Distribution line private change monitoring alarm device based on image recognition

Abstract

The invention relates to the technical field of power equipment maintenance, in particular to a distribution line private change monitoring alarm device based on image recognition, which comprises an image acquisition device and an image processing device. And the image acquisition device comprises a micro camera arranged on the door plate. When the door is opened or closed, the photographing control device starts the miniature camera to shoot a new circuit diagram when the door plate rotates to a set angle. The image processing device detects whether the new line diagram is the same as the existing line diagram or not, and if not, an alarm signal is sent out. When the distribution box or the power distribution cabinet is opened and closed, images before and after the cabinet door or the cabinet door is opened are obtained in a photographing mode, whether the phenomenon that a wire is changed privately occurs or not is determined in an image comparison mode, and if the phenomenon is found, an alarm signal is sent out in time; inspection personnel or maintenance personnel can check on site in time, and hidden dangers can be eliminated in time.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment maintenance, and in particular to a distribution line unauthorized modification monitoring and alarm device based on image recognition. Background Art

[0002] Distribution boxes and distribution cabinets are the main power equipment for power distribution. In order to meet the needs of power users and the requirements of safe operation, distribution boxes and distribution cabinets are carefully calculated and designed by designers. For example, the capacity of circuit breakers and the capacity of wires must operate within reasonable rated data. However, after the construction of substations and distribution stations is completed, some users will privately set some loads and connect the privately set loads to the inside of the distribution box or distribution cabinet. Or, privately swap the output circuits of two loads and other unreasonable operations. Both of the above operations will lead to a mismatch between the load and the capacity of the equipment, which will lead to heat accumulation due to overload or overload of the equipment. In the case of light, the service life of the equipment will be reduced, and in the case of severe cases, accidents such as fire will be caused, affecting the safe operation of the power supply system and the personal and property safety of personnel. Therefore, it is an urgent requirement to design a distribution line unauthorized modification monitoring and alarm device based on image recognition that can promptly detect unauthorized modification of the line. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a distribution line unauthorized modification monitoring and alarm device based on image recognition, which can timely detect unauthorized modification of the line.

[0004] The technical solution to the technical problem to be solved by the present invention is: a distribution line unauthorized modification monitoring and alarm device based on image recognition, including an image acquisition device and an image processing device; the image acquisition device includes a shooting bracket and a shooting control device arranged on the door panel, and a miniature camera is installed on the shooting bracket; when the door is opened or closed, the shooting control device starts the miniature camera to shoot a new circuit diagram when the door panel is rotated to a set angle; the image processing device detects whether the new circuit diagram is the same as the existing circuit diagram, and sends an alarm signal if they are different.

[0005] Better yet, the shooting bracket is provided with a micro pan-tilt for adjusting the angle of the micro camera, and an angle sensor is provided on the door axis to detect the angle of rotation of the door panel. When it is detected that the door panel has rotated to a set angle, the micro pan-tilt is started to control the micro camera toward the wiring line.

[0006] Preferably, a side wall fulcrum is provided on the side wall, a shooting bracket is provided in the middle of the door panel, and a door panel fulcrum is provided on the side of the door panel close to the rotating shaft; the shooting bracket also includes an upper pillar, a lower pillar, a roller, and a shooting platform, the shooting platform is hollow cylindrical and a through hole is opened on one side for the micro camera to collect images, the upper pillar and the lower pillar are connected to the shooting platform through the rotating shaft, the roller is provided on the rotating shaft, the side wall fulcrum, the door panel fulcrum and the roller are arranged on the same horizontal line, and a pull wire is provided on the roller, one end of the pull wire is fixedly connected to the side wall fulcrum, and the other end is connected to the door panel fulcrum through a telescopic device.

[0007] Preferably, the telescopic device is an elastic rope.

[0008] Preferably, the telescopic device is a wire roller provided with a torsion spring, and the wire roller is installed on the upper part of the door panel fulcrum.

[0009] Even better, start shooting when the door panel is opened at a 45-degree angle.

[0010] Preferably, the image processing process specifically includes the following steps:

[0011] Step 1: Identify the device nodes in the new image;

[0012] Step 2, identifying a wire segment connected to a device node in the new image as a node connection line and forming a detection line segment with the device node;

[0013] Step 3: compare all the detection line segments with the existing detection line segments one by one. If there is a group that is different, an alarm message is issued. If they are all the same, proceed to step 4.

[0014] Step 4: Compare the complete directions of all wires. If the direction of any wire is different from the existing direction, an alarm signal is issued.

[0015] Better yet, the actual length of the node connection line is set at 1-2cm.

[0016] Preferably, in step 4, the method for determining the direction of the wire includes the following steps:

[0017] First, number the four wires connected to the intersection;

[0018] Then, obtain the wire diameter of each wire. If the wire diameters are the same, the two wires with the same wire diameter are connected wires. If the wire diameters of the four wires are the same, obtain the wire textures, which include the wire colors and the patterns printed on the wires. If the wire textures are the same, the two wires with the same texture are connected wires. When the wire diameters and textures are the same:

[0019] Step 4.1, construct a plane coordinate system with the entire image area as the background;.

[0020] Step 4.2, obtaining the center position O of the intersection of the crossing wires;

[0021] Step 4.3, set two concentric circles with the center position as the center and obtain the intersection points of the concentric circles and the wires. The intersection points on each wire segment include the proximal point and the distal point;

[0022] Calculate the slope of the line segment formed by the proximal point and the center position of the four conductors; two adjacent conductors with similar slopes are connected conductors. When the difference in slope is less than the connectivity limit, they are determined to be two conductors with similar slopes.

[0023] Better yet, subtract the slope of the line segment formed by the pericenter and the apocenter from the slope of the line segment formed by the pericenter and the center position and take the absolute value; determine whether the absolute value is less than the trend limit; if it is greater than the trend limit, reduce the diameter of the concentric circle and execute steps 4.2 and 4.3 again.

[0024] The beneficial effects of the present invention are:

[0025] When the distribution box or cabinet is opened and closed, take photos to obtain images before and after the cabinet door is opened and closed, and determine whether the wires have been tampered with by private means through image comparison. If found, an alarm signal will be issued in time, and the auditors or maintenance personnel will go to the site to check and eliminate the hidden dangers in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic diagram of an image acquisition device according to an embodiment of the present invention.

[0027] Figure 2 It is a schematic diagram of the operating principle of an image acquisition device according to an embodiment of the present invention.

[0028] Figure 3 It is a schematic diagram of an image acquisition device according to an embodiment of the present invention.

[0029] Figure 4 The present invention is a schematic diagram of the installation of a shooting bracket and a micro camera according to an embodiment of the present invention.

[0030] In the figure: 240, miniature camera; 300, induction switch; 231, trigger; 103, reversing fulcrum; 230, pull wire; 220, roller; 210, shooting platform; 102, door panel fulcrum; 200, shooting bracket; 101, side wall fulcrum; DETAILED DESCRIPTION

[0031] In order to make the technical solutions and beneficial effects of the present invention clearer, the embodiments of the present invention are further explained in detail below.

[0032] The distribution line unauthorized modification monitoring and alarm device based on image recognition includes an image acquisition device and an image processing device. The image acquisition device takes an image of the distribution line inside the cabinet or the box when the distribution box or the distribution cabinet is opened or closed, that is, obtains the image information of the distribution line. The image processing device processes the acquired image, mainly performs image recognition and compares it with the existing distribution line image, and issues an alarm signal if an abnormality is found.

[0033] In order to take better pictures, the image acquisition device is set on the door panel. In order to better obtain the image, a certain shooting distance is required to obtain a more complete image. Therefore, the image acquisition device is set on the door panel so that it can be separated from the power distribution line by a certain distance when the door is opened, thereby achieving the effect of complete shooting. At the same time, it is necessary to set a function with camera direction adjustment to offset the change in camera angle caused by the rotation of the door panel, that is, when opening or closing the door, the micro camera 240 is started to shoot the new circuit diagram when the door panel is rotated to the set angle through the camera control device. This can be achieved in the following two ways. one

[0035] The photographing control device is a digital control device. The image acquisition device includes an angle sensor, a micro camera 240, a micro pan-tilt, and a controller. The micro pan-tilt and the micro camera 240 form a shooting bracket 200. The micro pan-tilt is installed in the middle of the door panel or on the edge of the door panel away from the door axis, and the micro camera 240 is set on the micro pan-tilt. The angle sensor, the micro camera 240, the micro pan-tilt and the controller are electrically connected, and the angle sensor is used to detect the rotation angle of the door panel, and the controller is used to realize automatic control.

[0036] A camera with a pan / tilt system can achieve changes in shooting angles, and can adjust the shooting range of up to approximately 360 degrees in three dimensions. A rotating platform and a servo are often used to control the angle. The application of pan / tilt systems in the field of cameras is already very mature, and will not be further explained here.

[0037] The angle sensor is used to detect the angle of rotation. Generally, the angle sensor is provided with a rotating shaft for realizing the rotation. In order to drive the rotation of the rotating shaft of the angle sensor, a 90-degree arc gear is provided at the position of the door panel near the door shaft, and one end of the arc gear is fixedly connected to the door panel. An angle sensor is provided on the side panel, and a gear meshing with the arc gear is provided on the rotating shaft of the angle sensor. The rotating shaft of the angle sensor is adjacent to the position where the arc gear is connected to the door panel. At this time, when the door is opened, the arc gear rotates along the axis of the door shaft driven by the door panel, and then drives the rotation of the angle sensor through the gear. The angle sensor is electrically connected to the controller. After the controller collects the rotation angle reaching the set angle, the angle of the micro camera 240 is adjusted. At this time, the camera is rotated to the opposite direction of the door panel according to the set angle, so that the camera faces the inside of the box or cabinet, that is, toward the location where the power distribution line is located. two

[0039] The image acquisition device includes a side wall fulcrum 101 arranged on the side wall of a box or cabinet, a shooting bracket 200 arranged in the middle of a box door or cabinet door, a door panel fulcrum 102 arranged near the rotating shaft of the box door or cabinet door, and a controller fixedly mounted on the door panel or side wall. In order to take a more complete picture, the shooting bracket 200 can be located in the middle in the vertical direction and at the edge away from one end of the rotating shaft.

[0040] The side wall fulcrum 101 is a column, and the length direction of the side wall fulcrum is perpendicular to the side wall of the box or cabinet. The door panel fulcrum 102 is a single column or a double column. The shooting bracket 200 includes an upper column and a lower column. A shooting platform 210 is arranged between the upper column and the lower column, and a miniature camera 240 is installed on the upper part of the shooting platform. The shooting platform is rotatably connected with the upper column and the lower column through a rotating shaft, and a roller 220 is arranged on the rotating shaft. The side wall fulcrum 101, the door panel fulcrum 102 and the roller are at the same horizontal height.

[0041] A pull wire 230 is wound around the roller 220. One end of the pull wire 230 is fixedly connected to the side wall fulcrum 101, and the other end is connected to the door panel fulcrum 102 through a telescopic device. The telescopic device is an elastic rope, in which case the door panel fulcrum 102 is an independent pillar, one end of the elastic rope is fixedly connected to the door panel fulcrum 102, and the other end is fixedly connected to the pull wire 230, wherein the length of the elastic rope is less than half of the distance between the shooting bracket 200 and the door panel fulcrum 102. Alternatively, the telescopic device is a wire roller provided with a torsion spring, and the wire roller is rotatably connected to the door panel fulcrum 102, in which case the door panel fulcrum 102 is a double pillar. The pull wire is wound around the wire roller for a length greater than one circle.

[0042] In the initial state, the camera faces the inside of the cabinet or box, and the door hinge is on the right side as an example. Figure 2As shown, when the door is opened, the door panel rotates counterclockwise around the rotation axis, and at the same time drives the door panel fulcrum 102 to rotate. Figure 2 It can be seen that during the rotation of the door panel support point 102, due to the relative change in the position of the door panel support point 102 and the shooting bracket 200, Figure 2 It can be seen that R2>R1, at this time, the change in the length of the pull wire 230 between the door panel fulcrum 102 and the side wall fulcrum 101 drives the telescopic device to stretch, and then drives the roller 220 to rotate. According to the actual size of the cabinet or box and the length of the side wall fulcrum 101 and the door panel fulcrum 102, the diameter of the roller is matched so that during the stretching and pulling process, the direction of the camera driven by the roller 220 is always kept toward the inside of the cabinet or box.

[0043] In order to better obtain the image, the shooting is started when the door panel is opened at an angle of 45 degrees. Correspondingly, an induction switch 300 is arranged at the position covered by the pull line between the door panel fulcrum 102 and the shooting bracket 200, and a trigger 231 is arranged on the pull line. A certain length is spaced between the initial position trigger and the induction switch, such as Figure 2 It can be seen that L1>L2. During the rotation of the door panel, the position interval between the trigger 231 and the induction switch 300 becomes smaller. When the door panel rotates to an angle of 45 degrees, the trigger overlaps the induction switch. The induction switch can select a magnetic induction switch, an eddy current induction switch or a Hall induction switch, and the corresponding trigger can select metal iron or a magnetic block. The induction switch, the controller and the camera are electrically connected. The induction switch is used to control the camera, and the shooting is started after the induction switch is closed. The power supply of this device can use the power supply inside the distribution box or the distribution cabinet.

[0044] The camera is most suitable to be located in the middle of the door panel edge. In order to reduce the influence of the pull wire 230 on daily maintenance, the side wall fulcrum 101, the door panel fulcrum 102 and the roller 220 can be arranged on the upper part of the door panel and the side wall, thereby avoiding affecting the operation of the maintenance personnel.

[0045] Alternatively, in order to reduce the impact of the pull wire 230 on the operating space of the maintenance personnel, a reversing fulcrum 103 is set near the door panel fulcrum 102, wherein the length of the reversing fulcrum 103 is greater than the length of the side wall fulcrum 101. This design can achieve the change of the relative position of the side wall fulcrum 101 and the reversing fulcrum 103 to achieve the pulling of the pull rope. The reversing fulcrum 103 is provided with a ring or a through hole, and the pull wire 230 passes through the ring or the through hole. Alternatively, a reversing pulley is set on the reversing fulcrum 103, and the pull wire 230 is wound on the reversing pulley, so that when the door panel is closed, the pull wire 230 close to the door panel and the pull wire 230 close to the side wall remain in a nearly vertical state, that is, the pull wire is finally close to the door panel and the side wall, thereby providing a wider operating space for the maintenance personnel.

[0046] The image processing device is arranged in the controller. The image processing device stores an image processing program module for processing the collected image data. The image processing process specifically includes the following steps.

[0047] Step 1: Perform image recognition processing on the newly photographed new circuit diagram. The object to be recognized is the device node, where the device node is the end point and the starting point of the wire in the new circuit diagram.

[0048] In actual distribution boxes and distribution cabinets, the nodes of internal equipment are mainly terminal blocks and input and output terminals of electrical components. The terminals of electrical components include open terminals, lead terminals of current transformers and voltage transformers, etc. In the present application, the device nodes also include some virtual nodes. For example, a section of wire passes through a baffle or the bottom of a device due to inappropriate wiring space. At this time, the new circuit diagram shows that the wire disappears at the obstructed position. Therefore, the position where the wire disappears, that is, the cut-off point of the wire, is also a device node. Although the device node does not have a wire connected to the terminal, it is still processed as a device node, and this device node is defined as a virtual device node.

[0049] In the early stage of system operation, different device images are input into the big data learning model to build an image recognition model, and then the device nodes are identified through the image recognition model. After obtaining the new circuit diagram, the device nodes can be identified through the image recognition model.

[0050] In addition to building an image recognition model for device nodes, a wire recognition model needs to be built before the system runs. The wire recognition can use image segmentation algorithms, such as edge detection or region growing algorithms, to extract the wire area in the wiring cabinet image. After the wire is recognized, the end point of the wire is found, and it is determined whether the end point of the wire is connected to the actual device node. If not, the end point of the wire is set to a virtual device node.

[0051] After the device nodes are identified, each device node is numbered and marked, wherein the marking content includes the pixel coordinates of the image area, the pixel information of the image area, and the image device type, and the above information corresponds to a unique number.

[0052] Step 2: Construct a detection line segment, where the detection line segment includes a device node and a wire segment connected to the device node, which is a node connection line. The actual length of the node connection line is set at 1-2 cm. According to the ratio of the image to the actual distribution box or distribution cabinet, the wire image of the corresponding length is intercepted in the image.

[0053] After the device node is identified, the wire image connected to the device node is identified around the device node, and the wire segment of the corresponding length is obtained as the node connection line, and the node connection line is numbered and marked. The marking content includes line type, wire diameter, corresponding device node, etc. At the same time, the number information of the node connection line is set in the marking information of the device node.

[0054] As described in step 1, a wire recognition model is provided, and the corresponding wire is recognized in step 1. After the wire is recognized, the end point of the wire, i.e., the virtual device node, is determined. A wire segment connecting the virtual device node and the device node constitutes a detection line segment.

[0055] At this point, a list of associations between devices and wires is formed, that is, a list of relationships between devices and wires connected to the devices. After each identification and establishment of the association list, the list data is saved.

[0056] Step 3: Compare the detection line segment with the existing detection line segment. In general, the distribution line will not be easily changed, so the equipment node is fixed and the wires connected to it are also fixed. By comparing, it can be determined whether the problem of unauthorized line modification has occurred.

[0057] The saved association list of the detection line segment includes the existing association list and the new association list. The existing association list is the association list saved after the last door opening and detection. The detection line segments in the two association lists are compared one by one to determine whether the wire connected to the device node has changed.

[0058] During detection, it is necessary to find the same device node and then determine whether the node connection lines connected to it are the same. The pixel coordinate information of the existing device node can be compared to determine which new device node the existing device node corresponds to, and then the existing node connection line can be obtained according to the tag information of the existing device node. Finally, it is determined whether the existing node connection line connected to the device node is the same as the new node connection line.

[0059] After all the detection line segments are compared one by one, if one detection line segment is different, an alarm message will be issued. If there is a difference, it means that the wire connected to the device node has been rerouted.

[0060] For virtual device nodes, the consistency of the device nodes must be determined first. Since the structure inside the distribution box or distribution cabinet is relatively compact, and if the lines are not intentionally modified, the displacement of the wires inside the distribution box or distribution cabinet is limited in the event of accidental contact, so a moving range is set. Determine the relative position of the new device node and the existing device node. If the relative position is less than the moving range, it is determined that the existing device node and the new device node belong to the same device node, and then determine the detection line segment corresponding to the device node for comparison.

[0061] If the comparison results of the existing detection line segment and the new detection line segment are the same, proceed to step 4 to perform an in-depth detection of the distribution line.

[0062] Step 4: In this step, the complete direction of the wire is compared. In this step, the complete path of a wire is identified and judged whether it is the same as the existing path, that is, whether the device nodes at both ends of the wire of the complete path are the same. If they are the same, it means that the wiring line has not changed.

[0063] In the actual distribution box, the problem of wire crossing is inevitable. In the image, there will also be multiple line segments of wire crossing. After the wires are identified in step 2, due to the problem of wire crossing, in addition to identifying the complete wire segment from the starting device node to the ending device node, multiple incomplete line segments will also be identified. The incomplete line segment refers to a device node at at least one end of which is a virtual device node. Therefore, it is necessary to identify the connectivity of the two pairs of single-segment wires at the intersection, that is, to identify which two of the four wires at the intersection are one wire. In order to determine the connectivity of the wires, the following steps are performed.

[0064] First, the four wires connected to the intersection are numbered. If the intersection of wires is most common, two wires cross at one point. The probability of three wires crossing at one point is also possible but less. The method of this embodiment is also applicable to three wires crossing at one point, that is, the six wires connected to the intersection need to be numbered. Taking four wires as an example, the four wires are numbered 1, 2, 3, and 4 respectively.

[0065] Then, obtain the wire diameter of each conductor. If the wire diameters are the same, the two wires with the same wire diameter are connected. In the power system, the wire diameters of wires with different voltage levels and different current carrying capacity are different. Therefore, the wire diameter can be determined by determining which two of the four wires at the intersection are one wire. After the wire is identified, the wire diameter can be measured based on the width of the wire image at the intersection, where the width of the wire can be represented by the cumulative calculation of pixels. If the width is the same, it means the wire diameter is the same.

[0066] If the wire diameters of the four wires are the same, the wire textures are obtained. If the wire textures are the same, the two wires with the same texture are connected wires. Through image extraction, the texture of the surface of the identified wire can be obtained, where the texture includes the color of the wire and the graphics printed on the wire. In the power system, in order to distinguish the guides with different functions and uses, the guides will be distinguished by textures. For example, the three-phase electricity A, B, and C are distinguished by yellow, green, and red, the neutral wire N is distinguished by blue or black, and the ground wire is identified by yellow and green textures. If the wire diameters are the same, the wire texture is judged to distinguish which two of the four wires are connected.

[0067] When the wire diameter and the texture are the same, the continuity identification method is started to identify the continuity of the wire, which specifically includes the following steps:

[0068] Step 4.1: Construct a plane coordinate system with the entire image area as the background. Constructing a plane coordinate system can facilitate data calculation.

[0069] Step 4.2: Get the center position O of the intersection of the crossing wires, that is, the center point. First, get the intersection point, and then extract its coordinates in the coordinate system.

[0070] In order to calculate better, the identified wires are processed first. After the wires are identified, they are long strips with a certain width, which can be seen as an area surrounded by two parallel curves. The center line of the parallel curves composed of two edges can be used as the substitute line of the wire, and the coordinates of each point are extracted. Then the intersection of the two substitute lines is the center position O.

[0071] Step 4.3, set two concentric circles with the center position as the center and obtain the intersection points of the concentric circles and the wires. The intersection points on each wire segment include the proximal point A and the distal point B. In addition to setting two concentric circles, three concentric circles can also be set. As the number of concentric circles increases, the complexity of the calculation will increase. In this embodiment, two concentric circles are taken as an example.

[0072] After setting two concentric circles, eight points are generated after they intersect with the four defined wires, namely A1, B1, A2, B2, A3, B3, A4, B4. A1 is the intersection of the inner circle and wire No. 1, that is, the near center of wire No. 1, and A2 is the intersection of the outer circle and wire No. 1, that is, the far center of wire No. 1. The definitions of the intersection points of other wires are the same.

[0073] Calculate the adjacent slope K of the line segment composed of the pericentric point and the center position, that is, the slope of the line segment OA. The slopes K1, K2, K3, and K4 of the four line segments composed of the four wire segments OA1, OA2, OA3, and OA4 are calculated. Among the four slopes, the two wires with similar adjacent slopes are connected wires. When the difference of the adjacent slopes is less than the connected limit, they are determined to be two wires with similar slopes. Then extract two of the four slopes in turn and calculate the difference, and finally find two slopes with similar adjacent slopes.

[0074] In the present application, when making a comparison, the absolute value of the difference between adjacent slopes is within 0.1 as a similar slope. Among them, 0.1 is the connectivity limit, which is used as a parameter in the actual processing process and can be set according to the actual situation. For example, for a large-capacity distribution cabinet or distribution box, that is, the current flowing through the wire is large, the wire used is relatively hard and not easy to bend. At this time, the connectivity limit can be set smaller. If it is a distribution box or distribution cabinet used for control, the wires inside are mostly control wires, which are easily squeezed to cause large arc bending. At this time, the connectivity limit can be set larger. After calculating the load connectivity limit requirements for one group, calculate whether the difference between the remaining two slopes meets the limit. If it also meets the limit, it means that the two groups of two segments that meet the calculation results obtained according to the above division are connected, that is, the two segments are one wire.

[0075] During the calculation process, due to the irregular bending shape of the wire, it may not be possible to find two adjacent slopes with similar slopes. At this time, it is necessary to verify the change trend of the four segments of wire. At this time, the slope Ky, that is, the slope of segment AB, is calculated and verified. Finally, the slopes of the four segments are calculated to be Ky1, Ky2, Ky3, and Ky4.

[0076] Calculate the absolute value of the difference between the verification slope and the adjacent slope, that is, subtract the slope Ky of the line segment composed of the proximal point and the distal point from the adjacent slope K of the line segment composed of the proximal point and the center position and take the absolute value |Ky-K|. Determine whether the absolute value is less than the trend limit. If it is greater than the trend limit, reduce the diameter of the concentric circle and execute steps 4.2 and 4.3 again. Among them, the trend limit and the connectivity limit are similar, both of which are used as a parameter in the calculation process. Appropriate limits can be set according to the actual distribution system. If it is less than the trend limit, it means that the change trend of the conductor is similar, that is, the curvature of the bend is small and the change is relatively gentle. If it is greater than the trend limit, it means that the curve changes greatly. At this time, when comparing the adjacent slopes, the adjacent slopes may not reflect the actual change trend of the conductor, which in turn affects the judgment result.

[0077] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Through the above description, relevant staff can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the contents of the specification. All so-called equal changes and modifications of the shapes, structures, features and spirits described in the scope of the claims of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A distribution line tampering monitoring and alarm device based on image recognition, characterized in that: It includes an image acquisition device and an image processing device; The image acquisition device comprises a shooting bracket (200) and a shooting control device arranged on the door panel, wherein a micro camera (240) is installed on the shooting bracket (200); when the door is opened or closed, the shooting control device activates the micro camera (240) to shoot a new route map when the door panel rotates to a set angle; The image processing device detects whether the new circuit diagram is the same as the existing circuit diagram, and sends out an alarm signal if they are different.

2. The distribution line tampering monitoring and alarm device based on image recognition according to claim 1 is characterized in that: The shooting bracket (200) is provided with a micro-pan head for adjusting the angle of the micro-camera (240), and an angle sensor is provided on the door axis for detecting the angle of rotation of the door panel. When it is detected that the door panel has rotated to a set angle, the micro-pan head is activated to control the micro-camera (240) to face the wiring line.

3. The distribution line tampering monitoring and alarm device based on image recognition according to claim 1 is characterized in that: A side wall fulcrum (101) is arranged on the side wall, a shooting bracket (200) is arranged in the middle of the door panel, and a door panel fulcrum (102) is arranged on the side of the door panel close to the rotating shaft; the shooting bracket (200) further comprises an upper support, a lower support, a roller (220), and a shooting platform (210); the shooting platform (210) is hollow and cylindrical and has a through hole on one side for collecting images by a micro camera (240); the upper support and the lower support are connected to the shooting platform (210) via a rotating shaft; the roller (220) is arranged on the rotating shaft; the side wall fulcrum (101), the door panel fulcrum (102) and the roller (220) are arranged on the same horizontal line; and a pull wire (230) is arranged on the roller (220); one end of the pull wire (230) is fixedly connected to the side wall fulcrum (101), and the other end is connected to the door panel fulcrum (102) via a telescopic device.

4. The distribution line tampering monitoring and alarm device based on image recognition according to claim 3 is characterized in that: The telescopic device is an elastic rope.

5. The distribution line tampering monitoring and alarm device based on image recognition according to claim 3 is characterized in that: The telescopic device is a wire roller provided with a torsion spring, and the wire roller is installed on the upper part of the door panel support point (102).

6. The distribution line tampering monitoring and alarm device based on image recognition according to any one of claims 1 to 5, characterized in that: Start shooting when the door panel is opened at a 45-degree angle.

7. The distribution line tampering monitoring and alarm device based on image recognition according to claim 6 is characterized in that: The image processing process specifically includes the following steps: Step 1: Identify the device nodes in the new image; Step 2, identifying a wire segment connected to a device node in the new image as a node connection line and forming a detection line segment with the device node; Step 3: compare all the detection line segments with the existing detection line segments one by one. If there is a group that is different, an alarm message is issued. If they are all the same, proceed to step 4. Step 4: Compare the complete directions of all wires. If the direction of any wire is different from the existing direction, an alarm signal is issued.

8. The distribution line tampering monitoring and alarm device based on image recognition according to claim 7 is characterized in that: The actual length of the node connection line is set at 1-2cm.

9. The distribution line tampering monitoring and alarm device based on image recognition according to claim 7 is characterized in that: In step 4, the method for determining the direction of the wire includes the following steps: First, number the four wires connected to the intersection; Then, obtain the wire diameter of each wire. If the wire diameters are the same, the two wires with the same wire diameter are connected wires. If the wire diameters of the four wires are the same, obtain the wire textures, which include the wire colors and the patterns printed on the wires. If the wire textures are the same, the two wires with the same texture are connected wires. When the wire diameters and textures are the same: Step 4.1, construct a plane coordinate system with the entire image area as the background;. Step 4.2, obtaining the center position O of the intersection of the crossing wires; Step 4.3, set two concentric circles with the center position as the center and obtain the intersection points of the concentric circles and the wires. The intersection points on each wire segment include the proximal point and the distal point; Calculate the slope of the line segment formed by the proximal point and the center position of the four conductors; two adjacent conductors with similar slopes are connected conductors. When the difference in slope is less than the connectivity limit, they are determined to be two conductors with similar slopes.

10. The distribution line tampering monitoring and alarm device based on image recognition according to claim 9 is characterized in that: Subtract the slope of the line segment formed by the pericenter and the apocenter from the slope of the line segment formed by the pericenter and the center position and take the absolute value; determine whether the absolute value is less than the trend limit. If it is greater than the trend limit, reduce the diameter of the concentric circle and execute steps 4.2 and 4.3 again.

Images