Safety early warning methods, devices, equipment and media

By combining the first camera and the second camera, we can identify whether the window is equipped with a guardrail and open state, and process the maximum pixel value of the thermal image, solving the problem that the existing system cannot accurately identify children's dangerous situations, achieving fast and accurate Safety warning.

CN114187738BActive Publication Date: 2024-11-29ZHEJIANG UNIVIEW TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202010857843.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-08-24
Publication Date
2024-11-29
Estimated Expiration
2040-08-24

AI Technical Summary

Technical Problem

The existing safety warning system cannot accurately identify whether a child is in a dangerous situation, the false alarm rate is high and the warning time is insufficient, resulting in users being unable to deal with the dangerous situation in a timely manner.

Method used

The window image is collected through the first camera to determine whether the guardrail and the opening status are installed. If it is different, the thermal image image is collected using the second camera to obtain the maximum pixel value. If the preset value is exceeded, the child will be identified and the warning information will be sent. .

Benefits of technology

It realizes accurate identification of whether a child is in a dangerous situation without the need to be renovated, and sends warning information to users as soon as possible, improving the accuracy and timeliness of safety warnings.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114187738B_ABST
    Figure CN114187738B_ABST
Patent Text Reader

Abstract

The embodiment of the present invention discloses a safety warning method, device, equipment and medium. The method includes: according to the first type of window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first type window images; determining whether the at least two frames of new first type window images are the same; if not, controlling the second camera to capture the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image; if the maximum pixel value is greater than the preset pixel value, identifying the second type of window image, and sending a warning message when identifying the presence of a child in the second type of window image. The embodiment of the present invention improves the accuracy and timeliness of safety warnings.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of security warning technology, and in particular, to a security warning method, device, equipment and medium. Background Art

[0002] With the rapid development of urbanization, more and more people are moving into high-rise or small high-rise residential areas in cities. When children are alone at home or parents are at home but have no time to take care of their children, there is a possibility that children will fall from balcony windows or room windows.

[0003] In order to reduce the probability of children falling, infrared detectors are generally installed at different locations of balcony windows or room windows. When the infrared detectors sense that a living creature has entered the detection area, they will trigger the balcony windows or room windows to close and trigger the alarm device to sound an alarm.

[0004] However, the above method can usually only detect whether the object entering the monitoring area is a living thing, and cannot determine whether the living thing is an adult, a child, or a pet, resulting in a high false alarm rate. In addition, when a child climbs through the window and an alarm is sounded through the alarm device, the warning time may be too short, resulting in the user being unable to arrive at the scene in time to deal with the dangerous situation. Summary of the invention

[0005] The embodiments of the present invention provide a safety warning method, device, equipment and medium, which can accurately identify whether a child is in a dangerous situation without modifying the window, and send warning information to the user as soon as possible when the child is in a dangerous situation, so that the user can arrive at the scene in time to deal with the dangerous situation, thereby improving the accuracy and timeliness of the safety warning.

[0006] In a first aspect, an embodiment of the present invention provides a security warning method, which is applied to a monitoring device, and the method includes:

[0007] When it is determined that the window is not equipped with a guardrail and is in an open state according to the first-category window image of the window in the monitoring area captured by the first camera, controlling the first camera to capture at least two frames of new first-category window images;

[0008] Determine whether the at least two frames of new first-category window images are the same;

[0009] If not, controlling the second camera to collect the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image;

[0010] If the maximum pixel value is greater than the preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent.

[0011] In a second aspect, an embodiment of the present invention provides a safety warning device, which is configured in a monitoring device and includes:

[0012] A first control module is configured to control the first camera to capture at least two frames of new first-category window images when it is determined that the window is not equipped with a guardrail and is in an open state based on the first-category window images of the windows in the monitoring area captured by the first camera;

[0013] A first determination module, used to determine whether the at least two frames of new first-type window images are the same;

[0014] a processing module, configured to, if not, control the second camera to collect a second type of window image of the window, and process the second type of window image to obtain a maximum pixel value in the processed second type of window image;

[0015] The second control module is used to identify the second type of window image if the maximum pixel value is greater than a preset pixel value, and send a warning message when it is identified that a child exists in the second type of window image.

[0016] In a third aspect, an embodiment of the present invention further provides an electronic device, including:

[0017] one or more processors;

[0018] a storage device for storing one or more programs,

[0019] When the one or more programs are executed by the one or more processors, the one or more processors implement any security warning method described in any one of the embodiments of the present invention.

[0020] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the security warning methods described in the embodiments of the present invention.

[0021] The technical solution disclosed in the embodiment of the present invention has the following beneficial effects:

[0022] According to the first-class window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, the first camera is controlled to capture at least two frames of new first-class window images, and when it is determined that the at least two frames of new first-class window images are different, the second camera is controlled to capture the second-class window image of the window, and the second-class window image is processed to obtain the maximum pixel value in the processed second-class window image. If the maximum pixel value is greater than the preset pixel value, the second-class window image is identified, and when a child is identified in the second-class window image, an early warning message is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the window, and when in a dangerous situation, an early warning message is sent to the user as soon as possible, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flowchart of a safety early warning method provided by an embodiment of the present invention;

[0024] Figure 1 (a) is a schematic diagram of determining an adjusted focal length of a first camera provided by an embodiment of the present invention;

[0025] Figure 1 (b) is a schematic diagram of a second type of window image provided by an embodiment of the present invention;

[0026] Figure 2 is a flow chart of another safety warning method provided by an embodiment of the present invention;

[0027] Figure 3 is a flowchart of another safety warning method provided by an embodiment of the present invention;

[0028] Figure 4 It is a flowchart of another safety warning method provided by an embodiment of the present invention;

[0029] Figure 5 It is a structural schematic diagram of a safety warning device provided by an embodiment of the present invention;

[0030] Figure 6 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention, rather than to limit the embodiments of the present invention. It is also necessary to explain that, for ease of description, only parts related to the embodiments of the present invention are shown in the accompanying drawings, rather than all structures.

[0032] The security warning method, device, equipment and medium of the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0033] Figure 1 It is a flowchart of a safety warning method provided by an embodiment of the present invention. The embodiment of the present invention is applicable to the scenario of sending a safety warning to the user when a child appears in the window area. The method can be executed by the safety warning device provided by the embodiment of the present invention, and can also be implemented in hardware and / or software. The device can be integrated into an electronic device, wherein the electronic device is preferably a monitoring device. In the embodiment of the present invention, the monitoring device can be a camera with a data processing function. Among them, the camera includes a first camera and a second camera. As Figure 1 As shown, the method specifically includes the following:

[0034] S101, based on the first type of window images of the windows in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first type of window images.

[0035] In the embodiment of the present invention, the first camera refers to an ordinary camera; correspondingly, the first type of window image refers to an ordinary window image. The first camera is preferably an ordinary network camera.

[0036] The monitoring area refers to any side area of ​​a residential building. For example, monitoring the south side area of ​​Building 10, or monitoring the north side area of ​​Building 11, etc., which are not specifically limited here. The window refers to a balcony window or a room window, etc.

[0037] Before executing S101, the first camera installed on the outside of the roof can be started, and the first camera can be controlled to sequentially scan the windows of each floor in the monitoring area. The sequential scanning can be from top to bottom or from bottom to top. When scanning a window on any floor in the monitoring area, the current pitch angle of the first camera is obtained, and the floor information of the window is determined according to the current pitch angle of the first camera.

[0038] Specifically, the current pitch angle of the first camera can be obtained by determining the current pitch angle of the first camera based on the rotation distance of the first motor responsible for horizontal rotation of the camera and the second motor responsible for vertical rotation of the camera in the gimbal carrying the first camera.

[0039] After obtaining the current pitch angle of the first camera, the preset mapping relationship can be queried according to the current pitch angle of the first camera to determine the floor information of the window. The preset mapping relationship is to control the first camera to scan the windows on each floor in the monitoring area before controlling the monitoring device to execute the security warning solution to obtain the pitch angle of the first camera at the window on each floor. According to the obtained pitch angle of the first camera at the window on each floor, the floor information of the window corresponding to each pitch angle is calculated according to the preset calculation formula. Then, a mapping relationship is established according to each pitch angle and the floor information.

[0040] The floor information of the window corresponding to each pitch angle can be calculated according to the following formula (1).

[0041]

[0042] Among them, floor i represents the floor information corresponding to the i-th pitch angle, floormax represents the maximum number of floors, and L represents the vertical distance between the first camera and the vertical wall of the residential building where it is located. represents the i-th pitch angle, h' represents the height deviation of the first camera relative to the highest floor, Represents the average adjacent floor height.

[0043] After calculating the floor information of the window corresponding to each pitch angle, the mapping relationship between different pitch angles and floor information can be established. Furthermore, the mapping relationship can be stored in the basic information of the residents, providing a basis for determining the floor information according to the pitch angle of the first camera in the future, so that there is no need to calculate the floor information according to the pitch angle of the first camera every time, which improves the speed of determining the floor information and reduces unnecessary data processing. Among them, the basic information of the residents is an information table of the basic information of each household in each residential building pre-counted by the property staff, and the basic information of the residents can usually be stored in an electronic device. The basic information of the residents may include but is not limited to: the floor information of the residents, the number of permanent residents and the age of the population, and the rules of entry and exit of the residents of each household.

[0044] Furthermore, when scanning windows on any floor, this embodiment can also adjust the focal length of the first camera so that the window is displayed at the center position of the first camera to ensure that the imaging quality of the first type of windows captured by the first camera is optimal.

[0045] Specifically, when adjusting the focal length of the first camera, Figure 1 (a) and formula (2) to determine the focal length of the first camera;

[0046]

[0047] Where, l represents the distance from the first camera to the window, l' represents the imaging distance of the window, f represents the focal length of the first camera, and H orignal Indicates the original height of the window, H image represents the imaging height of the window, L represents the vertical distance between the first camera and the vertical wall of the residential building, Indicates the pitch angle of the first camera.

[0048] After adjusting the focal length of the first camera, the embodiment of the present invention can also control the adjusted first camera to collect the first type of window images of the windows in the monitoring area, and determine whether the window is installed with a guardrail based on the collected first type of window images.

[0049] Among them, the first type of window image can be identified and processed through the image processing program of the first camera to determine whether the window is equipped with a guardrail; and / or the first type of window image captured by the first camera is sent to the monitoring display screen, so that the property staff can determine whether the window is equipped with a guardrail based on the first type of window image displayed on the display screen.

[0050] In an embodiment of the present invention, when the property staff determines whether the window is installed with a guardrail based on the first-category window image displayed on the display screen, the property staff may query the basic information of the resident based on the building and floor information carried on the first-category window image displayed on the display screen to determine whether the window is installed with a guardrail; or the property staff may observe the first-category window image displayed on the display screen with the naked eye to determine whether the window is installed with a guardrail, etc., which is not specifically limited here.

[0051] If it is determined that the window is equipped with a guardrail based on the first-class window image captured by the first camera, the first camera is controlled to end the image acquisition of the window and continue to scan the next window in the monitoring area; if it is determined that the window is not equipped with a guardrail based on the first-class window image captured by the first camera, this embodiment can further determine whether the window is in an open state based on the first-class window image. Optionally, a pre-trained first recognition model can be used to determine whether the window is in an open state based on the first-class window image captured by the first camera; or a preset algorithm can be used to determine whether the window is in an open state based on the first-class window image captured by the first camera. Among them, if the window is identified as being in a closed state, it means that the current window area is safe, and the first camera is controlled to end the image acquisition of the window and continue to scan the next window in the monitoring area; if the window is identified as being in an open state, it means that there is danger in the current window area, and the first camera is controlled to continue to monitor the window to collect at least two frames of new first-class window images of the window, laying a foundation for the subsequent determination of whether there is a child in the window area.

[0052] In the embodiment of the present invention, the first recognition model refers to a model used to recognize whether the window corresponding to the first type of window image is in an open state or a closed state. Specifically, the first recognition model can be trained by the following steps:

[0053] S11. Obtain a first training sample set.

[0054] The first training sample set includes window images in a closed state and window images in an open state.

[0055] Optionally, the first training sample set may be obtained from a massive image library on the Internet, or may be obtained based on the first camera capturing windows in an open state or windows in a closed state, which is not specifically limited here.

[0056] S12. Training a first initial recognition model according to the window images in the closed state and the window images in the open state in the first training sample set to generate a first recognition model.

[0057] In the embodiment of the present invention, the first initial recognition model may be a support vector machine (SVM) or other machine learning algorithms, etc., which are not specifically limited here.

[0058] Optionally, the window image in the closed state and the window image in the open state can be input as input values ​​into the first initial recognition model, so as to extract the Histogram of Oriented Gradient (HOG) features from the window image in the closed state and the window image in the open state respectively through the first initial recognition model, and then continuously train according to the extracted HOG features until the recognition result of the trained first initial recognition model reaches the recognition accuracy. Among them, the recognition accuracy can be set according to the recognition requirements, for example, set to 90% or 95%. In the embodiment of the present invention, the recognition accuracy is set to 90% by default.

[0059] That is to say, when the accuracy of the recognition result of the first initial recognition model after training reaches 90%, the trained first initial recognition model can be used as the first recognition model of the embodiment of the present invention.

[0060] In order to ensure the performance of the first recognition model, the embodiment of the present invention trains the first initial recognition model. When obtaining the first recognition model, it is necessary not only to ensure that the accuracy of the first recognition model reaches the preset recognition accuracy, but also to ensure that the first recognition model recognizes and processes the first type of window image as quickly as possible. Optionally, obtaining the first recognition model may include the following multiple situations:

[0061] In case one, corresponding thresholds can be set in advance for the accuracy and speed of the first recognition model. Then, after training the first initial recognition model, if the trained first initial model is used to recognize the first type of window image, and the accuracy of the recognition result reaches the recognition accuracy threshold, and the recognition speed also reaches the speed threshold, then the trained first initial model can be determined as the first recognition model.

[0062] Preferably, the embodiment of the present invention can determine the first initial recognition model after training that can achieve a preset recognition accuracy for the recognition result of the first type of window image and has the fastest recognition speed for the first type of window image as the first recognition model.

[0063] Generally, the requirements for the recognition accuracy and recognition speed of the first type of window image may be different. Therefore, in this embodiment, different weights may be assigned to the accuracy and speed of the first recognition model according to application requirements to obtain the first recognition model under different requirements, as shown in the following cases 2 and 3:

[0064] Case 2: When the accuracy requirement is high, the embodiment of the present invention can configure a weight for the recognition accuracy that is greater than the recognition speed. For example, an accuracy threshold can be configured for the recognition accuracy. Then, after training the first initial recognition model, if the trained first initial model is used to recognize the first type of window image and the accuracy of the recognition result reaches the accuracy threshold, the trained first initial recognition model can be determined as the first recognition model.

[0065] Case three, when the recognition speed requirement is high, the embodiment of the present invention can configure a weight for the recognition speed that is greater than the recognition accuracy. For example, a speed threshold can be configured for the recognition speed. Then, after training the first initial recognition model, if the trained first initial model is used to recognize the first type of window image and the recognition speed of the recognition result reaches the speed threshold, the trained first initial recognition model can be determined as the first recognition model.

[0066] It should be noted that the above-mentioned several cases of obtaining the first recognition model are only used as exemplary descriptions of the embodiments of the present invention and are not intended to be specific limitations of the present invention.

[0067] S102, determining whether the at least two frames of new first-category window images are the same.

[0068] Optionally, it can be determined whether at least two frames of new first-category window images captured by the first camera are the same based on an inter-frame difference method; or other methods can be used to determine whether at least two frames of new first-category window images captured by the first camera are equal, which is not specifically limited in the present invention.

[0069] Specifically, when it is determined that at least two frames of the new first-category window images are the same, it means that the scene corresponding to the window currently scanned by the first camera is a static scene, that is, there is no moving object. When it is determined that at least two frames of the new first-category window images are different, it means that the scene corresponding to the window currently scanned by the first camera is an active scene, that is, there is a moving object. Among them, when it is determined that the scene corresponding to the window currently scanned by the first camera is a static scene, it means that the current window area is safe, and the first camera is controlled to end the image acquisition of the window and continue to scan the next window in the monitoring area.

[0070] S103: If not, control the second camera to collect the second type of window image of the window, and process the second type of window image to obtain the maximum pixel value in the processed second type of window image.

[0071] The second camera is a thermal imaging camera, and correspondingly, the second type of window image is a thermal imaging window image. The second camera is preferably a thermal imaging network camera.

[0072] Optionally, when it is determined that at least two frames of new first-category window images captured by the first camera are different, it means that the scene corresponding to the window currently scanned by the first camera is an activity scene. Since the activity scene may include human activities, pet activities or other non-biological activities, in order to determine whether the activity scene belongs to a biological activity scene or a non-biological activity scene. The embodiment of the present invention can control the second camera to capture the second-category window image of the window, and process the second-category window image captured by the second camera to obtain the maximum pixel value in the processed second-category window image. Then, based on the maximum pixel value and the preset pixel value, it is determined whether the scene corresponding to the window is a biological activity scene or a non-biological activity scene. Among them, the preset pixel value refers to a judgment standard value used to determine whether the activity scene is a biological activity scene or a non-biological activity scene. It can be set specifically according to actual needs, and the embodiment of the present invention does not impose specific restrictions on this. The second-category window image captured by the specific second camera can be as follows Figure 1 (b) as shown.

[0073] As an optional implementation, in an embodiment of the present invention, the second type of window image is processed to obtain the maximum pixel value in the processed second type of window image, which can be grayscale processing of the second type of window image to obtain a grayscale image, and the maximum grayscale value is obtained from the grayscale image, and the maximum grayscale value is determined as the maximum pixel value. The second type of window image is a color image with three color channels. The three color channels are: red channel (R), green channel (G) and blue channel (B).

[0074] The second type of window image is gray-scaled to obtain a gray-scale image. A component method may be used to take the brightness of the three components in the second type of window image as the gray-scale values ​​of the three gray-scale images, and a gray-scale image of any one component may be selected as the gray-scale image of the second type of window image as needed; or a maximum value method may be used to take the maximum value of the brightness of the three components in the second type of window image as the gray-scale value of the gray-scale image, and a gray-scale image of the second type of window image may be obtained according to the gray-scale value; or an average value method or a weighted average method may be used to obtain a gray-scale image of the second type of window image, and the embodiments of the present invention do not specifically limit this.

[0075] It should be noted that the above-mentioned biological activity scene refers to the active objects being living objects such as people or pets. Non-biological activity scene refers to the active objects being objects without life characteristics, such as swaying plants, fluttering curtains, or fluttering clothes on the balcony.

[0076] S104: If the maximum pixel value is greater than a preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent.

[0077] When a child is identified in the second type of window image, the warning information is sent to users such as residents and property personnel. In the embodiment of the present invention, the resident may be an owner or a tenant, etc., which is not specifically limited here.

[0078] When it is determined that the maximum pixel value is greater than the preset pixel value, it means that the activity scene corresponding to the window is a biological activity scene. Because the organisms in the biological activity scene may include humans and pets, in order to determine whether there are people in the activity scene and whether the people are children. The embodiment of the present invention can use other methods such as a preset algorithm to identify the second type of window image captured by the second camera. Among them, if it is identified that there are children in the second type of window image, it means that there are children in the window area, and because the window is not equipped with a guardrail and is in an open state, it is determined that the child is in a dangerous area, and it is necessary to send an early warning message to the user so that the user can take timely measures to reduce the occurrence of safety accidents. If it is identified that there are no children in the second type of window image, it means that the window area is safe, and the next window can continue to be scanned and monitored.

[0079] When sending warning information to the user, the contact information of the resident of the window can be obtained from the stored basic information of the resident, and the warning information can be sent to the resident according to the contact information of the resident; the monitoring image of the window is sent to the monitoring display screen used by the property staff, so that the monitoring display screen displays the monitoring image. The warning information sent to the resident can be information such as "your child is playing near the window and there is danger. Please take the child to a safe area as soon as possible."

[0080] In the embodiment of the present invention, warning information can be sent to the residents at the same time, and the monitoring image of the window can be sent to the monitoring display screen used by the property staff; or, the monitoring image of the window can be sent to the monitoring display screen used by the property staff first, and when the confirmation information of the property is not obtained within the specified time, the warning information can be sent to the residents, etc. The residents can reply to the reply information that they are at home or not at home according to the warning information, so that the monitoring equipment can perform the warning operation according to the reply information that the residents are not at home, and the property staff can determine whether there is an adult in the resident's home based on the basic information of the resident according to the monitoring image time displayed on the monitoring display screen, and when it is determined that there is no adult, timely collect remedial measures to reduce the probability of children falling accidents.

[0081] The technical solution provided by the embodiment of the present invention is to control the first camera to collect at least two frames of new first-class window images when it is determined that the window is not equipped with a guardrail and is in an open state according to the first-class window images of the windows in the monitoring area collected by the first camera, and to control the second camera to collect the second-class window images of the windows when it is determined that the at least two frames of new first-class window images are different, and to process the second-class window images to obtain the maximum pixel value in the processed second-class window images. If the maximum pixel value is greater than the preset pixel value, the second-class window images are identified, and when the second-class window images are identified to contain children, warning information is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the windows, and to send warning information to the user as soon as possible when the child is in a dangerous situation, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning.

[0082] Figure 2 FIG. 1 is a flow chart of another safety warning method provided by an embodiment of the present invention. Based on the above embodiment, this embodiment optimizes “identifying the second type of window image and determining whether there is a child in the second type of window image”. Figure 2 As shown, the method specifically includes:

[0083] S201, based on the first-category window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first-category window images.

[0084] S202: Determine whether the at least two frames of new first-category window images are the same.

[0085] S203: If not, control the second camera to collect the second type of window image of the window, and process the second type of window image to obtain the maximum pixel value in the processed second type of window image.

[0086] S204: If the maximum pixel value is greater than a preset pixel value, identifying the second type of window image in a preset manner to obtain at least one portrait in the second type of window image.

[0087] Optionally, the second type of window image can be identified according to a preset second recognition model to obtain at least one portrait in the second type of window image, or other methods can be used to identify the second type of window image, which are not specifically limited here. When at least one portrait in the second type of window image is obtained, the area where each portrait is located is also framed to lay the foundation for subsequently determining the portrait area of ​​each portrait.

[0088] In the embodiment of the present invention, the second recognition model refers to a model for identifying whether there is a child in the second type of window image. Specifically, the second recognition model can be trained by the following steps:

[0089] S21. Obtain a second training sample set.

[0090] The second training sample set includes window images with human portraits and window images without human portraits.

[0091] Optionally, the second training sample set may be obtained from a massive image library on the Internet, or may be obtained by capturing window images with portraits and window images without portraits by a second camera, which is not specifically limited here.

[0092] S22. Train the second initial recognition model according to the window images with human portraits and the window images without human portraits in the second training sample set to generate a second recognition model.

[0093] In the embodiment of the present invention, the second initial recognition model may be SVM or other machine learning algorithms, etc., which are not specifically limited here.

[0094] Optionally, the window image with a portrait and the window image without a portrait can be input as input values ​​into the second initial recognition model, so as to extract the Histogram of Oriented Gradient (HOG) features from the window image with a portrait and the window image without a portrait respectively through the second initial recognition model, and then continuously train according to the extracted HOG features until the recognition result of the trained second initial recognition model reaches the recognition accuracy. Among them, the recognition accuracy can be set according to the recognition requirements, for example, set to 90% or 95%. In the embodiment of the present invention, the recognition accuracy is set to 90% by default.

[0095] That is to say, when the accuracy of the recognition result of the trained second initial recognition model reaches 90%, the trained second initial recognition model can be used as the second recognition model of the embodiment of the present invention.

[0096] In order to ensure the performance of the second recognition model, the embodiment of the present invention trains the second initial recognition model. When obtaining the second recognition model, it is necessary not only to ensure that the accuracy of the second recognition model reaches the preset recognition accuracy, but also to ensure that the second recognition model recognizes and processes the second type of window image as quickly as possible. Optionally, obtaining the second recognition model may include the following multiple situations:

[0097] In case one, corresponding thresholds can be set in advance for the accuracy and speed of the second recognition model. Then, after the second initial recognition model is trained, if the trained second initial model is used to recognize the second type of window image, and the accuracy of the recognition result reaches the recognition accuracy threshold, and the recognition speed also reaches the speed threshold, then the trained second initial model can be determined as the second recognition model.

[0098] Preferably, the embodiment of the present invention can determine the second initial recognition model after training that can achieve a preset recognition accuracy for the recognition result of the second type of window images and has the fastest recognition speed for the second type of window images as the second recognition model.

[0099] Generally, the requirements for the recognition accuracy and recognition speed of the second type of window images may be different. Therefore, in this embodiment, different weights may be assigned to the accuracy and speed of the second recognition model according to application requirements to obtain the second recognition model under different requirements, as shown in the following cases 2 and 3:

[0100] Case 2: When the accuracy requirement is high, the embodiment of the present invention can configure a weight for the recognition accuracy that is greater than the recognition speed. For example, an accuracy threshold can be configured for the recognition accuracy. Then, after training the second initial recognition model, if the trained second initial model is used to recognize the second type of window image and the accuracy of the recognition result reaches the accuracy threshold, the trained second initial recognition model can be determined as the second recognition model.

[0101] Case three, when the recognition speed requirement is high, the embodiment of the present invention can configure a weight for the recognition speed that is greater than the recognition accuracy. For example, a speed threshold can be configured for the recognition speed. Then, after training the second initial recognition model, if the trained second initial model is used to recognize the second type of window image and the recognition speed of the recognition result reaches the speed threshold, the trained second initial recognition model can be determined as the second recognition model.

[0102] It should be noted that the above-mentioned several cases of obtaining the second recognition model are only used as exemplary descriptions of the embodiments of the present invention and are not intended to be specific limitations of the present invention.

[0103] S205: Determine whether there is a child in the second-type window image according to the portrait area of ​​each portrait.

[0104] Optionally, after identifying at least one portrait included in the second-category window image, the second-category window image with the portrait in the frame can be subtracted from a preset pixel value to obtain a second-category window image that only includes portraits and animals. Since after the preset pixel value is subtracted from the second-category window image with the portrait in the frame, the framed portrait boundary line may have holes or discontinuities, resulting in errors in the subsequent determination of the portrait area of ​​the portrait. The embodiment of the present invention can perform a completion process on the portrait area where the portrait boundary line has holes or discontinuities according to the pixel values ​​of the adjacent areas to the holes or discontinuities, so that the framed portrait boundary line is a continuous and closed whole, thereby providing conditions for determining the portrait area of ​​each portrait.

[0105] In the embodiment of the present invention, after obtaining the continuous and closed portrait regions, the portrait area of ​​each portrait region can be calculated, and then whether there is a child in the second type of window image can be determined based on the portrait area. The embodiment of the present invention can use the existing algorithm for calculating the area of ​​the region to calculate the portrait area of ​​the region occupied by each portrait, which will not be described in detail here.

[0106] In this embodiment, the portrait area of ​​each portrait can be compared with the portrait area threshold to determine whether the portrait area of ​​each portrait is less than the portrait area threshold, and whether there is a child in the second type of window image according to the comparison result. Specifically, if the portrait area of ​​any portrait is less than the portrait area threshold, it is determined that there is a child in the second type of window image; if the portrait area of ​​each portrait is greater than or equal to the portrait area threshold, it is determined that there is no child in the second type of window image.

[0107] The portrait area threshold refers to the adult portrait area threshold. In the embodiment of the present invention, the adult portrait area threshold can be set according to different adult postures, which is not specifically limited here, such as standing posture, squatting posture or walking posture.

[0108] S206: When it is identified that a child exists in the second type of window image, a warning message is sent.

[0109] The technical solution provided by the embodiment of the present invention is, according to the first-class window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, the first camera is controlled to capture at least two frames of new first-class window images, and when it is determined that the at least two frames of new first-class window images are different, the second camera is controlled to capture the second-class window image of the window, and the second-class window image is processed to obtain the maximum pixel value in the processed second-class window image. If the maximum pixel value is greater than the preset pixel value, the second-class window image is identified in a preset manner to obtain at least one portrait in the second-class window image, and then according to the portrait area of ​​each portrait, it is determined whether there is a child in the second-class window image, and when the second-class window image is identified to exist a child, an early warning message is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the window, and when in a dangerous situation, an early warning message is sent to the user as soon as possible, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning. In addition, by obtaining at least one portrait based on the recognition of the second type of window image, it is determined whether there is a child in the second type of window image according to the area of ​​the portrait, thereby meeting the requirements for accurate recognition of children and adults in different forms, thereby improving the accuracy of child recognition.

[0110] Figure 3 FIG. 1 is a flow chart of another safety warning method provided by an embodiment of the present invention. This embodiment is further optimized on the basis of the above embodiment. Figure 3 As shown, the method specifically includes the following:

[0111] S301, based on the first-category window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first-category window images.

[0112] S302: Determine whether the at least two frames of new first-category window images are the same.

[0113] S303: If not, control the second camera to collect the second type of window image of the window, and process the second type of window image to obtain the maximum pixel value in the processed second type of window image.

[0114] S304: If the maximum pixel value is greater than the preset pixel value, the second type of window image is identified, and the safety protection level and the danger level are determined according to the identification result.

[0115] Optionally, based on the identification results, the security protection level and the danger level are determined, which may include:

[0116] The greater the possibility that an adult exists in the second type of window image, the higher the safety protection level and the lower the danger level; the more children exist in the second type of window image, the lower the safety protection level and the higher the danger level. Specifically, it can be set as follows: if it is identified that there are adults in the second type of window image, the safety protection level is determined to be level three protection and the danger level is level zero danger; if it is identified that there may be adults in the second type of window image, the safety protection level is determined to be level two protection and the danger level is level one danger; if it is identified that there are no adults in the second type of window image and the number of children is one, the safety protection level is determined to be level one protection and the danger level is level two danger; if it is identified that there are no adults in the second type of window image and the number of children is at least two, the safety protection level is determined to be level one protection and the danger level is level three danger.

[0117] It should be noted that, in the embodiment of the present invention, when the first type of window image captured by the first camera determines that the window is in a closed state, the security protection level can be determined to be the third level of protection level.

[0118] S305: When the security protection level is less than or equal to the danger level, a warning message is sent.

[0119] Specifically, when the security protection level is level one protection and the danger level is level one danger, a warning message is sent to the user; or, when the security protection level is level one protection and the danger level is level two danger, a warning message is sent to the user, and so on.

[0120] That is to say, in this embodiment, if the number of children in the second type of window image is identified as at least one and there may be adults, a warning message is sent to the user; or, if the number of children in the second type of window image is identified as at least one and there are no adults, a warning message is sent to the user.

[0121] The technical solution provided by the embodiment of the present invention is to control the first camera to collect at least two frames of new first-class window images when it is determined that the window is not equipped with a guardrail and is in an open state according to the first-class window images of the windows in the monitoring area collected by the first camera, and to control the second camera to collect the second-class window images of the windows when it is determined that the at least two frames of new first-class window images are different, and to process the second-class window images to obtain the maximum pixel value in the processed second-class window images. If the maximum pixel value is greater than the preset pixel value, the second-class window images are identified, and when the second-class window images are identified to contain children, warning information is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the window, and to send a warning information to the user in the first time when the child is in a dangerous situation, so that the user can rush to the scene in time to deal with the dangerous situation, thereby improving the accuracy and timeliness of the safety warning.

[0122] Figure 4This is a flow chart of another safety warning method provided by an embodiment of the present invention. This embodiment is further optimized on the basis of the above embodiment. Specifically, after sending the warning information to the user, the relevant content of "updating the warning record table of the residents of the window" is added. Figure 4 As shown, the method specifically includes:

[0123] S401, based on the first-category window image of the window in the monitoring area captured by the first camera, when it is determined that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first-category window images.

[0124] S402: Determine whether the at least two frames of new first-category window images are the same.

[0125] S403: If not, control the second camera to collect the second type of window image of the window, and process the second type of window image to obtain the maximum pixel value in the processed second type of window image.

[0126] S404: If the maximum pixel value is greater than a preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent.

[0127] S405, updating the resident warning record list of the window, and sending the updated warning record list to the resident.

[0128] Among them, by sending the updated early warning record form to the residents, the residents can strengthen the care of their children and the danger prevention education according to the updated early warning record form.

[0129] Optionally, after sending the warning information to the user, the embodiment of the present invention can also determine the resident warning record table of the window, and update the resident warning record table based on the warning information sent this time. Then, the updated resident warning record table is sent to the resident, so that the resident can understand the time period when their children are likely to appear in the window area, and the dangerous actions they do in the window area, etc., to provide children with risk prevention education and strengthen the care of children, thereby reducing the occurrence of child falling accidents.

[0130] The technical solution provided by the embodiment of the present invention provides a method for accurately identifying whether a child is in a dangerous situation without modifying the window, and sending warning information to the user immediately when the child is in a dangerous situation, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning. In addition, the updated resident warning record form is sent to the residents, so that the residents can strengthen the child care and danger prevention education, and provide conditions for avoiding child falling accidents.

[0131] Figure 51 is a schematic diagram of the structure of a safety warning device provided by an embodiment of the present invention. The safety warning device is configured in a monitoring device. Figure 5 As shown, the safety warning device 500 according to the embodiment of the present invention includes: a first control module 510 , a first determination module 520 , a processing module 530 and a second control module 540 .

[0132] The first control module 510 is used to control the first camera to capture at least two frames of new first-category window images when it is determined that the window is not equipped with a guardrail and is in an open state according to the first-category window images of the windows in the monitoring area captured by the first camera;

[0133] A first determination module 520, configured to determine whether the at least two frames of new first-type window images are the same;

[0134] A processing module 530 is used for, if not, controlling the second camera to collect the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image;

[0135] The second control module 540 is configured to identify the second type of window image if the maximum pixel value is greater than a preset pixel value, and send a warning message when a child is identified in the second type of window image.

[0136] As an optional implementation of the embodiment of the present invention, the safety warning device 500 further includes: a third control module and a fourth control module;

[0137] The third control module is used to control the first camera to scan the windows of each floor in the monitoring area in sequence;

[0138] The fourth control module is used to determine the floor information of the window according to the current pitch angle of the first camera when scanning a window on any floor, and adjust the focal length of the first camera so that the window is displayed at the center position of the first camera.

[0139] As an optional implementation of the embodiment of the present invention, the fourth control module is specifically configured to:

[0140] According to the current pitch angle of the first camera, a preset mapping relationship is queried to determine the floor information of the window.

[0141] As an optional implementation of the embodiment of the present invention, the security warning device 500 further includes: a second determination module and a third determination module;

[0142] The second determination module is used to control the first camera to collect a first type of window image of a window in the monitoring area, and determine whether the window is installed with a guardrail according to the first type of window image;

[0143] The third determination module is used to determine whether the window is in an open state according to the first type of window image if no.

[0144] As an optional implementation of the embodiment of the present invention, the processing module 530 is specifically configured to:

[0145] Performing grayscale processing on the second type of window image to obtain a grayscale image;

[0146] A maximum grayscale value is obtained from the grayscale image, and the maximum grayscale value is determined as a maximum pixel value.

[0147] As an optional implementation of the embodiment of the present invention, the second control module 540 includes: an identification subunit and a determination subunit;

[0148] The recognition subunit is used to recognize the second type of window image in a preset manner to obtain at least one human portrait in the second type of window image;

[0149] The determination subunit is used to determine whether there is a child in the second type of window image according to the portrait area of ​​each portrait.

[0150] As an optional implementation manner of the embodiment of the present invention, the determination subunit is specifically used for:

[0151] Determine whether the portrait area of ​​each portrait is smaller than the portrait area threshold;

[0152] If the portrait area of ​​any portrait is smaller than the portrait area threshold, it is determined that there is a child in the second type of window image;

[0153] If the portrait area of ​​each portrait is greater than or equal to the portrait area threshold, it is determined that there is no child in the second type of window image.

[0154] As an optional implementation of the embodiment of the present invention, the security warning device 500 further includes: a fourth determination module;

[0155] The fourth determination module is used to determine the safety protection level and the danger level according to the recognition result.

[0156] As an optional implementation manner of the embodiment of the present invention, the fourth determining module is specifically configured to:

[0157] If it is identified that there is an adult in the second type of window image, the safety protection level is determined to be level three protection and the danger level is level zero danger;

[0158] If it is identified that there may be an adult in the second type of window image, the safety protection level is determined to be level 2 protection and the danger level is level 1 danger;

[0159] If it is determined that there is no adult in the second type of window image and the number of children is one, the safety protection level is determined to be level one protection and the danger level is level two danger;

[0160] If it is identified that there is no adult in the second type of window image and the number of children is at least two, the safety protection level is determined to be level one protection and the danger level is level three danger.

[0161] As an optional implementation of the embodiment of the present invention, the second control module 540 is specifically configured to:

[0162] If it is identified that there is at least one child in the second type of window image, and there may be an adult, sending a warning message to the user;

[0163] Alternatively, if it is identified that there is at least one child in the second type of window image and no adult exists, a warning message is sent to the user.

[0164] As an optional implementation of the embodiment of the present invention, the second control module 540 is specifically configured to:

[0165] Sending an early warning message to the resident according to the resident contact information of the window obtained from the resident basic information;

[0166] The monitoring image of the window is sent to a monitoring display screen used by property personnel so that the monitoring display screen displays the monitoring image.

[0167] As an optional implementation of the embodiment of the present invention, the security warning device 500 further includes: an update module;

[0168] The updating module is used to update the resident warning record list of the window and send the updated warning record list to the resident.

[0169] As an optional implementation of the embodiment of the present invention, the first camera is a common camera; the second camera is a thermal imaging camera;

[0170] Correspondingly, the first type of window images are ordinary window images, and the second type of window images are thermal imaging window images.

[0171] It should be noted that the above explanation of the embodiment of the safety warning method is also applicable to the safety warning device of this embodiment, and the implementation principle is similar, which will not be repeated here.

[0172] The technical solution provided by the embodiment of the present invention is to control the first camera to collect at least two frames of new first-class window images when it is determined that the window is not equipped with a guardrail and is in an open state according to the first-class window images of the windows in the monitoring area collected by the first camera, and to control the second camera to collect the second-class window images of the windows when it is determined that the at least two frames of new first-class window images are different, and to process the second-class window images to obtain the maximum pixel value in the processed second-class window images. If the maximum pixel value is greater than the preset pixel value, the second-class window images are identified, and when the second-class window images are identified to contain children, warning information is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the windows, and to send warning information to the user as soon as possible when the child is in a dangerous situation, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning.

[0173] In order to achieve the above objective, an embodiment of the present invention further provides an electronic device.

[0174] Figure 6 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 6 A block diagram of an exemplary electronic device 600 suitable for implementing embodiments of the present invention is shown. Figure 6 The electronic device 600 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0175] like Figure 6 As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include, but are not limited to: one or more processors or processing units 610, a system memory 620, and a bus 630 connecting different system components (including the system memory 620 and the processing unit 610).

[0176] Bus 630 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0177] The electronic device 600 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 600, including volatile and non-volatile media, removable and non-removable media.

[0178] The system memory 620 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 621 and / or cache memory 622. The electronic device 600 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 623 may be used to read and write non-removable, non-volatile magnetic media ( Figure 6 not shown, usually called a "hard drive"). Although Figure 6 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical medium) may be provided. In these cases, each drive may be connected to bus 630 via one or more data medium interfaces. Memory 620 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of various embodiments of the present invention.

[0179] A program / utility 624 having a set (at least one) of program modules 625 may be stored, for example, in the memory 620, such program modules 625 including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 625 generally perform the functions and / or methods of the embodiments described herein.

[0180] The electronic device 600 may also communicate with one or more external devices 640 (e.g., keyboard, pointing device, display 641, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., network card, modem, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public network, such as the Internet) via a network adapter 660. As shown, the network adapter 660 communicates with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0181] The processing unit 610 executes various functional applications and data processing by running the programs stored in the system memory 620, for example, implementing the security warning method provided in the embodiment of the present invention, including:

[0182] When it is determined that the window is not equipped with a guardrail and is in an open state according to the first-category window image of the window in the monitoring area captured by the first camera, controlling the first camera to capture at least two frames of new first-category window images;

[0183] Determine whether the at least two frames of new first-category window images are the same;

[0184] If not, controlling the second camera to collect the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image;

[0185] If the maximum pixel value is greater than the preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent.

[0186] It should be noted that the above explanation of the embodiment of the safety warning method is also applicable to the electronic device of this embodiment, and its implementation principle is similar, which will not be repeated here.

[0187] The technical solution provided by the embodiment of the present invention is to control the first camera to collect at least two frames of new first-class window images when it is determined that the window is not equipped with a guardrail and is in an open state according to the first-class window images of the windows in the monitoring area collected by the first camera, and to control the second camera to collect the second-class window images of the windows when it is determined that the at least two frames of new first-class window images are different, and to process the second-class window images to obtain the maximum pixel value in the processed second-class window images. If the maximum pixel value is greater than the preset pixel value, the second-class window images are identified, and when the second-class window images are identified to contain children, warning information is sent. Thus, a method is provided to accurately identify whether a child is in a dangerous situation without modifying the windows, and to send warning information to the user as soon as possible when the child is in a dangerous situation, so that the user can rush to the scene to deal with the dangerous situation in time, thereby improving the accuracy and timeliness of the safety warning.

[0188] In order to achieve the above objective, the present invention also proposes a computer-readable storage medium.

[0189] The computer-readable storage medium provided in the embodiment of the present invention stores a computer program, and when the program is executed by a processor, the security warning method according to the embodiment of the present invention is implemented, and the method includes:

[0190] When it is determined that the window is not equipped with a guardrail and is in an open state according to the first-category window image of the window in the monitoring area captured by the first camera, controlling the first camera to capture at least two frames of new first-category window images;

[0191] Determine whether the at least two frames of new first-category window images are the same;

[0192] If not, controlling the second camera to collect the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image;

[0193] If the maximum pixel value is greater than the preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent.

[0194] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.

[0195] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0196] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0197] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0198] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A safety early warning method, It is characterized in that Applied to monitoring equipment, the method comprises: When it is determined that the window is not equipped with a guardrail and is in an open state according to the first-category window image of the window in the monitoring area captured by the first camera, controlling the first camera to capture at least two frames of new first-category window images; Determine whether the at least two frames of new first-category window images are the same; If not, controlling the second camera to collect the second type of window image of the window, and processing the second type of window image to obtain the maximum pixel value in the processed second type of window image; If the maximum pixel value is greater than a preset pixel value, the second type of window image is identified, and when a child is identified in the second type of window image, a warning message is sent; Among them, the first camera is an ordinary camera, the first type of window images are ordinary window images, the second camera is a thermal imaging camera, and the second type of window images are thermal imaging window images.

2. The method according to claim 1, It is characterized in that The method further comprises: before determining, based on the first-category window image of the window in the monitoring area captured by the first camera, that the window is not equipped with a guardrail and is in an open state, controlling the first camera to capture at least two frames of new first-category window images; Controlling the first camera to scan the windows of each floor in the monitoring area in sequence; When a window on any floor is scanned, the floor information of the window is determined according to the current pitch angle of the first camera, and the focal length of the first camera is adjusted so that the window is displayed at the center position of the first camera.

3. The method according to claim 2, It is characterized in that The determining the floor information of the window according to the current pitch angle of the first camera includes: According to the current pitch angle of the first camera, a preset mapping relationship is queried to determine the floor information of the window.

4. The method according to claim 2, It is characterized in that After adjusting the focal length of the first camera, the method further includes: Controlling the first camera to collect first-type window images of windows in a monitoring area, and determining whether the windows are equipped with guardrails according to the first-type window images; If not, determine whether the window is in an open state based on the first-type window image.

5. The method according to claim 1, It is characterized in that Processing the second-type window image to obtain a maximum pixel value in the processed second-type window image includes: Performing grayscale processing on the second type of window image to obtain a grayscale image; A maximum grayscale value is obtained from the grayscale image, and the maximum grayscale value is determined as a maximum pixel value.

6. The method according to claim 1, It is characterized in that The identifying the second type of window image comprises: Identify the second type of window image in a preset manner to obtain at least one human portrait in the second type of window image; It is determined whether there is a child in the second-category window image according to the portrait area of ​​each portrait.

7. The method according to claim 6, It is characterized in that The determining, based on the portrait area of ​​each portrait, whether there is a child in the second-type window image comprises: Determine whether the portrait area of ​​each portrait is smaller than the portrait area threshold; If the portrait area of ​​any portrait is smaller than the portrait area threshold, it is determined that there is a child in the second type of window image; If the portrait area of ​​each portrait is greater than or equal to the portrait area threshold, it is determined that there is no child in the second type of window image.

8. A safety warning device, It is characterized in that Configured in monitoring equipment, including: A first control module is configured to control the first camera to capture at least two frames of new first-category window images when it is determined that the window is not equipped with a guardrail and is in an open state based on the first-category window images of the windows in the monitoring area captured by the first camera; A first determination module, used to determine whether the at least two frames of new first-type window images are the same; a processing module, configured to, if not, control the second camera to collect a second type of window image of the window, and process the second type of window image to obtain a maximum pixel value in the processed second type of window image; a second control module, configured to identify the second type of window image if the maximum pixel value is greater than a preset pixel value, and send a warning message when a child is identified in the second type of window image; Among them, the first camera is an ordinary camera, the first type of window images are ordinary window images, the second camera is a thermal imaging camera, and the second type of window images are thermal imaging window images.

9. An electronic device, It is characterized in that include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the security warning method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the security warning method as described in any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Children high-altitude protection method based on deep learning model

    CN110826406A

  • Early warning range determination method, device, equipment and system

    CN111369761A