number of training samples. Through the semantic segmentation network, each pixel in the image can be classified as faulty or normal, thereby achieving pixel-level accuracy detection.
[0165] As in the technical solution of the above embodiment, by using the pre-trained recognition model, when performing fault detection of image blocks or lamp tubes in the lighting equipment image, a suitable algorithm and recognition model can be selected according to the specific situation. Fault detection with different accuracy and efficiency can be achieved, and can be selected and adjusted according to needs, which has higher applicability in actual use.
[0166] In one embodiment, when a lamp failure is detected in a lighting device, the lamp needs to be located in order to troubleshoot the failure. For this purpose, the present application also provides an embodiment for locating the faulty lamp tube, in which the control system recognizes and analyzes the lighting device image and its lighting parameters to automatically detect the faulty lamp tube, and uses the solved camera parameters to perform triangulation to obtain the actual position of the faulty lamp tube, thereby providing important technical support for troubleshooting, reducing the time and labor intensity of manual inspection, facilitating rapid troubleshooting, and thus improving the maintenance efficiency of the lighting equipment.
[0167] Specifically, the lighting equipment fault detection method of the present application may also include the following:
[0168] C1: Using at least two cameras to respectively capture a lighting equipment image including the lighting equipment to identify the faulty lamp tube and its position information in the lighting equipment image.
[0169] like Figure 1 In the present invention, the technical solutions of camera X1 and camera X2 are used. The corresponding location information can be identified respectively through the images of lighting equipment taken by the two cameras.
[0170] C2: solving the camera parameters of the camera according to the coordinate data of the preset calibration point image in the lighting equipment image.
[0171] For example, the position of a faulty lamp captured by two cameras X1 and X2 in the lighting equipment images of cameras X1 and X2 is (x1, y1), and the position on the image of camera X2 is (x2, y2); the positions and angles of cameras X1 and X2 are known, and the trigonometric law can be used to determine the actual position of the faulty lamp.
[0172] In one embodiment, the method for solving the camera parameters of the camera includes the following steps:
[0173] a. Select several position points at the lighting equipment site as calibration points, and determine the coordinate parameters of the calibration points in the lighting equipment site coordinate system.
[0174] For example, several vertices inside the audio-visual room can be selected as calibration points. The three-dimensional coordinates of these calibration points are known in the audio-visual room coordinate system and are represented by P i =(X i ,Y i ,Z i ,), where i represents the number of the calibration point.
[0175] b. Determine the two-dimensional coordinates of each calibration point in the lighting equipment image; specifically, obtain the calibration point image inside the audio-visual room captured by the camera through the lighting equipment image, and determine the two-dimensional coordinates of the calibration point on the calibration point image, expressed as pi =(x i ,y i ).
[0176] c. For each calibration point, establish a set of equations for camera parameters according to its two-dimensional coordinates;
[0177] Exemplarily, the equation group of the camera parameters corresponding to each calibration point is:
[0178]
[0179] Where f is the focal length of the camera and (u,v) is the principal point of the image.
[0180] d. Use the equations established by all the calibrated points and solve them by the least squares method to obtain the camera parameters.
[0181] Exemplarily, the camera parameters to be solved are:
[0182]
[0183] Among them, p i =(x i ,y i ) are the two-dimensional coordinates of the calibration points, and M represents the matrix.
[0184] As in the technical solution of the above-mentioned embodiment, the coordinate data of the preset calibration points are used in combination with the lighting equipment image to finally solve the camera parameters of the camera, providing accurate position point data for triangulation positioning, thereby accurately locating the position of the faulty lamp tube.
[0185] Furthermore, after solving the camera parameters, in order to ensure the accuracy of the camera parameters, the present application also provides an error evaluation technical solution, which may specifically include the following.
[0186] Use the solved camera parameters to calculate the coordinate error between the predicted 2D coordinates and the actual 2D coordinates of each calibration point; if the coordinate error is within the available range, the calibration is successful; otherwise, reselect the calibration point to solve the camera parameters or adjust the camera's position and posture.
[0187] For example, the coordinate error can be expressed as:
[0188]
[0189] in, To predict the two-dimensional coordinates, p i is the actual two-dimensional coordinate.
[0190] As in the technical solution of the above-mentioned embodiment, the accuracy of the camera parameters can be ensured through error evaluation processing, thereby ensuring the accuracy of the triangulation position point data.
[0191] In one embodiment, in order to improve the accuracy of calculating camera parameters, other influencing factors need to be considered, such as the distortion of the image taken by the camera, the shape and size of the faulty lamp, etc. Accordingly, the distortion of the image taken by the camera can be corrected before calculation, and then the calculation can be performed in combination with the shape and size of the faulty lamp.
[0192] For example, in the calibration of triangulation positioning, the size of the audio-visual room and the position of the camera are fixed, and there may be a certain error in the observation angle. The observation angle of the camera can be calibrated according to the vertices of the audio-visual room. The camera can be calibrated using the geometric characteristics of the audio-visual room to determine the camera's internal parameters (such as focal length, principal point, etc.) and external parameters (such as the camera's position and posture).
[0193] As in the technical solution of the above-mentioned embodiment, through the calibration scheme of the observation angles in cameras X1 and X2, the intrinsic and extrinsic parameters of the cameras can be accurately determined, providing a more accurate data basis for triangulation positioning, thereby improving the accuracy of detection and positioning of faulty lamps.
[0194] C3: Perform triangulation based on the position information of the faulty lamp tube identified by the two lighting equipment images and the corresponding camera parameters to obtain the actual position of the faulty lamp tube.
[0195] Specifically, triangulation positioning is performed using the solved camera parameters. The position of the damaged lamp tube in the image taken by camera X1 is (x1, y1), and the position in the image taken by camera X2 is (x2, y2). The actual position (X, Y, Z) of the lamp tube is calculated using the triangulation rule based on the position and angle of the camera.
[0196] In one embodiment, a method for obtaining the actual position of a faulty lamp may include the following steps:
[0197] e. Acquire a first position of the faulty lamp identified by the lighting equipment image of the first camera, and acquire a second position of the faulty lamp identified by the lighting equipment image of the second camera;
[0198] f. respectively obtaining the coordinate positions and observation angles of the first camera and the second camera in the on-site coordinate system of the lighting equipment;
[0199] g. Calculate the three-dimensional coordinates of the faulty lamp in the lighting equipment on-site coordinate system based on the first position, the coordinate position and the observation angle of the first camera, the second position, the coordinate position and the observation angle of the second camera and the trigonometric rule.
[0200] Specifically, using trigonometric rules to calculate the three-dimensional coordinates of the faulty lamp in the lighting equipment on-site coordinate system may include:
[0201]
[0202] Among them, (x A ,y A ) is the first position, (x B ,y B ) is the second position, (X, Y, Z) is the three-dimensional coordinate of the faulty lamp in the lighting equipment on-site coordinate system;
[0203] d A =f×tan(θ A )
[0204] d B =f×tan(θ B )
[0205]
[0206] Among them, d A is the distance between the first camera and the faulty lamp, d B is the distance between the second camera and the faulty lamp, θ A and θ B These are the observation angles of the faulty lamp in the first camera and the second camera respectively.
[0207] The technical solution of the above embodiment identifies the faulty lamp tube and its position information in the lighting equipment image by performing fault analysis on at least two of the lighting equipment images, and determines the actual position of the faulty lamp tube by using the triangulation positioning method in combination with the solved camera parameters; the actual position of the lamp tube can be accurately found, providing important technical support for troubleshooting and facilitating rapid troubleshooting.
[0208] In order to improve the maintenance efficiency of lighting equipment and ensure the stability and consistency of lighting effects when a lamp fails, the user can get the best user experience; to this end, after positioning the lamp, the present application also provides a technical solution for adjusting the lamp.
[0209] In one embodiment, the fault detection method of the lighting device of the present application may further include:
[0210] D1: obtaining the lighting effect of the scene within a certain range of the faulty lamp in real time according to the lighting equipment image taken by the camera;
[0211] D2: According to the actual position of the faulty lamp, several lamps around it are selected as target lamps; the brightness and chromaticity of the target lamps are adjusted, and the adjusted lighting effect is observed in real time according to the lighting equipment image taken by the camera until the adjusted lighting effect reaches a preset consistency range.
[0212] As in the technical solution of the above-mentioned embodiment, the consistency of lighting effects is achieved by adjusting the lamps around the faulty lamp, ensuring the stability and consistency of the lighting effects of the lighting equipment, enhancing the user experience, and through automatic detection and adjustment, it can ensure that the lighting effects on site are always in the best state, providing users with a continuous, stable, high-quality entertainment experience.
[0213] An embodiment of the lighting control system is described below.
[0214] refer to Figure 3 As shown, Figure 3 It is an electrical structure diagram of a lighting control system of an embodiment, which includes: a camera and a control device; wherein the control device is connected to the lighting device; the camera is used to capture the image of the lighting device; the control device is used to receive the image of the lighting device, and is configured to execute the lighting device fault detection method of any of the aforementioned embodiments to perform fault detection on the lighting device.
[0215] Exemplarily, a corresponding algorithm module may be configured in the control device, which executes the relevant functions provided in the embodiments of the present application, thereby providing a more convenient use for the user; Figure 4 As shown, Figure 4 This is an example of an algorithm module diagram for controlling the operation of a device, which mainly includes:
[0216] Camera module: used to control the deployed cameras to capture images of lighting equipment and capture details of lighting effects.
[0217] Image processing and analysis module: used to analyze and process the lighting equipment images taken by the camera, analyze the lighting effects of the lighting equipment images, etc.
[0218] Algorithm execution module: used to execute the algorithm for fault detection of faulty areas and faulty lamps and the adjustment of the lighting effects of the lamps.
[0219] Lighting control module: used to control lighting effect parameters such as brightness and chromaticity of each lamp.
[0220] User interface module: provides a friendly operation interface, displays the results of various fault detections and related information for adjusting the lighting effects of the lamps, etc.
[0221] The technical solution of the above-mentioned embodiment automatically detects faults by utilizing the image of the lighting equipment captured by the camera, quickly detects the fault status of the lighting equipment, reduces the time and labor intensity of manual inspection, and thus improves the maintenance efficiency of the lighting equipment; by dividing the image of the lighting equipment into multiple image blocks for detection, the fault area can be quickly and accurately located, thereby providing important technical support for quickly determining the fault location and troubleshooting, and further improving the maintenance efficiency of the lighting equipment; by detecting the status of the lamp tube, the faulty lamp tube can be quickly found, and the faulty lamp tube is automatically located, thereby reducing the frequent manual maintenance needs and reducing the labor cost of maintenance; by adjusting the lamp tubes around the faulty lamp tube to achieve consistency of lighting effects, ensure the stability and consistency of the lighting effects of the lighting equipment, enhance the user experience, and through automatic detection and adjustment, it can ensure that the lighting effects on site are always in the best state, providing users with a continuous, stable and high-quality entertainment experience.
[0222] The technical solution provided by the embodiments of the present application has the following technical effects:
[0223] Improve maintenance efficiency: It can quickly and accurately detect the fault status of lighting equipment with different brightness and chromaticity, locate the faulty area or faulty lamp, greatly reducing the time and labor intensity of manual inspection, thereby improving the overall maintenance efficiency.
[0224] Enhanced user experience: Automatically detect and adjust solutions to ensure that lighting effects are always optimal, providing users with a continuous, stable, high-quality entertainment experience.
[0225] Extend the service life of equipment: Timely fault detection and maintenance of lighting equipment can effectively avoid equipment damage caused by long-term fault operation, thereby extending the service life of the equipment.
[0226] Save maintenance costs: Automatic detection reduces the need for frequent manual maintenance and reduces maintenance labor costs; at the same time, through timely fault detection and processing, it can also reduce the replacement or repair costs caused by equipment failure.
[0227] Improve the stability of lighting effects: Through the function of automatically adjusting lighting effects, when a local faulty lamp is detected, the brightness and chromaticity of the surrounding lamps can be automatically adjusted to ensure the stability and consistency of the overall lighting effect.
[0228] Environmental protection and energy saving: Accurate fault detection and automatic adjustment functions can avoid unnecessary energy waste, bring higher energy efficiency to entertainment venues, and also help reduce carbon footprint and achieve more environmentally friendly operations.
[0229] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A method for detecting a fault of a lighting device, characterized in that: include: When the lighting device is in a lighting state, using a camera to capture a lighting device image including the lighting device; Analyze the lighting equipment image and calculate the corresponding real-time brightness feature, and determine the real-time brightness feature according to a preset brightness threshold to obtain a brightness detection result; Using a color histogram comparison method to detect chromaticity features of the lighting equipment image to obtain a chromaticity detection result; The fault state of the lighting equipment is determined according to the brightness detection result and the chromaticity detection result.
2. The fault detection method for lighting equipment according to claim 1, characterized in that: Analyzing the lighting equipment image and calculating the corresponding real-time brightness feature, and determining the real-time brightness feature according to a preset brightness threshold to obtain a brightness detection result, including: Performing color space conversion on the lighting equipment image; Performing local contrast enhancement processing on the converted lighting equipment image; Extract brightness features of lighting equipment images; The extracted brightness features are equalized, and pixels whose brightness exceeds a preset brightness threshold are marked as brightness abnormal points.
3. The fault detection method for lighting equipment according to claim 2, characterized in that: Extract the brightness features of the lighting equipment image, including: Where L represents the average brightness of the lighting device image, I ij is the pixel value of the lighting device image at position (x, j), W and H are the width and height of the lighting device image respectively, w ij is the weight at position (x, j); The equalization processing of the extracted brightness features includes: I enhanced =CLAHE(Play) Wherein, Igray(x,y) is the grayscale image of the lighting equipment image, CLAHE is the application of contrast-limited adaptive histogram equalization algorithm to enhance image contrast, and x and y represent the horizontal and vertical coordinates of the image, respectively.
4. The lighting equipment fault detection method according to claim 1, characterized in that: The color histogram comparison method is used to detect the chromaticity feature of the lighting equipment image to obtain the chromaticity detection result, including: Obtaining a normalized histogram of the lighting device image; Calculating the cross entropy of the normalized histogram of the lighting device image according to a preset standard color histogram; The chromaticity detection result is obtained by comparing the cross entropy with the preset threshold.
5. The lighting equipment fault detection method according to claim 4, characterized in that: Calculating the cross entropy of the normalized histogram of the lighting device image according to a preset standard color histogram includes: P={p1,p2,...,p3} Q={q1,q2,...,q3} Among them, H(P,Q) is the cross entropy, P is the preset standard color histogram, and Q is the normalized histogram of the lighting equipment image.
6. The lighting equipment fault detection method according to any one of claims 1 to 5, characterized in that: Also includes: Divide the lighting equipment image into a plurality of image blocks, and extract identification feature information of each image block; wherein the identification feature information includes the brightness mean and brightness standard deviation and the chromaticity mean and chromaticity standard deviation of the image block; Identifying a faulty image block from each image block according to the identification feature information; The fault area of the lighting device is located according to the fault image block.
7. The lighting equipment fault detection method according to claim 6, characterized in that: Identifying a faulty image block from each image block according to the identification feature information includes: Selecting a matching lighting fault detection model according to the recognition accuracy, inputting the recognition feature information of the image block into the lighting fault detection model for fault classification and recognition to obtain a fault image block; The lighting fault detection model is a pre-trained classification model for identifying image brightness and chromaticity faults.
8. The lighting equipment fault detection method according to claim 7, characterized in that: The lighting fault detection model includes a binary classification SVM model of a support vector machine, and the SVM model training process includes: subject.to.y i (w·x i +b)≥1-ζ i ,ζ i ≥0 Among them, w and b are the parameters of the SVM model, C is the penalty term coefficient, and f(·) represents the SVM classification function; The lighting fault detection model includes a CNN classifier using a convolutional neural network, and the training process of the CNN classifier includes: in, Represents the predicted category probability, N represents the number of training samples, and y represents the output category probability; or The lighting fault detection model includes a semantic segmentation network based on deep learning, and the melody process of the semantic segmentation network includes: Among them, L(·) represents the loss function, Represents the category of each pixel predicted by the semantic segmentation network, y i represents the true label, and N represents the number of training samples.
9. The lighting equipment fault detection method according to claim 6, characterized in that: Identifying a faulty image block from each image block according to the identification feature information includes: Detect the brightness and chromaticity of each image block and calculate the corresponding brightness error and chromaticity error; Among them, L ij Indicates brightness and C ij Indicates chromaticity; ΔL ij =|L ij -L 标准 | represents the brightness error of the image block, ΔC ij =|C ij -C 标准 | represents the chrominance error of the image block; The brightness error and the chromaticity error are judged by using a preset recognition threshold to determine the faulty image block.
10. The lighting equipment fault detection method according to claim 6, characterized in that: Also includes: Obtaining standard state parameters of a lamp in the lighting device, wherein the standard state parameters represent the brightness and chromaticity corresponding to the normal operation of the lamp; Acquire actual state parameters corresponding to the lamp tube in the lighting device; wherein the actual state parameters are the brightness and chromaticity corresponding to the lamp tube in the lighting device image; The faulty lamp tube in the fault state is determined by using the preset state parameter threshold to judge the state parameter deviation value of the lamp tube.
11. The lighting equipment fault detection method according to claim 10, characterized in that: Determining the faulty lamp tube in a faulty state by using a preset state parameter threshold to determine the state parameter deviation value of the lamp tube includes: ΔC=|C R -C S | ΔL=|L R -L S | Among them, ΔL represents the brightness deviation value, ΔC represents the chromaticity deviation value, L R Indicates the brightness of the standard state parameter, L S Indicates the brightness of the actual state parameter; C R Indicates the chromaticity of the standard state parameter, C S Indicates the chromaticity of the actual state parameter.
12. The lighting equipment fault detection method according to claim 10, characterized in that: Also includes: Using at least two cameras to respectively capture a lighting equipment image including the lighting equipment to identify the faulty lamp tube and its position information in the lighting equipment image; Determining the camera parameters of the camera according to the coordinate data of the preset calibration point image in the lighting equipment image; The actual position of the faulty lamp is obtained by performing triangulation based on the position information of the faulty lamp identified by the two lighting equipment images and the corresponding camera parameters.
13. The lighting equipment fault detection method according to claim 12, characterized in that: Also includes: According to the lighting equipment image captured by the camera, the lighting effect of the scene within a certain range of the faulty lamp is obtained in real time; According to the actual position of the faulty lamp tube, several lamp tubes around it are selected as target lamp tubes; the brightness and chromaticity of the target lamp tubes are adjusted, and the adjusted lighting effect is observed in real time according to the lighting equipment image taken by the camera until the adjusted lighting effect reaches a preset consistency range.
14. A lighting control system, characterized in that: include: A camera and a control device; wherein the control device is connected to a lighting device; The camera is used to capture images of lighting equipment; The control device is used to receive a lighting device image and execute the lighting device fault detection method according to claims 1 to 13 to perform fault detection on the lighting device.
Citation Information
Cited By
Urban lighting fault positioning method and device
CN120352111A
