A high-power LED lamp bead group control system based on wireless communication
Through a high-power LED light bead group control system based on wireless communication, using camera cameras and HSV image processing technology, visual monitoring of road traffic lights is realized, solving the problem of difficulty in timely detection of light beads in the existing technology, and improving the working efficiency of traffic lights.
Patent Information
- Application Number
- CN202510140927.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-08
- Publication Date
- 2025-06-03
- Estimated Expiration
- 2045-02-08
AI Technical Summary
It is difficult for the existing technology to visually monitor the fault of road traffic lights, resulting in damage to the lamp beads not being discovered in time, affecting the working effect of traffic lights.
A high-power LED light bead group control system based on wireless communication is adopted. By capturing the camera, acquiring pictures, performing HSV image processing, and selecting three target pixels, red, yellow and green. Through the overlap analysis of multiple pictures, an RYG distribution map is constructed, the signal light area is identified, and fault judgment is made.
It realizes rapid fault diagnosis of traffic lights, can detect damage to light beads in time, improves the working efficiency of traffic lights, and is suitable for situations where the number of damage to light beads does not reach the fault error limit but has a significant impact.
Smart Images

Figure CN119580521B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of light source detection and control, and particularly relates to a high-power LED lamp bead group control system based on wireless communication. Background Art
[0002] Road traffic signal lights refer to devices that, at road intersections or other traffic control areas, control the traffic flow on the road by changing the lights of three different colors, red, yellow, and green, to ensure traffic safety and order. Road traffic signal lights are usually composed of a specific combination of monochromatic LED lamp bead groups. When the lamp bead group lights up, it displays a pattern of a specified color to assist in directing, guiding, and controlling traffic.
[0003] Since road traffic signal lights need to be used for a long time and are used in outdoor terrace locations, the LED lamp bead groups in them are prone to lamp bead damage. In the prior art, when the number of damaged lamp beads reaches a certain amount, a fault error signal will be automatically triggered for fault handling. However, when the number of damaged lamp beads does not reach the preset quantity threshold, visual inspection is required to detect the fault, resulting in a failure to respond in a timely manner and poor working performance of the traffic signal lights. Therefore, there is an urgent need for a lamp bead group control system that can replace manual labor and visually monitor the faults of road traffic signal lights. Summary of the Invention
[0004] In view of the above-mentioned drawbacks of the prior art, the present invention provides a high-power LED lamp bead group control system based on wireless communication, which can effectively solve the problem of difficult visual monitoring of faults in road traffic signal lights in the prior art.
[0005] To achieve the above object, the present invention is realized through the following technical solutions:
[0006] The present invention provides a high-power LED lamp bead group control system based on wireless communication, including a control center, and further including:
[0007] A wireless communication unit: establishing a wireless communication connection between the control center, the traffic signal light group, and the capture camera. When the traffic signal light group generates a traffic signal change, a light shear signal is generated and sent to the control center;
[0008] A signal light recognition unit, which captures a photo when the control center obtains the light shear signal, screens out the red pixel points, yellow pixel points, and green pixel points in the captured photo, collectively referred to as target pixel points, obtains an RYG distribution map based on the distribution of the target pixel points, and assigns a value to each individual pixel point, where:
[0009] Non-target pixel points are assigned a value of 0, and red pixel points, yellow pixel points, and green pixel points are assigned non-zero integers;
[0010] Based on multiple consecutive RYG distribution maps, an overlapping analysis is performed to construct an assignment sequence. Based on the number q of distinct non-zero elements in the assignment sequence, the corresponding pixel points are marked as pixel points with q color changes. The pixel points with 1 color change are screened out, and multiple region circles are constructed. Based on the distribution of the pixel points with 1 color change in the region circles, the pending region circles are screened out. Based on the spatial distribution of multiple pending region circles, the signal light region is screened out, and the signal light region corresponds to a traffic signal light;
[0011] The lamp type discrimination unit divides the colors of the traffic signal lights based on the light source hues corresponding to the pixel points within the signal light region, calculates the discrimination threshold based on the pattern areas of different pattern signal lights, and discriminates the types of traffic signal lights according to the discrimination threshold;
[0012] The fault judgment unit performs fault analysis on the traffic signal lights corresponding to each signal light region.
[0013] Furthermore, the process of screening target pixel points is as follows:
[0014] Obtain the currently captured picture by the capture camera and record it as the captured picture. Perform HSV color space conversion processing on the captured picture to obtain the HSV captured picture. Any pixel point in the HSV captured picture corresponds to a hue, saturation, and lightness;
[0015] Preset a red hue value range, a yellow hue value range, and a green hue value range. Each hue value range corresponds to the hue value range of a color light in the HSV captured picture. Extract the pixel points in the HSV captured picture whose hue values are within the red hue value range, the yellow hue value range, and the green hue value range, and record them as red pixel points, yellow pixel points, and green pixel points respectively. The three types of pixel points are collectively referred to as target pixel points. Eliminate the non-target pixel points in the HSV captured picture to obtain the RYG distribution map, and each RYG distribution map corresponds to a shooting moment.
[0016] Furthermore, the process of constructing the assignment sequence is as follows:
[0017] Construct a rectangular coordinate system with one corner of the captured picture as the coordinate origin and record it as the overlapping coordinate system. The unit length of the overlapping coordinate system is one pixel point. Obtain the coordinate positions of all pixel points in the captured picture in the overlapping coordinate system and record them as pixel coordinates;
[0018] Assign values to each pixel coordinate based on the RYG distribution map. When the pixel point corresponding to the pixel coordinate is a non-target pixel point, assign 0. When the pixel point corresponding to the pixel coordinate is a red pixel point, assign 1. When the pixel point corresponding to the pixel coordinate is a yellow pixel point, assign 2. When the pixel point corresponding to the pixel coordinate is a green pixel point, assign 3;
[0019] Based on multiple consecutive RYG distribution maps, the assignment sequence corresponding to each pixel coordinate is denoted as , and the arrangement order of the elements in the assignment sequence is consistent with the sorting of the RYG distribution maps. k is the shooting serial number of the RYG distribution maps.
[0020] Furthermore, the process of obtaining the distinct quantity is as follows:
[0021] Extract the non-zero elements in the assignment sequence in turn and perform subtraction operations with all the previous non-zero elements. When the operation results are all not equal to 0, record the non-zero element as a distinct non-zero element, and count the number of distinct non-zero elements in the assignment sequence as the distinct quantity.
[0022] Furthermore, the process of constructing the regional circle is as follows:
[0023] Screen out the single-color-changing pixel points and remove the discrete pixel points among them to obtain the single-color-changing distribution map, and record the remaining pixel points as the first target points;
[0024] Obtain the horizontal distance between the traffic signal light group and the capture camera and denote it as , obtain the vertical height difference between the traffic signal light group and the capture camera and denote it as , obtain the mask diameter of the traffic signal light and denote it as , obtain the interval between traffic signal lights and denote it as , obtain the angle between the main optical axis of the capture camera lens and the vertical plane and denote it as ;
[0025] Substitute into the formula for calculation to obtain the regional width threshold R, where are all preset weight coefficients;
[0026] Construct multiple regional circles with a diameter equal to the regional width threshold in the single-color-changing distribution map.
[0027] Furthermore, the process of discrete judgment of pixel points is as follows:
[0028] Calculate the distance between any one of the first target points and each of the other first target points and take the minimum value as the discrete judgment value. There is a preset discrete judgment threshold. When the discrete judgment value of any one of the first target points is greater than or equal to the discrete judgment threshold, mark the first target point as a discrete point.
[0029] Furthermore, the process of traffic signal light area screening is as follows:
[0030] Calculate the interval threshold , and the calculation formula of the interval threshold is as follows:
[0031] , where is a preset weight coefficient;
[0032] When the area ratio of the first target point within the regional circle is greater than or equal to a preset ratio threshold, and the distance between the nearest first target point outside the regional circle and the regional circle is less than the interval threshold, the regional circle is denoted as a to-be-determined regional circle;
[0033] Obtain the centers of multiple to-be-determined regional circles, calculate the distance between the centers of adjacent to-be-determined regional circles and compare it with the center distance threshold. The center distance threshold is equal to the sum of the regional width threshold and the interval threshold. When the distance between the centers of adjacent to-be-determined regional circles is less than or equal to the center distance threshold, they are denoted as a group of adjacent regional circles;
[0034] When there are two groups of adjacent regional circles among any three to-be-determined regional circles, the three to-be-determined regional circles are denoted as a group of regional circle combinations. Obtain the maximum angle of the triangle formed by the centers of the three to-be-determined regional circles in the regional circle combination as the judgment angle. When the judgment angle is greater than or equal to a preset angle threshold, the three regions corresponding to the three to-be-determined regional circles are denoted as signal light regions.
[0035] Furthermore, the process of distinguishing the types of traffic signal lights is as follows:
[0036] There are a preset arrow-shaped standard pattern and a circular standard pattern. Calculate the pattern areas of the arrow-shaped standard pattern and the circular standard pattern respectively and denote them as , substitute into the formula for calculation to obtain the discrimination threshold , where DF is the preset ratio threshold;
[0037] Calculate the area ratio of the first target point in the signal light region as the ratio judgment value. When the ratio judgment value in the signal light region is less than the discrimination threshold, the signal light region is marked as an arrow signal light region, corresponding to an arrow indicator light;
[0038] When the ratio judgment value in the signal light region is greater than or equal to the discrimination threshold, the signal light region is marked as a circular signal light region, corresponding to a circular indicator light.
[0039] Furthermore, the fault analysis process of the fault judgment unit is as follows:
[0040] Denote the monitored signal light region as the target region. Obtain the color of the traffic signal light corresponding to the target region as the target color. Obtain the most recent captured picture and count the number of target color pixel points in the target region as the quantity parameter value. When the quantity parameter value is greater than 0 and less than or equal to the quantity lower limit value, a damage signal is generated;
[0041] When the quantity parameter value is greater than the quantity lower limit value and less than the quantity detection threshold, pattern abnormality detection is performed, where:
[0042] quantity lower limit value The calculation formula is , where is a preset weight coefficient, S is the pattern area of the standard pattern of the pattern signal light corresponding to the target area, and P is a preset proportional threshold;
[0043] Quantity detection threshold The calculation formula is , where is a preset detection trigger ratio value.
[0044] Furthermore, the pattern abnormality detection process is as follows:
[0045] When the type of the traffic signal light corresponding to the target area is a circular indicator light, a maintenance signal is generated;
[0046] When the type of the traffic signal light corresponding to the target area is an arrow indicator light, a reference arrow contour is constructed, a plurality of uniformly distributed reference points are drawn in the reference arrow contour, all target color pixel points within the current target area are marked in the reference arrow contour, the discrete pixel points are removed, and the distance between each reference point and the nearest target color pixel point is calculated and recorded as the distribution distance;
[0047] When the distribution distance is greater than or equal to a preset distribution threshold, the reference point is recorded as a discrete reference point, the proportion of the discrete reference points in all reference points is calculated to obtain the coverage ratio, and when the coverage ratio is less than or equal to the coverage threshold, a pattern missing signal is generated.
[0048] The technical solution provided by the present invention has the following beneficial effects compared with the known prior art:
[0049] 1. The present invention performs HSV image processing on the captured pictures of the capture camera, screens out the target pixel points of red, yellow, and green, and through the overlapping analysis of the target pixel points in multiple captured pictures, can further screen out the pixel points with single-color changes. Based on the distribution characteristics of the single-color changed pixel points and the arrangement characteristics when the traffic signal lights are installed, the position area of the traffic signal lights in the picture can be quickly locked, so that the corresponding fault diagnosis of the traffic signal lights can be carried out based on the image changes within this area. It can not only timely detect the fault manifestations of the traffic signal lights, but also make full use of the image data information obtained by the capture camera, enrich the uses of the capture camera, and is faster and more convenient than the fault judgment methods in the prior art.
[0050] 2. The present invention can further analyze the pattern of the traffic signal light based on the pixel distribution characteristics in the target area of the recently captured picture, so as to timely detect the incomplete indication pattern caused by the damage of the lamp beads in the traffic signal light. The present invention makes a fault judgment from the perspective of the actual working use of the traffic signal light, which is more advanced and intelligent compared with the existing traffic signal light lamp bead fault detection means, and is especially applicable to the situation where the number of damaged lamp beads does not reach the fault reporting limit value, but the actual working effect is affected. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.
[0052] Figure 1 It is the overall module block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0054] The following further describes the present invention with reference to the embodiments.
[0055] Refer to Figure 1 , a high-power LED lamp bead group control system based on wireless communication, which acts on the traffic signal light group and the capture camera at the traffic road intersection. The traffic signal light group includes multiple traffic signal lights indicating different traffic signals (such as red light, green light, yellow light, straight, left turn, and right turn, etc.). Each traffic signal light includes multiple high-power LED lamp beads. The capture camera is arranged opposite the traffic signal light group for the traffic signal light group. In a specific embodiment, the capture camera directly uses the existing traffic violation capture cameras in the prior art. These cameras can usually capture pictures including the traffic signal light group, so as to control the traffic signal light group based on the picture.
[0056] It should be noted that the relatively arranged traffic light group and snapshot camera are a group of control objects. The present invention only analyzes and controls the same group of control objects, that is, any traffic light group and snapshot camera mentioned below belong to the same group of control objects.
[0057] At least the control center, and also:
[0058] Wireless communication unit: establishes wireless communication connection between the control center and the traffic light group and snapshot camera. The control center acts as a processor to receive, process and forward data from the traffic lights and snapshot cameras. The wireless communication connection includes but is not limited to BLE Mesh, Bluetooth, Zigbee and other communication methods. When the traffic light group generates a traffic signal change (for example, a red light changes to a green light, a green light changes to a yellow light, or a yellow light changes to a green light), a light shear signal is generated and sent to the control center;
[0059] The traffic light recognition unit generates a snapshot command when the control center obtains the light cutting edge signal, sends it to the snapshot camera, controls the snapshot camera to take photos, and automatically recognizes the traffic lights in the picture based on the photos taken by the snapshot camera. The recognition steps are as follows:
[0060] Step 1: Get the picture currently captured by the snapshot camera and record it as the snapshot picture. Perform HSV color space conversion on the snapshot picture to obtain an HSV snapshot picture. Any pixel in the HSV snapshot picture corresponds to a hue, saturation, and brightness. There are preset red hue value intervals, yellow hue value intervals, and green hue value intervals. Each hue value interval corresponds to a hue value range of a color light in the HSV snapshot picture (for example, the hue value range of the green light in the picture after HSV color space conversion is between 45° and 75°). Extract the pixels in the HSV snapshot picture whose hue values are in the red hue value interval, the yellow hue value interval, and the green hue value interval, and record them as red pixels, yellow pixels, and green pixels, respectively. The three kinds of pixels are collectively referred to as target pixels. Eliminate non-target pixels in the HSV snapshot picture to obtain a RYG (red, yellow, and green) distribution map. Each RYG (red, yellow, and green) distribution map corresponds to a shooting moment (i.e., the time of the capture).
[0061] It should be noted that no matter when the snapshot is taken, at least one traffic light will be lit in the captured picture, that is, when the traffic light group is operating normally, at least one of the red, yellow and green lights will be lit at any time. Therefore, there will be at least one lit traffic light in the picture taken at any time.
[0062] Step 2: Obtain multiple continuous RYG distribution maps and record them as And perform overlapping analysis. k is the shooting serial number of the RYG distribution map. Assign values to each pixel point based on the hues of the pixel points at the same position in different RYG distribution maps (the RYG distribution map contains four hues, namely red R, yellow Y, green G, and blank NULL, and each hue corresponds to an assignment), and construct an assignment sequence, where:
[0063] Take one corner of the captured image as the origin of the coordinate system to construct a rectangular coordinate system and denote it as the overlapping coordinate system. The unit length of the overlapping coordinate system is one pixel point. Obtain the coordinate positions of all pixel points in the captured image in the overlapping coordinate system and denote them as pixel coordinates. Assign values to each pixel coordinate based on the RYG distribution map. When the pixel point corresponding to the pixel coordinate is a non-target pixel point, the assignment is 0. When the pixel point corresponding to the pixel coordinate is a red pixel point, the assignment is 1. When the pixel point corresponding to the pixel coordinate is a yellow pixel point, the assignment is 2. When the pixel point corresponding to the pixel coordinate is a green pixel point, the assignment is 3. Based on multiple consecutive RYG distribution maps, obtain the assignment sequence corresponding to each pixel coordinate and denote it as , and the arrangement order of the elements in the assignment sequence is consistent with the sorting of the RYG distribution map.
[0064] Step 3: Based on the number of distinct non-zero elements (i.e., elements with assignments not equal to 0) in the assignment sequence corresponding to the pixel point, mark it as a pixel point with q color changes, where q is the number of distinct non-zero elements. For example, when there is only one type of numerical element in the assignment sequence, it means that the corresponding pixel point only produces one assignment change in multiple consecutive RYG distribution maps, which also means that the corresponding pixel point is a light source with single-color change (possibly corresponding to a single-color signal light in the traffic light). By determining that the pixel point is a pixel point with q color changes, the color change of the pixel point can be analyzed, so as to distinguish based on the number of changes of different pixel points, in order to screen out pixel points with similar characteristics to the traffic light changes for further comparison.
[0065] The process of obtaining the number of distinct elements is as follows:
[0066] Extract the non-zero elements in the assignment sequence in turn and perform subtraction operations with all the previous non-zero elements. When the operation results are all not equal to 0, denote this non-zero element as a distinct non-zero element, and count the number of distinct non-zero elements in the assignment sequence as the number of distinct elements.
[0067] Step 4: Screen out the pixels that change color once and remove the discrete pixels among them to obtain the distribution map of pixels that change color once. Denote the remaining pixels as the first target points. Based on the distribution of the first target points, determine the positions of the monochromatic change signal lights. It should be noted that a traffic signal light group generally consists of traffic signal lights of three colors: red, green, and yellow. Each individual traffic signal light generates a light source of one color and switches between on and off, which causes the corresponding pixels of the traffic signal light group to appear as pixels that change color once in the captured picture. Therefore, the pixels that change color once can be used as the features for preliminarily screening the area of the traffic signal light group.
[0068] Among them:
[0069] Obtain the horizontal distance between the traffic signal light group and the capture camera and denote it as ; obtain the vertical height difference between the traffic signal light group and the capture camera and denote it as ; obtain the mask diameter of the traffic signal light and denote it as ; obtain the interval between traffic signal lights and denote it as (the interval refers to the center interval); obtain the angle between the main optical axis of the capture camera's lens and the vertical plane and denote it as ; substitute into the formula for calculation to obtain the regional width threshold R, where are all preset weight coefficients. Construct multiple regional circles with a diameter equal to the regional width threshold in the distribution map of pixels that change color once;
[0070] It should be noted that the settings of the capture camera and the traffic signal lights have strict standards, and the installation data will be recorded during the installation process, such as the installation height, installation position, and installation angle. In addition, the specifications (mask diameter) of the traffic signal lights also need to be recorded in advance. Therefore, all the data in this step are existing directly inputtable data, which do not require secondary collection and can be directly used as known data.
[0071] Calculate the interval threshold through the formula , where is a preset weight coefficient. When the area ratio of the first target points in the regional circle (i.e., the number of the first target points divided by the total number of pixels in the regional circle) is greater than or equal to the preset ratio threshold, and the distance between the nearest first target point outside the regional circle and the regional circle is less than the interval threshold, denote this regional circle as the pending regional circle;
[0072] Obtain the centers of multiple circles of undetermined areas, calculate the distances between the centers of adjacent circles of undetermined areas and compare them with the center distance threshold. The center distance threshold is equal to the sum of the area width threshold and the interval threshold. When the distance between the centers of two adjacent circles of undetermined areas is less than or equal to the center distance threshold, they are recorded as a group of adjacent area circles. When there are two groups of adjacent area circles among any three circles of undetermined areas, the three circles of undetermined areas are recorded as a group of area circle combinations. Obtain the maximum angle of the triangle formed by the centers of the three circles in the area circle combination as the judgment angle. When the judgment angle is greater than or equal to the preset angle threshold, in a specific embodiment, the angle threshold is equal to 160°, the three areas corresponding to the three circles of undetermined areas are recorded as signal light areas.
[0073] It should be noted that according to GB14886 "Installation Specification for Road Traffic Signal Lights" and GB14887 "Road Traffic Signal Lights", the traffic signal lights (red, green, and yellow lights) installed in China currently are arranged strictly in groups of three, either vertically or horizontally. Therefore, by analyzing the arrangement of the centers of the three circles of undetermined areas in the area circle combination, it is possible to determine whether the three circles of undetermined areas are traffic signal lights that meet the set standards, so as to determine the position of the traffic signal lights in the captured picture.
[0074] More specifically, the discrete judgment process of pixel points is as follows:
[0075] Calculate the distance between any one first target point and each of the other first target points and take the minimum value as the discrete judgment value. There is a preset discrete judgment threshold. When the discrete judgment value of any one first target point is greater than or equal to the discrete judgment threshold, mark the first target point as a discrete point.
[0076] Through the analysis of multiple steps in the signal light recognition unit, it is possible to quickly lock the position area of the traffic signal light in the captured picture based on the captured picture of the capture camera, so as to perform corresponding fault diagnosis on the traffic signal light based on the image changes in this area. It can not only detect the fault manifestations of the traffic signal light in a timely manner, but also make full use of the image data information obtained by the capture camera, enrich the functions of the capture camera, and is faster and more convenient than the existing fault judgment methods.
[0077] The light type discrimination unit marks the traffic signal lights corresponding to the signal light areas as red signal lights, yellow signal lights, and green signal lights based on the light source hues corresponding to the pixel points in the signal light areas. Calculate the discrimination threshold based on the pattern areas of different pattern signal lights, and distinguish the types of traffic signal lights corresponding to the signal light areas according to the discrimination threshold, and mark them as arrow indicator lights or circular indicator lights.
[0078] It should be understood that the pixel points within the signal light area are all single-color-changing pixel points, and each single-color-changing pixel point corresponds to one of the light source hues of red, green, and yellow. Therefore, the corresponding light source hue within the signal light area can be determined, that is, the light source color of the traffic signal light.
[0079] Where:
[0080] There are preset arrow-shaped standard patterns and circular standard patterns (referring to the standard patterns of signal lights with different patterns under the same lampshade specification. When the lampshade specification changes, the standard patterns are enlarged or reduced proportionally). Calculate the pattern areas of the arrow-shaped standard pattern and the circular standard pattern respectively and denote them as , and substitute them into the formula to calculate the discrimination threshold , where DF is the proportion threshold;
[0081] Calculate the area proportion of the first target point in the signal light area and denote it as the proportion judgment value. When the proportion judgment value in the signal light area is less than the discrimination threshold, the signal light area is marked as an arrow signal light area, corresponding to an arrow indicator light. When the proportion judgment value in the signal light area is greater than or equal to the discrimination threshold, the signal light area is marked as a circular signal light area, corresponding to a circular indicator light.
[0082] The fault judgment unit independently monitors each signal light area, analyzes the pixel points in the signal light area based on the traffic signal light color and type corresponding to the signal light area, judges whether the traffic signal light has abnormal display, and further analyzes the degree of the fault. Where:
[0083] Denote the monitored signal light area as the target area, obtain the color of the traffic signal light corresponding to the target area as the target color, obtain the most recent captured picture and count the number of target color pixel points in the target area as the quantity parameter value. When the quantity parameter value is greater than 0 and less than or equal to the quantity lower limit value, generate a damage signal;
[0084] When the quantity parameter value is greater than the quantity lower limit value and less than the quantity detection threshold, perform pattern abnormality detection.
[0085] Quantity lower limit value The calculation formula is , where is the preset weight coefficient, S is the pattern area of the standard pattern of the pattern signal light corresponding to the target area (refer to in the lamp type discrimination unit), and P is the preset proportion threshold (taking a value of 50% in a specific embodiment, according to the regulations on the failure detection function of light-emitting diodes (LEDs) in GB14887 "Road Traffic Signal Lights");
[0086] Quantity detection threshold The calculation formula is , where is a preset detection trigger ratio value (taking the value of 90% in a specific embodiment).
[0087] The pattern abnormality detection process is as follows:
[0088] When the type of the traffic signal corresponding to the target area is a circular indicator light, a maintenance signal is generated;
[0089] When the type of the traffic signal corresponding to the target area is an arrow indicator light, a reference arrow contour is constructed (obtained based on the contour of the target pixel points when the traffic signal corresponding to the target area is working normally). A plurality of evenly distributed reference points are drawn in the reference arrow contour, and all the target color pixel points in the current target area are marked in the reference arrow contour. The discrete pixel points are removed (refer to step four in the signal lamp recognition unit). The distance between each reference point and the nearest target color pixel point is calculated and recorded as the distribution distance. When the distribution distance is greater than or equal to the preset distribution threshold, the reference point is recorded as a discrete reference point. The proportion of the discrete reference points in all the reference points is calculated to obtain the coverage ratio. When the coverage ratio is less than or equal to the coverage threshold, a pattern missing signal is generated, indicating that there is an obvious missing part in the signal lamp pattern composed of the remaining LED lamps that can be normally lit in this traffic signal, and the signal indication function cannot be normally completed.
[0090] It should be noted that the pattern abnormality detection step can further analyze the pattern of the traffic signal, so as to timely detect the incomplete indication pattern caused by the damage of the lamp beads in the traffic signal, so that the staff can replace them as soon as possible to ensure the normal use of this traffic signal. The present invention can judge the fault from the perspective of the actual working use of the traffic signal. Compared with the existing traffic signal lamp bead fault detection means, it is more advanced and intelligent, especially suitable for the situation where the number of damaged lamp beads does not reach the fault reporting limit value, but the actual working effect is affected.
[0091] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the protection scope of the technical solutions of the embodiments of the present invention.
Claims
1. A high-power LED lamp bead group control system based on wireless communication, including a control center, characterized in that: Also includes: Wireless communication unit: establishes wireless communication connection between the control center and the traffic light group and snapshot camera. When the traffic light group generates a traffic signal change, it generates a light shear signal and sends it to the control center. The signal light recognition unit takes a snapshot when the control center obtains the light shear signal, and selects the red pixels, yellow pixels, and green pixels in the snapshot, which are collectively referred to as target pixels. The RYG distribution map is obtained based on the distribution of target pixels, and a single pixel is assigned a value, where: Non-target pixels are assigned a value of 0, and red pixels, yellow pixels, and green pixels are assigned a non-zero integer value; Based on multiple continuous RYG distribution maps, overlapping analysis is performed to construct an assignment sequence. Based on the mutually different number q of non-zero elements in the assignment sequence, the corresponding pixel points are marked as q-time color-changing pixels, and the pixels that change color once are screened out to construct multiple area circles. Based on the distribution of the pixels that change color once in the area circles, the undetermined area circles are screened out. Based on the spatial distribution of the multiple undetermined area circles, the signal light area is screened out, and the signal light area corresponds to a traffic light. The process of constructing the area circle is as follows: Filter out the pixels that change color once and remove the discrete pixels to obtain the distribution map of color change once, and record the remaining pixels as the first target points; Get the horizontal distance between the traffic light group and the capture camera and record it as , get the vertical height difference between the traffic light group and the capture camera and record it as , get the mask diameter of the traffic light and record it as , get the interval between traffic lights and record it as , get the angle between the main optical axis of the camera lens and the vertical plane and record it as ; Substitute into the formula Calculate in and get the region width threshold R, where All are preset weight coefficients; Construct multiple area circles with diameters equal to the area width threshold in the one-time color change distribution map; A light type distinguishing unit, which divides the traffic lights into different colors based on the light source hue corresponding to the pixel points in the traffic light area, calculates a distinguishing threshold based on the pattern area of the traffic lights with different patterns, and distinguishes the types of traffic lights according to the distinguishing threshold; The fault judgment unit performs fault analysis on the traffic lights corresponding to each traffic light area.
2. According to the wireless communication-based high-power LED lamp group control system of claim 1, it is characterized in that: The target pixel screening process is as follows: The picture currently captured by the capture camera is obtained as the captured picture, and the captured picture is converted into an HSV color space to obtain an HSV captured picture. Any pixel point in the HSV captured picture corresponds to a hue, saturation, and brightness; There are preset red hue value intervals, yellow hue value intervals and green hue value intervals. Each hue value interval corresponds to the hue value range of a color light in the HSV snapshot image. Pixel points whose hue values in the HSV snapshot image are in the red hue value interval, the yellow hue value interval and the green hue value interval are extracted and recorded as red pixels, yellow pixels and green pixels respectively. The three types of pixels are collectively referred to as target pixels. Non-target pixels in the HSV snapshot image are eliminated to obtain a RYG distribution map. Each RYG distribution map corresponds to a shooting moment.
3. According to the wireless communication-based high-power LED lamp group control system of claim 1, it is characterized in that: The construction process of the assignment sequence is as follows: A rectangular coordinate system is constructed with one of the corners of the captured image as the origin of the coordinates and recorded as the overlapping coordinate system. The unit length of the overlapping coordinate system is one pixel. The coordinate positions of all pixels in the captured image in the overlapping coordinate system are obtained and recorded as pixel coordinates; Based on the RYG distribution map, each pixel coordinate is assigned a value. When the pixel point corresponding to the pixel coordinate is a non-target pixel point, the value is assigned to 0. When the pixel point corresponding to the pixel coordinate is a red pixel point, the value is assigned to 1. When the pixel point corresponding to the pixel coordinate is a yellow pixel point, the value is assigned to 2. When the pixel point corresponding to the pixel coordinate is a green pixel point, the value is assigned to 3. Based on multiple continuous RYG distribution maps, the value series corresponding to each pixel coordinate is obtained and recorded as , k is the shooting sequence number of the RYG distribution map, and the order of elements in the assigned sequence is consistent with the order of the RYG distribution map.
4. According to the wireless communication-based high-power LED lamp group control system of claim 1, it is characterized in that: The process of obtaining different quantities is as follows: The non-zero elements in the assigned sequence are extracted in sequence and subtracted from all the previous non-zero elements. When the operation results are not equal to 0, the non-zero element is recorded as a mutually distinct non-zero element. The number of mutually distinct non-zero elements in the assigned sequence is counted and recorded as the mutually distinct number.
5. The high-power LED lamp bead group control system based on wireless communication according to claim 1, characterized in that: The discrete judgment process of pixels is as follows: The distance between any first target point and each other first target point is calculated and the minimum value is taken as the discrete judgment value. A discrete judgment threshold is preset. When the discrete judgment value of any first target point is greater than or equal to the discrete judgment threshold, the first target point is marked as a discrete point.
6. A high-power LED lamp bead group control system based on wireless communication according to claim 5, characterized in that: The signal light area screening process is as follows: Calculate interval threshold , the interval threshold calculation formula is as follows: ,in is the preset weight coefficient; When the area ratio of the first target point in the area circle is greater than or equal to the preset ratio threshold, and the distance between the first target point outside the area circle and the area circle is less than the interval threshold, the area circle is recorded as a pending area circle; Get the centers of multiple pending area circles, calculate the center distance between two adjacent pending area circles and compare it with the center distance threshold. The center distance threshold is equal to the sum of the area width threshold and the interval threshold. When the center distance between two adjacent pending area circles is less than or equal to the center distance threshold, they are recorded as a group of adjacent area circles. When there are two groups of adjacent area circles among any three area circles to be determined, the three area circles to be determined are recorded as a group of area circle combination, and the maximum angle of the triangle formed by the centers of the three area circles to be determined in the area circle combination is obtained and recorded as the judgment degree; when the judgment degree is greater than or equal to the preset degree threshold, the three areas corresponding to the three area circles to be determined are recorded as traffic light areas.
7. A high-power LED lamp bead group control system based on wireless communication according to claim 6, characterized in that: The process of distinguishing the types of traffic lights is as follows: There are preset arrow-shaped standard patterns and circular standard patterns. The pattern areas of the arrow-shaped standard patterns and circular standard patterns are calculated and recorded as , substitute into the formula The distinction threshold is calculated in , where DF is the preset proportion threshold; The area ratio of the first target point in the traffic light area is calculated and recorded as a ratio judgment value. When the ratio judgment value in the traffic light area is less than the distinction threshold, the traffic light area is marked as an arrow traffic light area, corresponding to an arrow indicator light; When the proportion judgment value in the traffic light area is greater than or equal to the distinction threshold, the traffic light area is marked as a circular traffic light area, corresponding to a circular indicator light.
8. A high-power LED lamp bead group control system based on wireless communication according to claim 7, characterized in that: The fault analysis process of the fault judgment unit is as follows: The monitored traffic light area is recorded as the target area, the color of the traffic light corresponding to the target area is obtained as the target color, the most recently captured image is obtained and the number of target color pixels in the target area is counted as the quantity parameter value, and when the quantity parameter value is greater than 0 and less than or equal to the quantity lower limit value, a damage signal is generated; When the quantity parameter value is greater than the quantity lower limit and less than the quantity detection threshold, pattern anomaly detection is performed, where: Quantity lower limit The calculation formula is ,in is a preset weight coefficient, S is the pattern area of the standard pattern of the pattern signal light corresponding to the target area, and P is a preset ratio threshold; Quantity detection threshold The calculation formula is ,in It is the preset detection trigger ratio value.
9. A high-power LED lamp bead group control system based on wireless communication according to claim 8, characterized in that: The pattern anomaly detection process is as follows: When the type of the traffic light corresponding to the target area is a round indicator light, a maintenance signal is generated; When the type of traffic light corresponding to the target area is an arrow indicator light, a reference arrow outline is constructed, a plurality of evenly distributed reference points are drawn in the reference arrow outline, all target color pixel points in the current target area are marked in the reference arrow outline, discrete pixel points are removed, and the distance between each reference point and the nearest target color pixel point is calculated and recorded as the distribution distance; When the distribution distance is greater than or equal to the preset distribution threshold, the reference point is recorded as a discrete reference point, and the ratio of the discrete reference point to all reference points is calculated to obtain the coverage ratio. When the coverage ratio is less than or equal to the coverage threshold, the generated pattern lacks a signal.
Citation Information
Patent Citations
Video vehicle characteristic detection system and detection method based on combination of long exposure and short exposure
CN104134352A
Program and device for discriminating traffic light
JP2011216051A