Intelligent Internet of Things operating system and control method thereof

Abstract

The invention relates to the field of intelligent Internet of Things, in particular to an intelligent Internet of Things operating system and a control method thereof.The intelligent Internet of Things operating system is provided with a plurality of image acquisition devices arranged in groups and comprises a shooting module used for shooting real-time images of a plurality of to-be-monitored points and generating corresponding image data, and a generation module used for filtering the image data and sending the filtered image data to a server; the filtering module is used for filtering the filtering data and generating corresponding filtering data and a shielding rate learning model, the learning module is used for preprocessing the filtering data, selecting a plurality of index characteristics for learning and generating a shielding rate distribution diagram of a real-time image, the alarm module is used for receiving the shielding rate distribution diagram and comparing the shielding rate distribution diagram with a shielding rate threshold value to generate a corresponding comparison result, and the object driving module is used for driving the object to move. The method and the device are used for adjusting the sheltering object of the to-be-monitored point according to the comparison result, the sheltering problem of the camera of the to-be-monitored point is solved to a certain extent, the problem that the image is not clear due to factors such as fog is reduced, the recognition accuracy of an Internet of Things operating system is improved, and the actual experience of a user is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent Internet of Things, and in particular to an intelligent Internet of Things operating system and a control method thereof. Background Art

[0002] Security cameras are semiconductor imaging devices with the advantages of high sensitivity, strong light resistance, low distortion, small size, long life, and vibration resistance. Most existing security cameras do not have anti-blocking functions and are easily blocked by falling debris or dust, making it impossible for security cameras to collect video images normally, losing their security protection and monitoring functions, and posing a large potential safety hazard. In addition, since most existing security cameras are installed at a high position, dust easily accumulates on the surface after long-term use, making them difficult to clean, which brings inconvenience to the cleaning of security cameras. Therefore, ensuring that security cameras can collect video images normally, play a role in security protection and monitoring, and avoiding being blocked by falling debris or dust that affects the security camera's shooting line of sight has become an important issue.

[0003] Chinese patent application publication number: CN115695979A discloses an anti-obstruction camera for Internet of Things monitoring, which relates to the technical field of security monitoring equipment. It includes a camera and a mounting base detachably mounted on the camera, the mounting base includes a lower support plate, a sleeve plate, a rectangular side rod and a U-shaped clamping rod detachably mounted on the rectangular side rod. A threaded rod is provided on the U-shaped clamping rod for rotation, and a threaded adapter hole is provided on the threaded rod for rotation with the threaded rod. A tightening disc is fixedly installed on the side of the threaded rod close to the sleeve plate, and a cleaning device for anti-obstruction cleaning of the camera is provided on the mounting base. The invention has the effect of clearing debris falling on the camera, effectively cleaning the surface of the camera, preventing dust from adhering to and blocking the camera's shooting line of sight, and facilitating disassembly and installation for most cameras.

[0004] Chinese patent application publication number: CN112153258A discloses an anti-obstruction monitoring camera based on intelligent manufacturing of Internet of Things engineering, including a housing, the inner wall of the housing is connected to a fixed rod, the inner side of the fixed cylinder is slidably connected to a first moving rod, the upper side of the first moving rod is connected to a support plate, the lower side of the support plate is fixedly connected to a connecting plate, the lower side of the connecting plate is fixedly connected to a first track, the inner side of the first track is slidably connected to a first driving block, and the upper side of the driving block is fixedly connected to a connecting frame. The anti-obstruction monitoring camera based on intelligent manufacturing of Internet of Things engineering, when the driving block moves to the ramp of the first track, the driving block drives the first track to move upward along the ramp, thereby driving the support plate fixedly connected to the connecting plate to move outward, so that the support plate moves outside the camera. This structure achieves the effect that the volume of the hidden anti-obstruction device camera does not increase, and the built-in anti-obstruction device increases the service life.

[0005] However, the above method has the following problems: it is impossible to intelligently control the system in real time through the Internet of Things, which results in the inability to effectively adjust the monitoring equipment when the monitored object is blocked, resulting in reduced operating efficiency. Summary of the invention

[0006] To this end, the present invention provides an intelligent Internet of Things operating system and a control method thereof, so as to overcome the problem in the prior art that the system cannot be intelligently regulated in real time through the Internet of Things, resulting in the inability to effectively adjust the monitoring equipment when the monitored object is blocked, thereby reducing the operating efficiency.

[0007] To achieve the above object, the present invention provides an intelligent Internet of Things operating system, wherein the system is provided with a plurality of image acquisition devices arranged in groups, and for a single image acquisition device, the system comprises: A shooting module, which is used to shoot real-time images of several points to be monitored and generate corresponding image data; A generating module, connected to the shooting module, for filtering the image data and generating corresponding filtering data and a corresponding occlusion rate learning model; A learning module, connected to the generating module, for preprocessing the filtered data, generating corresponding preprocessed data, selecting several index features of the preprocessed data, learning the preprocessed data according to the index features, and generating an occlusion rate distribution map of the real-time image; an alarm module, connected to the learning module, for receiving the occlusion rate distribution map, comparing the occlusion rate distribution map with an occlusion rate threshold, and generating a corresponding comparison result; an object-removing module, which is connected to the alarm module and is used to adjust the obstruction of the point to be monitored according to the comparison result; Wherein, the real-time image is a real-time monitoring image of the point to be monitored; The occlusion rate learning model is trained and generated on the filtered data according to the filtered data set; The filtering data set includes corresponding filtering data generated by the image data of the monitored point at different periods; The preprocessed data is to fit the corresponding image according to the filtered data, and generate a number of preprocessed data with a sampling rate of a standard sampling rate, and the learning module generates a corresponding indicator feature according to each preprocessed data; The standard sampling rate is a sampling rate that can be recognized by the learning module, and for a single sampling, the corresponding standard sampling rate is a single sampling rate.

[0008] Furthermore, the shooting module includes: A plurality of image sensor groups are arranged at corresponding points to be monitored, and are used to respectively capture real-time images of the points to be monitored; A converter, connected to the image sensor group, for converting the real-time image into corresponding image data; A transmitter is connected to the converter and is used to collect the image data and transmit it to the generation module.

[0009] Furthermore, the generating module receives the image data, the filter filters the image data, and generates corresponding filtering data. Wherein, a pixel threshold is set in the filter to collect pixels of the image data and filter the image data exceeding the pixel threshold.

[0010] Furthermore, the generating module stores the filtered data, and trains the filtered data according to the filtered data set to generate a corresponding occlusion rate learning model.

[0011] Further, the learning module receives the occlusion rate learning model, and preprocesses the filtered data through a preprocessor to generate corresponding preprocessed data, the learner selects several indicator features of the preprocessed data, learns the preprocessed data according to the occlusion rate learning model, and generates a corresponding occlusion rate distribution map; The indicator features include the proportion of obstructions in the real-time image, time and / or the location of the point to be monitored; The preprocessed data is learned by converting the preprocessed data into the occlusion rate distribution map, extracting the index feature in the occlusion rate distribution map, and annotating the occlusion rate distribution map based on the index feature.

[0012] Further, the preprocessor preprocesses the filtered data to generate the corresponding preprocessed data, and preprocessing the filtered data further includes: Based on the data collection characteristics of the filtered data set, data cleaning is performed to obtain cleaned data of the filtered data set, and the cleaned data is standardized; The standardization process is to segment the filtered data according to a standard sampling rate.

[0013] Furthermore, the alarm module receives the occlusion rate distribution map, calculates the occlusion rate of each monitored point in the occlusion rate distribution map, compares the occlusion rate with the occlusion rate threshold, and when the occlusion rate is less than the occlusion rate threshold, the alarm module does not issue an adjustment instruction to the object-driving module.

[0014] Furthermore, when the occlusion rate is greater than the occlusion rate threshold, the regulator in the alarm module issues an adjustment instruction to the object-driving module, and the object-driving module receives the adjustment instruction and adjusts the angle of the image sensor group.

[0015] Furthermore, the image sensor group captures the real-time image of the point to be monitored again, and learns the occlusion rate of the real-time image until the occlusion rate is less than the occlusion rate threshold.

[0016] On the other hand, the present invention provides an intelligent Internet of Things control method, comprising: The generating module iteratively optimizes the occlusion rate learning model according to the filtered data set until the loss value converges to a preset value; Testing the current occlusion rate learning model using the filtered data set; When the test result meets the preset test result, the parameters of the occlusion rate learning model are saved.

[0017] Compared with the prior art, the present invention is provided with a plurality of grouped image acquisition devices, including: a shooting module, used for shooting real-time images of a plurality of points to be monitored and generating corresponding image data, a generation module, used for filtering the image data and generating corresponding filtering data and an occlusion rate learning model, a learning module, used for preprocessing the filtering data and selecting a plurality of indicator features for learning to generate an occlusion rate distribution map of the real-time image, an alarm module, used for receiving the occlusion rate distribution map and comparing it with the occlusion rate threshold to generate a corresponding comparison result, and an object driving module, used for adjusting the occlusions of the monitoring points according to the comparison results, thereby solving the occlusion problem of the camera of the monitoring points to a certain extent, reducing the problem of unclear images caused by factors such as fog, improving the recognition accuracy of the Internet of Things operating system, and improving the actual experience of users.

[0018] Furthermore, by setting an image sensor group in the shooting module to convert the real-time image into image data, the occlusion rate learning model is facilitated to learn the pre-processed data, thereby improving the accuracy of the system in identifying the image occlusion rate.

[0019] Furthermore, by filtering the image data in the generation module and comparing the pixels of the image data with the pixel threshold, the corresponding filtered data is obtained, thereby eliminating the errors in the learning results caused by incorrect processing of the filtered data during the learning process, making the learning results more accurate.

[0020] Furthermore, by using the filtered data set to train the filtered data and generate the corresponding occlusion rate learning model, the learning of the occlusion rate is realized, which facilitates the real-time adjustment of the shooting module and solves the occlusion problem of the camera at the monitoring point to a certain extent.

[0021] Furthermore, by preprocessing the filtered data, corresponding preprocessed data is obtained to improve the recognition efficiency of noise, so that effective features of the preprocessed data can be quickly selected, thereby further improving the recognition accuracy of the IoT operating system.

[0022] Furthermore, by setting an occlusion rate threshold and adjusting the image sensor group, the problem of unclear images caused by factors such as fog is reduced, thereby improving the actual user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a schematic diagram of the structure of a single image acquisition device in an intelligent Internet of Things operating system of the present invention; Figure 2 A side view of the vehicle to be monitored that is blocked; Figure 3 A top view of the vehicle to be monitored; Figure 4 This is a flow chart of an intelligent Internet of Things control method of the present invention; Among them: 1. Vehicles to be monitored; 2. Blocked vehicles. DETAILED DESCRIPTION

[0024] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0025] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0026] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0027] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0028] See also Figure 1 As shown, it is a structural schematic diagram of a single image acquisition device in an intelligent Internet of Things operating system of the present invention, including: A shooting module, which is used to shoot real-time images of several points to be monitored and generate corresponding image data; A generation module, which is connected to the shooting module, is used to filter the image data and generate corresponding filtering data and a corresponding occlusion rate learning model; A learning module, which is connected to the generation module, is used to pre-process the filtered data, generate corresponding pre-processed data, select several index features of the pre-processed data, learn the pre-processed data according to the index features, and generate an occlusion rate distribution map of the real-time image; An alarm module, which is connected to the learning module, is used to receive the occlusion rate distribution map, compare the occlusion rate distribution map with the occlusion rate threshold, and generate a corresponding comparison result; The object-removing module is connected to the alarm module and is used to adjust the obstruction of the monitoring point according to the comparison result; Among them, the occlusion rate learning model is trained and generated based on the filtered data set; The filtering data set includes corresponding filtering data generated by the image data of the monitoring point at different periods; The preprocessed data is to fit the corresponding image according to the filtered data and generate a number of preprocessed data with a sampling rate of the standard sampling rate. The learning module generates corresponding indicator features according to each preprocessed data; The standard sampling rate is a sampling rate that can be recognized by the learning module, and for a single sampling, the corresponding standard sampling rate is a single sampling rate.

[0029] Preferably, setting the sampling rate to 10 per second has a better learning effect on the preprocessed data, and for the system described in this application, the corresponding learning effect is optimal.

[0030] By providing a plurality of grouped image acquisition devices, including: a shooting module, used for shooting real-time images of a plurality of points to be monitored and generating corresponding image data, a generating module, used for filtering the image data and generating corresponding filtering data and an occlusion rate learning model, a learning module, used for preprocessing the filtering data and selecting a plurality of index features for learning, generating an occlusion rate distribution diagram of the real-time image, an alarm module, used for receiving the occlusion rate distribution diagram and comparing it with the occlusion rate threshold value to generate a corresponding comparison result, and an object driving module, used for adjusting the occlusion of the point to be monitored according to the comparison result, the occlusion problem of the camera of the point to be monitored is solved to a certain extent, the problem of unclear images caused by factors such as fog is reduced, the recognition accuracy of the Internet of Things operating system is improved, and the actual experience of users is improved.

[0031] Specifically, the shooting module includes a plurality of image sensor groups arranged at corresponding points to be monitored, and is used to respectively shoot real-time images of the plurality of points to be monitored and output them as image data.

[0032] By setting an image sensor group in the shooting module to convert the real-time image into image data, the occlusion rate learning model is facilitated to learn the preprocessed data, thereby improving the accuracy of the system in identifying the image occlusion rate.

[0033] Specifically, the generation module receives the image data, the filter filters the image data, and generates corresponding filtered data. The filter is provided with a pixel threshold for collecting pixels of the image data and filtering the image data exceeding the pixel threshold.

[0034] Embodiment 1: Set the pixel threshold to 1000px. If the pixel of image A is 800px, which is less than the pixel threshold of 1000px, the image A data is filtered. If the pixel of image B is 1200px, which is greater than the pixel threshold of 1000px, the image B data is retained to generate the corresponding filtered data.

[0035] By filtering the image data in the generation module and comparing the pixels of the image data with the pixel threshold, the corresponding filtered data is obtained, which eliminates the errors in the learning results caused by incorrect processing of the filtered data during the learning process, making the learning results more accurate.

[0036] Specifically, the generation module stores the filtered data, and trains the filtered data according to the filtered data set to generate a corresponding occlusion rate learning model.

[0037] By using the filtered data set to train the filtered data and generate the corresponding occlusion rate learning model, the learning of the occlusion rate is realized, which facilitates the real-time adjustment of the shooting module and solves the occlusion problem of the camera at the monitoring point to a certain extent.

[0038] Specifically, the learning module receives the occlusion rate learning model, and preprocesses the filtered data through the preprocessor to generate corresponding preprocessed data. The learner selects several indicator features of the preprocessed data, learns the preprocessed data according to the occlusion rate learning model, and generates a corresponding occlusion rate distribution map. Among them, the indicator characteristics include the proportion of obstructions in the real-time image, time and / or the location of the point to be monitored; The preprocessed data is learned by converting the preprocessed data into an occlusion rate distribution map, extracting indicator features in the occlusion rate distribution map, and annotating the occlusion rate distribution map based on the indicator features.

[0039] By preprocessing the filtered data, the corresponding preprocessed data is obtained to improve the recognition efficiency of noise, so that the effective features of the preprocessed data can be quickly selected, thereby further improving the recognition accuracy of the IoT operating system.

[0040] Specifically, the preprocessor preprocesses the filtered data to generate corresponding preprocessed data, and the preprocessing of the filtered data also includes: Based on the data collection characteristics of the filtered data set, data cleaning is performed to obtain the cleaned data of the filtered data set, and the cleaned data is standardized; The standardization process is to divide the filtered data according to the standard sampling rate.

[0041] Specifically, the alarm module receives the occlusion rate distribution map, calculates the occlusion rate of each monitored point in the occlusion rate distribution map, and compares the occlusion rate with the occlusion rate threshold. When the occlusion rate is less than the occlusion rate threshold, the alarm module does not issue an adjustment instruction to the object driving module.

[0042] Specifically, when the occlusion rate is greater than the occlusion rate threshold, the regulator in the alarm module issues an adjustment instruction to the object driving module, and the object driving module receives the adjustment instruction and adjusts the angle of the image sensor group.

[0043] Specifically, the image sensor group captures the real-time image of the monitoring point again, and learns the occlusion rate of the real-time image until the occlusion rate is less than the occlusion rate threshold.

[0044] Embodiment 2: Assuming the occlusion rate threshold is 20%, the alarm module calculates the occlusion rate of the monitored point A in the occlusion rate distribution diagram to be 10%. If 10% is less than the occlusion rate of 20%, the alarm module will not issue an adjustment instruction to the object-driving module.

[0045] Embodiment 3: Assuming the occlusion rate threshold is 20%, the alarm module calculates that the occlusion rate of the monitored point B in the occlusion rate distribution diagram is 25%. 25% is greater than the occlusion rate of 20%. The regulator in the alarm module issues an adjustment instruction to the object driving module. The object driving module receives the adjustment instruction and adjusts the angle of the image sensor group.

[0046] Please refer to Figure 2 and Figure 3 As shown, they are the side view and top view of the vehicle to be monitored, and the occlusion rate of the vehicle to be monitored is 80%: Among them, in the side view, the image sensor group captures the monitored vehicle 1 behind the obstructing vehicle 2, and the occlusion rate is higher than the occlusion rate threshold. The regulator in the alarm module issues an adjustment instruction to the object driving module, and the object driving module receives the adjustment instruction and adjusts the angle of the image sensor group. The image sensor group captures a top view of the monitored vehicle 1, and the obstructing vehicle 2 does not constitute an occlusion of the monitored vehicle 1, and the occlusion rate is 0%.

[0047] In a specific implementation, a camera module is provided on the image sensor. When the object driver adjusts the image sensor group, the position of the camera module in the horizontal or vertical direction is changed to adjust the angle of the captured image. For example, by tilting the camera module upward or downward, the vertical field of view of the captured image can be changed, and by moving the camera to the left or right, the horizontal field of view of the captured image can be changed. Through the learning results of the real-time image, the camera module can be adjusted in real time to ensure the clarity of the image.

[0048] By setting the occlusion rate threshold and adjusting the image sensor group, the problem of unclear images caused by factors such as fog is reduced, thereby improving the user's actual experience.

[0049] See also Figure 4 As shown, it is a flow chart of an intelligent Internet of Things control method of the present invention, including: Step S1, the generation module iteratively optimizes the occlusion rate learning model according to the filtered data set until the loss value converges to a preset value; Step S2, testing the current occlusion rate learning model by filtering the data set; Step S3: When the test result meets the preset test result, the parameters of the occlusion rate learning model are saved.

[0050] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent Internet of Things operating system, characterized in that: The system is provided with a plurality of image acquisition devices arranged in groups, and a single image acquisition device includes: A shooting module, which is used to shoot real-time images of several points to be monitored and generate corresponding image data; A generating module, connected to the shooting module, for filtering the image data and generating corresponding filtering data and a corresponding occlusion rate learning model; A learning module, connected to the generating module, for preprocessing the filtered data, generating corresponding preprocessed data, selecting several index features of the preprocessed data, learning the preprocessed data according to the index features, and generating an occlusion rate distribution map of the real-time image; an alarm module, connected to the learning module, for receiving the occlusion rate distribution map, comparing the occlusion rate distribution map with an occlusion rate threshold, and generating a corresponding comparison result; an object-removing module, which is connected to the alarm module and is used to adjust the obstruction of the point to be monitored according to the comparison result; Wherein, the occlusion rate learning model is trained and generated on the filtered data according to the filtered data set; The filtering data set includes corresponding filtering data generated by the image data of the monitored point at different periods; The preprocessed data is to fit the corresponding image according to the filtered data, and generate a number of preprocessed data with a sampling rate of a standard sampling rate, and the learning module generates a corresponding indicator feature according to each preprocessed data; The standard sampling rate is a sampling rate that can be recognized by the learning module, and for a single sampling, the corresponding standard sampling rate is a single sampling rate.

2. The intelligent Internet of Things operating system according to claim 1, characterized in that: The shooting module includes a plurality of image sensor groups arranged at corresponding points to be monitored, and is used to respectively shoot real-time images of the plurality of points to be monitored and output them as image data.

3. The intelligent Internet of Things operating system according to claim 2, characterized in that: The generating module receives the image data, the filter filters the image data, and generates corresponding filtering data, Wherein, a pixel threshold is set in the filter to collect pixels of the image data and filter the image data exceeding the pixel threshold.

4. The intelligent Internet of Things operating system according to claim 3, characterized in that: The generation module stores the filtered data, and trains the filtered data according to the filtered data set to generate a corresponding occlusion rate learning model.

5. The intelligent Internet of Things operating system according to claim 4, characterized in that: The learning module receives the occlusion rate learning model, and preprocesses the filtered data through a preprocessor to generate corresponding preprocessed data. The learner selects several indicator features of the preprocessed data, learns the preprocessed data according to the occlusion rate learning model, and generates a corresponding occlusion rate distribution map. The indicator features include the proportion of obstructions in the real-time image, time and / or the location of the point to be monitored; The preprocessed data is learned by converting the preprocessed data into the occlusion rate distribution map, extracting the index feature in the occlusion rate distribution map, and annotating the occlusion rate distribution map based on the index feature.

6. The intelligent Internet of Things operating system according to claim 5, characterized in that: The preprocessor preprocesses the filtered data to generate the corresponding preprocessed data, and the preprocessing of the filtered data further includes: Based on the data collection characteristics of the filtered data set, data cleaning is performed to obtain cleaned data of the filtered data set, and the cleaned data is standardized; The standardization process is to segment the filtered data according to a standard sampling rate.

7. The intelligent Internet of Things operating system according to claim 6, characterized in that: The alarm module receives the occlusion rate distribution map, calculates the occlusion rate of each monitored point in the occlusion rate distribution map, and compares the occlusion rate with the occlusion rate threshold. When the occlusion rate is less than the occlusion rate threshold, the alarm module does not issue an adjustment instruction to the object-driving module.

8. The intelligent Internet of Things operating system according to claim 7, characterized in that: When the occlusion rate is greater than the occlusion rate threshold, the regulator in the alarm module issues an adjustment instruction to the object-driving module, and the object-driving module receives the adjustment instruction and adjusts the angle of the image sensor group.

9. The intelligent Internet of Things operating system according to claim 8, characterized in that: The image sensor group captures the real-time image of the point to be monitored again, and learns the occlusion rate of the real-time image until the occlusion rate is less than the occlusion rate threshold.

10. An intelligent Internet of Things control method according to any one of claims 1 to 9, characterized in that: include: The generating module iteratively optimizes the occlusion rate learning model according to the filtered data set until the loss value converges to a preset value; Testing the current occlusion rate learning model using the filtered data set; When the test result meets the preset test result, the parameters of the occlusion rate learning model are saved.

Images