A method and system for detecting loose installation of cameras based on video intelligent analysis

Abstract

The present invention relates to a method and system for detecting loose camera installation based on video intelligent analysis, the method comprising the following steps: after the camera is installed, the maximum jitter amplitude of the camera in a normal operation scenario is initialized and calibrated, and the jitter amplitude value of the normal jitter of the picture when the camera is firmly installed is obtained, and the jitter amplitude value is saved accordingly; during operation and maintenance, the jitter amplitude value of the camera picture jitter is calculated regularly, and the calculated camera picture jitter amplitude value is subtracted from the pre-saved normal picture jitter amplitude value corresponding to the camera to obtain a change difference, and the change difference is compared with the pre-saved alarm threshold corresponding to the camera, and if the alarm threshold is exceeded, an alarm is issued. The present invention can quickly screen out cameras with potential safety hazards, such as detached cameras and loose cameras, and perform targeted maintenance, greatly reducing the maintenance workload.

Description

Technical Field

[0001] The invention belongs to the technical field of video monitoring, and in particular relates to a method and a system for detecting loose installation of a camera based on video intelligent analysis. Background Art

[0002] The existing installation methods of cameras in rail transit mostly adopt a single bracket type, hanging type or ceiling type installation method. In places with relatively harsh environments such as beside the track or on the top of the elevated station, the camera is in a state of vibration and corrosion for a long time, and it is easy for the installation anchor bolts and other fixings to loosen, corrode and rust, causing the camera to fall, which will not only cause damage to the camera, but also affect driving safety and the personal safety of pedestrians, posing certain safety hazards.

[0003] During operation, cameras installed in tunnels or on elevated roads are periodically affected by wind pressure and vibration from passing vehicles, causing the image to shake. The amplitude of the shake is related to factors such as vehicle speed, bracket rigidity, arm length, and looseness of the installation surface, among which the looseness of the installation surface has the greatest impact.

[0004] Therefore, manual inspections are required to check the installation firmness of the cameras on a regular basis, which adds a lot of workload to operations and maintenance. Summary of the invention

[0005] The purpose of the present invention is to overcome at least one defect in the prior art and provide a method and system for detecting loose installation of cameras based on video intelligent analysis, which is intended to solve the problems of heavy maintenance workload of cameras and difficulty in detecting loose installations for operation and maintenance personnel.

[0006] The technical solution of the present invention is implemented as follows: The present invention discloses a method for detecting loose installation of a camera based on video intelligent analysis, comprising the following steps:

[0007] During operation and maintenance, the jitter amplitude value of the camera image is calculated regularly, and the calculated jitter amplitude value of the camera image is subtracted from the pre-saved jitter amplitude value of the normal image jitter corresponding to the camera to obtain the change difference. The change difference is compared with the pre-saved alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

[0008] Furthermore, after the camera is installed, the maximum jitter amplitude of the camera in a normal operation scenario is initialized and calibrated to obtain a jitter amplitude value of normal image jitter when the camera is firmly installed, and save it accordingly.

[0009] Furthermore, the maximum jitter amplitude of the camera in a normal operation scenario is initialized and calibrated to obtain a jitter amplitude value of normal picture jitter when the camera is firmly installed, specifically including: when the camera is firmly installed, the original picture taken by the camera is processed by an anti-shake algorithm to obtain a cropped and reduced picture, and the pixel size difference a1 of the X component and the pixel size difference b1 of the Y component between the cropped and reduced picture and the original picture are calculated.

[0010] Furthermore, the amplitude value of the camera screen jitter is calculated periodically, specifically including: periodically capturing an original monitoring picture at a certain moment from the video clip shot by the camera, processing the original monitoring picture with an anti-shake algorithm to obtain a cropped and reduced picture, and calculating a pixel size difference a2 of the X component and a pixel size difference b2 of the Y component between the cropped and reduced picture and the original monitoring picture.

[0011] Further, the calculated pixel size difference a2 of the X component is subtracted from the pre-stored jitter amplitude value a1 of the X component when the picture corresponding to the camera is normally jittering to obtain the change difference of the X component, and then the change difference of the X component is compared with the pre-stored first alarm threshold corresponding to the camera. If the first alarm threshold is exceeded, an alarm is issued. The calculated pixel size difference b2 of the Y component is subtracted from the pre-stored jitter amplitude value b1 of the Y component when the picture corresponding to the camera is normally jittering to obtain the change difference of the Y component, and then the change difference of the Y component is compared with the pre-stored second alarm threshold corresponding to the camera. If the second alarm threshold is exceeded, an alarm is issued.

[0012] Furthermore, after subtracting the calculated camera image jitter amplitude value from the pre-saved normal image jitter amplitude value corresponding to the camera to obtain the change difference, the change difference is divided by the pre-saved normal image jitter amplitude value corresponding to the camera to obtain the change percentage, and the change percentage is compared with the alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

[0013] The present invention also discloses a system for detecting loose installation of a camera based on video intelligent analysis, comprising a camera and an analysis server, wherein the camera is used to shoot video images, and the analysis server is used to obtain video image data shot by the camera, calculate the jitter amplitude value of the camera image jitter, subtract the calculated camera image jitter amplitude value from the pre-saved jitter amplitude value of the normal image jitter corresponding to the camera to obtain a change difference, compare the change difference with the pre-saved alarm threshold value corresponding to the camera, and if the alarm threshold value is exceeded, an alarm is issued.

[0014] Furthermore, the analysis server is used to subtract the calculated camera image jitter amplitude value from the pre-saved normal image jitter amplitude value corresponding to the camera to obtain a change difference, and then divide the change difference by the pre-saved normal image jitter amplitude value corresponding to the camera to obtain a change percentage, and compare the change percentage with the alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

[0015] Furthermore, the system for detecting loose camera installation based on video intelligent analysis of the present invention also includes a video surveillance system, wherein the camera is used to transmit the captured video images to the video surveillance system, and the analysis server is used to obtain video data from the video surveillance system for computational analysis.

[0016] Furthermore, the camera is installed and fixed on the mounting frame by a traditional installation method, and a protective rope is connected between the camera and the mounting frame, one end of the protective rope is connected to the camera, and the other end of the protective rope is connected to the mounting frame; the camera has a network cable / optical cable interface for connecting to the network cable / optical cable, the camera is connected to the video surveillance system via the network cable / optical cable, the network cable / optical cable interface of the camera is plugged into and matched with the connector of the network cable / optical cable, when the camera falls, the connector of the network cable / optical cable falls off from the network cable / optical cable interface of the camera, causing the camera to go offline, triggering the video surveillance system to generate an alarm.

[0017] The present invention has at least the following beneficial effects: the present invention introduces a screen crossing line (train passing) alarm and a screen anti-shake algorithm, and captures and outputs a screen shot of a train passing video clip processed by the anti-shake algorithm, and a screen image that has been cropped and reduced is obtained. The difference in pixel size between this image and the original image can be calculated, and it can be used as a judgment parameter for the camera shaking amplitude. As long as the screen shaking amplitude is initialized and calibrated once when a train passes by, a normal shaking constant is obtained, and then the screen shaking amplitude during the train passing is detected regularly to determine whether the shaking amplitude exceeds the initial constant, and the difference is compared with the alarm threshold. If it exceeds, an alarm prompt is issued.

[0018] The method of the present invention can timely detect the installation firmness of the video surveillance camera of the operating line at a relatively low cost, achieve targeted manual inspection and review, and greatly reduce the work pressure and workload of manual regular inspection.

[0019] The camera double protection installation structure of the present invention, on the basis of adopting the traditional installation method to install and fix the camera, further arranges a protective steel wire to be connected with the camera, so as to provide redundant protection for the installation of the camera. When the camera installation fixture is loose and causes the camera to fall, the protective steel wire can hold the camera and the mounting seat, so that the camera will not fall onto the track to affect driving, or fall from a height to affect personal safety, which greatly improves the safety of the installed camera and reduces the maintenance workload.

[0020] The camera double protection installation structure of the present invention only needs to set a fixed support ear on the camera and the mounting base to connect with one end of the protection wire, and fix the protection wire on the building on the other side. It is simple to implement, with very little added cost and strong feasibility.

[0021] The present invention arranges a steel wire and a stainless steel tie for fixing the network cable / optical cable at the interface of the camera, ensuring that when the camera falls, the network cable / optical cable falls off and the camera is offline. The network manager of the video monitoring system can quickly find the offline camera and accurately determine the faulty camera, so as to perform targeted maintenance on the camera.

[0022] After the implementation of the present invention, this installation method, on the one hand, greatly improves the safety of the installed cameras and basically eliminates safety hazards. On the other hand, it can quickly screen out cameras with safety hazards such as detached cameras and loose cameras, and perform targeted maintenance, greatly reducing the maintenance workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 A flow chart of a method for detecting loose camera installation based on video intelligent analysis provided by an embodiment of the present invention;

[0025] Figure 2 A schematic diagram of the principle of detecting the amplitude jitter value during initialization calibration provided by an embodiment of the present invention;

[0026] Figure 3 A schematic diagram of the principle of detecting the jitter amplitude value at each moment during operation provided by an embodiment of the present invention;

[0027] Figure 4 A schematic diagram of a camera installation structure provided by an embodiment of the present invention.

[0028] In the accompanying drawings, 1 is a camera, 2 is a camera mounting base, 3 is a mounting bracket, 4 is a protective rope, 5 is a fixing lug, 6 is a network cable / optical cable, and 7 is a fixing rope. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] In the description of the present invention, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0031] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features; in the description of the present invention, unless otherwise specified, "multiple" and "several" mean two or more.

[0032] Embodiment 1

[0033] See also Figure 1 The embodiment of the present invention discloses a method for detecting loose installation of a camera based on video intelligent analysis, comprising the following steps:

[0034] After the camera is installed, the maximum jitter amplitude of the camera in normal operation scenarios (including when a train passes by) is initialized and calibrated to obtain the jitter amplitude value of normal image jitter when the camera is firmly installed, and save it accordingly.

[0035] During operation and maintenance, the jitter amplitude value of the camera screen jitter is calculated regularly, and the calculated jitter amplitude value of the camera screen jitter is subtracted from the pre-saved jitter amplitude value of the normal screen jitter corresponding to the camera to obtain the change difference value, and the change difference value is compared with the pre-saved alarm threshold value corresponding to the camera. If the alarm threshold value is exceeded, an alarm is issued. Furthermore, after the calculated jitter amplitude value of the camera screen jitter is subtracted from the pre-saved jitter amplitude value of the normal screen jitter corresponding to the camera to obtain the change difference value, the change difference value is divided by the pre-saved jitter amplitude value of the normal screen jitter corresponding to the camera to obtain the change percentage, and the change percentage is compared with the alarm threshold value corresponding to the camera. If the alarm threshold value is exceeded, an alarm is issued.

[0036] On-site inspection by operation and maintenance personnel can change the inspection mode from checking all cameras one by one to checking the installation status of specific cameras after an alarm, which greatly reduces the workload of operation and maintenance. The alarm threshold is obtained based on experiments and experience.

[0037] The present invention can calculate the jitter amplitude value of the camera image jitter regularly (at a certain frequency, which can be set as needed) for alarm analysis as needed, and can also calculate the jitter amplitude value of the camera image jitter for alarm analysis when a train passes by.

[0038] Furthermore, the maximum jitter amplitude of the camera in the normal operation scenario is initialized and calibrated to obtain the jitter amplitude value of the normal jitter of the image when the camera is firmly installed, specifically including: when the camera is firmly installed, the original image taken by the camera is processed by the anti-shake algorithm to obtain a cropped image, and the pixel size difference a1 of the X component and the pixel size difference b1 of the Y component between the cropped image and the original image are calculated, see Figure 2 , the center of the cropped image is the same as the center of the original image. Of course, the pre-saved amplitude values ​​a1 and b1 of the normal jitter of the image corresponding to the camera can be adjusted according to needs and experience, or may not be adjusted.

[0039] Further, the jitter amplitude value of the camera image jitter is calculated periodically, specifically including: periodically capturing an original monitoring picture at a certain moment from the video clip shot by the camera, processing the original monitoring picture with an anti-shake algorithm to obtain a cropped and reduced picture, and calculating the pixel size difference a2 of the X component and the pixel size difference b2 of the Y component between the cropped and reduced picture and the original monitoring picture, see Figure 3 , the center of the cropped image and the original surveillance image are the same point.

[0040] Further, the calculated pixel size difference a2 of the X component is subtracted from the pre-stored jitter amplitude value a1 of the X component when the picture corresponding to the camera is normally jittering to obtain the change difference of the X component, and then the change difference of the X component is compared with the pre-stored first alarm threshold corresponding to the camera. If the first alarm threshold is exceeded, an alarm is issued. The calculated pixel size difference b2 of the Y component is subtracted from the pre-stored jitter amplitude value b1 of the Y component when the picture corresponding to the camera is normally jittering to obtain the change difference of the Y component, and then the change difference of the Y component is compared with the pre-stored second alarm threshold corresponding to the camera. If the second alarm threshold is exceeded, an alarm is issued.

[0041] A specific embodiment is as follows: after obtaining the change difference of the X component and the change difference of the Y component, the change difference of the X component is divided by the pre-stored amplitude value a1 of the normal jitter of the image corresponding to the camera to obtain the change percentage of the X component, and the change percentage of the X component is compared with the first alarm threshold corresponding to the camera, and if it exceeds the first alarm threshold, an alarm is issued. The change difference of the Y component is divided by the pre-stored amplitude value b1 of the normal jitter of the image corresponding to the camera to obtain the change percentage of the Y component, and the change percentage of the Y component is compared with the second alarm threshold corresponding to the camera, and if it exceeds the second alarm threshold, an alarm is issued.

[0042] Furthermore, when the camera is firmly installed, an initial calibration of the image jitter amplitude is performed when a vehicle passes by, and the jitter amplitude value of the normal image jitter when the camera is firmly installed is obtained. During operation, the video is downloaded each time a vehicle passes by, and analyzed and processed to obtain the jitter amplitude X component and jitter amplitude Y component of the image jitter detected this time, and calculate the jitter amplitude value of the camera image jitter.

[0043] After the method of the embodiment of the present invention is implemented, this detection method can quickly screen out cameras with potential safety hazards, shorten the detection cycle, and on the other hand can also greatly reduce the workload of on-road maintenance.

[0044] Embodiment 2

[0045] The embodiment of the present invention discloses a system for detecting loose installation of a camera based on video intelligent analysis, comprising a camera and an analysis server, wherein the camera is used to shoot video images, and the analysis server is used to obtain video image data shot by the camera, calculate the jitter amplitude value of the camera image jitter, subtract the calculated camera image jitter amplitude value from the pre-saved jitter amplitude value of the normal image jitter corresponding to the camera to obtain a change difference, compare the change difference with the pre-saved alarm threshold corresponding to the camera, and if the alarm threshold is exceeded, an alarm is issued.

[0046] Furthermore, the analysis server is used to subtract the calculated camera image jitter amplitude value from the pre-saved normal image jitter amplitude value corresponding to the camera to obtain a change difference, and then divide the change difference by the pre-saved normal image jitter amplitude value corresponding to the camera to obtain a change percentage, and compare the change percentage with the alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

[0047] Furthermore, the system for detecting loose installation of a camera based on video intelligent analysis of the present invention also includes a video surveillance system, wherein the camera is used to transmit the captured video images to the video surveillance system, and the analysis server is used to obtain video data from the video surveillance system for calculation and analysis. The specific scheme for the calculation and analysis performed by the analysis server of this embodiment is shown in Example 1.

[0048] Further, see Figure 4 When the camera 1 has a camera mounting seat 2, the camera 1 is mounted on the camera mounting seat 2, the camera mounting seat 2 is fixed on the mounting frame 3 by a conventional mounting method, one end of the protective rope 4 is fixedly connected to the camera 1 and / or the camera mounting seat 2, and the other end of the protective rope 4 is fixedly connected to the mounting frame 3. When the camera 1 does not have a camera mounting seat 2 or the camera 1 and the camera mounting seat 2 are integrated, the camera 1 is fixedly mounted on the mounting frame 3 by a conventional mounting method, one end of the protective rope 4 is fixedly connected to the camera 1, and the other end of the protective rope 4 is fixedly connected to the mounting frame 3.

[0049] The camera has a network cable / optical cable interface for connecting to a network cable / optical cable. The camera is connected to a video surveillance system via the network cable / optical cable. The network cable / optical cable interface of the camera is plugged into and matched with a connector of the network cable / optical cable. When the camera falls, the connector of the network cable / optical cable falls off from the network cable / optical cable interface of the camera, causing the camera to go offline and triggering the video surveillance system to generate an alarm.

[0050] The video surveillance system of the present invention can adopt an existing structure. For example, a video surveillance system of an embodiment includes a video server, an optical fiber switch or an optical fiber transceiver, and a monitoring center. The camera 1 is connected to the video server via a network cable / optical cable 6, and the video server is connected to the monitoring center via an optical fiber switch or an optical fiber transceiver. Of course, the video surveillance system of the present invention is not limited to the above embodiment.

[0051] Furthermore, at least one of the camera 1 and the camera mounting base 2 is provided with a fixing lug 5 for fixedly connecting to the protection rope 4 ; and the fixing lug 5 is provided with a connecting hole for fixedly connecting to the protection rope 4 .

[0052] Furthermore, the mounting frame 3 is provided with a fixing lug 5 for fixedly connecting with the protection rope 4 ; the fixing lug 5 is provided with a connecting hole for fixedly connecting with the protection rope 4 .

[0053] Furthermore, the double-protection camera 1 installation structure of the present invention also includes a fixing rope 7, the camera 1 has a network cable / optical cable interface for connecting to the network cable / optical cable 6, the network cable / optical cable interface of the camera 1 is plugged into and matched with the connector of the network cable / optical cable 6, one end of the fixing rope 7 is connected to the network cable / optical cable 6, and the connection is close to the connector of the network cable / optical cable 6, and the other end of the fixing rope 7 is fixedly connected to the mounting bracket 3 to ensure that when the camera 1 falls, the connector of the network cable / optical cable 6 falls off from the network cable / optical cable interface of the camera, making the camera 1 offline.

[0054] Furthermore, the camera 1 is connected to the video surveillance system via a network cable / optical cable 6 .

[0055] Furthermore, the fixing rope 7 is fixedly connected to a fixing lug 5 provided on the mounting frame 3 ; and the fixing lug 5 is provided with a connecting hole for being fixedly connected to the fixing rope 7 .

[0056] Further, the network cable / optical cable 6 near the joint of the network cable / optical cable 6 is provided with a fixing piece for fixed connection with the fixing rope 7. For example, a stainless steel cable tie is tied at the interface where the camera 1 is connected to the network cable / optical cable 6, and is also fixed by steel wire.

[0057] Furthermore, the camera 1 has a network cable / optical cable interface for connecting to the network cable / optical cable 6, and the network cable / optical cable interface of the camera 1 is plugged into and matched with the connector of the network cable / optical cable 6. When the camera 1 falls, the length of the protective rope 4 can cause the connector of the network cable / optical cable 6 to fall off from the network cable / optical cable interface of the camera, causing the camera 1 to go offline.

[0058] The length of the protective rope 4 in the present invention is set according to actual needs. The length of the fixing rope 7 in the present invention is set according to actual needs. Figure 1 The structure of the protective rope 4 and the fixing rope 7 is merely illustrated, and the lengths of the protective rope 4 and the fixing rope 7 are not limited.

[0059] Furthermore, when the camera 1 falls, the length of the protective rope 4 can keep the camera 1 suspended in the air.

[0060] Furthermore, the protective rope 4 is made of corrosion-resistant material.

[0061] Preferably, the protective rope 4 is a steel wire rope.

[0062] Furthermore, conventional installation methods include bracket type, hanging type or ceiling type.

[0063] In addition to the traditional hanging and bracket-type installation and fixing solutions, the camera 1 of the present invention also provides a redundant protection installation method for the installation of the camera 1 through a protective steel wire (or other materials), and only one protective steel wire (applicable to the case where the base and the camera 1 are integrated) or two steel wires are respectively provided to fix the camera 1 and the base. The present invention fixes the steel wire (or other materials) with the cable tie to ensure that the cable of the camera 1 falls off, thereby triggering the network management alarm.

[0064] The above is the double protection installation method proposed by this patent, which is suitable for side wall bracket installation and hanging installation of camera 1. The steel wire can be replaced with other corrosion-resistant materials, and it is sufficient to bear the weight of camera 1. When the mounting fixture of camera 1 loosens and causes camera 1 to fall, the protective steel wire can hold camera 1 and the mounting base to prevent it from falling onto the track and affecting driving, or falling from a height and affecting personal safety. On the other hand, a steel wire is set to fix the network cable / optical cable 6 at the interface of camera 1 with a stainless steel cable tie to ensure that when camera 1 falls, the network cable / optical cable 6 falls off and the camera 1 will be offline, thereby triggering the video surveillance system to generate an alarm. The video surveillance system network manager can quickly find the offline camera 1 and can accurately determine the faulty camera 1, so as to perform targeted maintenance on the camera 1.

[0065] This patent also proposes a method for detecting loose camera installation by video analysis. By setting up a background video analysis server, the camera's screen jitter amplitude is detected. A normal value is set for each camera. By detecting the change in the jitter amplitude value during operation and maintenance, the occurrence of loose camera installation is found. After the alarm is triggered, the operation and maintenance personnel are arranged to conduct focused inspections, which greatly reduces the maintenance workload. This method only requires setting up an analysis server in the background to obtain video data from the existing video surveillance system through the GB / T28181 standard protocol for analysis to obtain the analysis results. It is simple to implement, with minimal additional costs, and is highly feasible.

[0066] The present invention is suitable for use in the installation of surveillance cameras 1 in locations with harsh environments such as rail transit tracks and the top of elevated stations. The installation method of the present invention only requires that a fixing lug 5 be provided on the camera 1 for connecting the protective steel wire, and the protective steel wire be fixed on the mounting frame 3 on the other side, such as a building. The implementation is simple, the cost is minimal, and the feasibility is strong.

[0067] After the implementation of the present invention, this installation structure can greatly improve the safety of the installed camera 1 and basically eliminate potential safety hazards. On the other hand, it can also perform targeted maintenance on the detached camera 1 and greatly reduce the maintenance workload.

[0068] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A system for detecting loose installation of cameras based on video intelligent analysis, characterized in that: The invention comprises a camera and an analysis server, wherein the camera is used to shoot video images, the analysis server is used to obtain video image data shot by the camera, and the analysis server is used to implement a method for detecting loose installation of a camera based on video intelligent analysis, which specifically comprises: During operation and maintenance, the jitter amplitude value of the camera screen is calculated regularly, specifically including: regularly capturing the original monitoring picture at a certain moment from the video clip shot by the camera, processing the original monitoring picture with an anti-shake algorithm to obtain a cropped and reduced picture, and calculating the pixel size difference a2 of the X component and the pixel size difference b2 of the Y component between the cropped and reduced picture and the original monitoring picture; Subtract the calculated camera image jitter amplitude value from the pre-stored normal image jitter amplitude value corresponding to the camera to obtain a change difference, and compare the change difference with the pre-stored alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued, specifically including: Subtract the calculated pixel size difference a2 of the X component from the pre-stored X component jitter amplitude value a1 when the image corresponding to the camera is normally jittering to obtain the change difference of the X component, then compare the change difference of the X component with the pre-stored first alarm threshold corresponding to the camera, if it exceeds the first alarm threshold, alarm is issued, subtract the calculated pixel size difference b2 of the Y component from the pre-stored Y component jitter amplitude value b1 when the image corresponding to the camera is normally jittering to obtain the change difference of the Y component, then compare the change difference of the Y component with the pre-stored second alarm threshold corresponding to the camera, if it exceeds the second alarm threshold, alarm is issued; After the camera is installed, the maximum jitter amplitude of the camera in the normal operation scenario is initialized and calibrated to obtain the jitter amplitude value of the normal jitter of the picture when the camera is firmly installed, and save it accordingly; Initialize and calibrate the maximum jitter amplitude of the camera in a normal operation scenario to obtain the jitter amplitude value of the normal jitter of the image when the camera is firmly installed, specifically including: when the camera is firmly installed, process the original image taken by the camera with an anti-shake algorithm to obtain a cropped image, and calculate the pixel size difference a1 of the X component and the pixel size difference b1 of the Y component between the cropped image and the original image; It also includes a video surveillance system, wherein the camera is used to transmit the captured video images to the video surveillance system, and the analysis server is used to obtain video data from the video surveillance system for calculation and analysis; The camera has a network cable / optical cable interface for connecting to a network cable / optical cable. The camera is connected to a video surveillance system via the network cable / optical cable. The network cable / optical cable interface of the camera is plugged into and matched with a connector of the network cable / optical cable. One end of a fixing rope is connected to the network cable / optical cable, and the connection is close to the connector of the network cable / optical cable. The other end of the fixing rope is fixedly connected to a mounting bracket to ensure that when the camera falls, the connector of the network cable / optical cable falls off from the network cable / optical cable interface of the camera, causing the camera to go offline, thereby triggering an alarm in the video surveillance system.

2. The system for detecting loose installation of cameras based on video intelligent analysis as claimed in claim 1, characterized in that: The analysis server is used to subtract the calculated camera image jitter amplitude value from the pre-saved normal image jitter amplitude value of the camera to obtain a change difference, then divide the change difference by the pre-saved normal image jitter amplitude value of the camera to obtain a change percentage, and compare the change percentage with the alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

3. The system for detecting loose installation of a camera based on video intelligent analysis according to claim 1 or 2, characterized in that: The camera is installed and fixed on the mounting frame by a traditional installation method. A protective rope is connected between the camera and the mounting frame. One end of the protective rope is connected to the camera, and the other end of the protective rope is connected to the mounting frame.

4. A method for detecting loose installation of cameras based on video intelligent analysis, characterized in that: The method comprises the following steps: during operation and maintenance, the jitter amplitude value of the camera screen jitter is calculated at regular intervals, specifically comprising: regularly capturing an original monitoring picture at a certain moment from a video clip shot by the camera, processing the original monitoring picture with an anti-shake algorithm to obtain a cropped and reduced screen picture, and calculating a pixel size difference a2 of an X component and a pixel size difference b2 of a Y component between the cropped and reduced screen picture and the original monitoring picture; Subtract the calculated camera image jitter amplitude value from the pre-stored normal image jitter amplitude value corresponding to the camera to obtain a change difference, and compare the change difference with the pre-stored alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued, specifically including: Subtract the calculated pixel size difference a2 of the X component from the pre-stored X component jitter amplitude value a1 when the image corresponding to the camera is normally jittering to obtain the change difference of the X component, then compare the change difference of the X component with the pre-stored first alarm threshold corresponding to the camera, if it exceeds the first alarm threshold, alarm is issued, subtract the calculated pixel size difference b2 of the Y component from the pre-stored Y component jitter amplitude value b1 when the image corresponding to the camera is normally jittering to obtain the change difference of the Y component, then compare the change difference of the Y component with the pre-stored second alarm threshold corresponding to the camera, if it exceeds the second alarm threshold, alarm is issued; After the camera is installed, the maximum jitter amplitude of the camera in the normal operation scenario is initialized and calibrated to obtain the jitter amplitude value of the normal jitter of the picture when the camera is firmly installed, and save it accordingly; The maximum jitter amplitude of the camera in a normal operation scenario is initialized and calibrated to obtain the jitter amplitude value of the normal jitter of the picture when the camera is firmly installed, specifically including: when the camera is firmly installed, the original picture taken by the camera is processed by an anti-shake algorithm to obtain a cropped and reduced picture, and the pixel size difference a1 of the X component and the pixel size difference b1 of the Y component between the cropped and reduced picture and the original picture are calculated.

5. The method for detecting loose installation of a camera based on video intelligent analysis as claimed in claim 4, characterized in that: After subtracting the calculated camera image jitter amplitude value from the pre-saved normal image jitter amplitude value corresponding to the camera to obtain the change difference, divide the change difference by the pre-saved normal image jitter amplitude value corresponding to the camera to obtain the change percentage, and compare the change percentage with the alarm threshold corresponding to the camera. If the alarm threshold is exceeded, an alarm is issued.

Images