Risk data acquisition method and system for house asset security risk assessment

Abstract

The invention provides a risk data acquisition method and system for house asset safety risk assessment. The method comprises the steps that an operator wears AR glasses integrated with a camera; the camera captures an image of the hidden danger area, and after a corresponding risk level is selected on the laser pen, track drawing calibration is carried out on a hidden danger target in the hidden danger area through the laser pen, and a captured image is obtained and sent to the intelligent device; inputting a feature description of an operator for a hidden danger target and converting the feature description into a hidden danger text sample; the intelligent device generates one or more hidden danger calibration graphs with calibration tracks according to the light spot tracks in the captured image, and an operator selects the corresponding hidden danger calibration graph as an optimal calibration graph; the risk level, the preferable calibration graph and the hidden danger character sample are associated and input into an evaluation table; through linkage of the AR glasses, the laser pen and the intelligent equipment, image acquisition and evaluation of the hidden danger target are carried out, and the accuracy and efficiency of acquiring and evaluating the hidden danger target by an operator are greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of risk assessment, and more specifically, relates to a risk data collection method and system for assessing the safety risks of housing assets. Background Technology

[0002] In modern society, safety risk assessment of buildings is of paramount importance, as it relates to people's lives and the stability of infrastructure. In practice, the assessment requires staff to go to the construction site, conduct research on various parts of the building, collect images of areas with potential hazards, and attach the corresponding risk level assessment results.

[0003] Current assessment methods typically involve operators carrying cameras, paper, pens, or other recording devices. When a potential hazard is observed, the operator takes a picture of the hazard and then records its risk level and characteristics using paper and pen or the recording device. The problem with this method is that manually taken photos may not clearly record the hazard, affecting the accuracy of subsequent assessments. Furthermore, the images cannot be directly correlated with the operator's assessment results; images must be sequentially correlated with corresponding hazard targets to create an assessment table, which is cumbersome, inefficient, and prone to errors during image correlation, leading to inaccurate assessments.

[0004] Therefore, overcoming the shortcomings of the existing technology is an urgent problem to be solved in this technical field. Summary of the Invention

[0005] The problem this invention aims to solve is how to improve the accuracy and efficiency of collecting and assessing building safety risks.

[0006] Firstly, a risk data collection method for assessing the safety risks of residential assets is provided, including: Operators wear AR glasses with integrated cameras; The camera captures images of the potential hazard area. Simultaneously, after selecting the corresponding risk level on the laser pointer, the camera uses the laser pointer to draw a trajectory to mark the potential hazard target in the potential hazard area. The camera then sends the image of the potential hazard area with the light spot trajectory as the captured image to the smart device. By using AR glasses or a laser pointer, operators can input their descriptions of potential hazards and convert them into text to obtain hazard text samples. The intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, and sends all hazard identification maps to the AR glasses for imaging. The operator selects the corresponding hazard identification map as the preferred identification map. The risk level, the preferred calibration map, and the text sample of the hidden danger are associated and entered together as the assessment result of the hidden danger target into the housing asset safety risk assessment template.

[0007] Preferably, the intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, specifically including: The intelligent device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image, and obtains the position and outline edge of the potential hazard target in the captured image; Based on the location of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edges of the hazard target in the captured image, resulting in one or more hazard calibration maps with calibration trajectories.

[0008] Preferably, the intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, specifically including: The characteristics of the potential hazard target are obtained through intelligent analysis based on the text sample of the hazard; The intelligent device intelligently identifies the potential hazard based on the light spot trajectory in the captured image and the characteristics of the potential hazard, and obtains the position and outline edge of the potential hazard in the captured image; Based on the location of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edges of the hazard target in the captured image, resulting in one or more hazard calibration maps with calibration trajectories.

[0009] Preferably, the intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, specifically including: The intelligent device generates a calibration contour on the captured image that is consistent with the light spot trajectory according to the contour of the light spot trajectory and the position of the light spot trajectory in the captured image, thereby obtaining the hazard calibration map.

[0010] Preferably, the risk data collection method for assessing the safety risks of residential assets further includes: Once the camera acquires the captured image, the smart device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image; The camera captures one or more secondary images of the potential hazard from different distances and / or different angles, and sends the secondary images to the smart device. The intelligent device identifies the location and outline of the potential hazard in the secondary image; Based on the location and contour edge of the potential hazard in the secondary image, multiple calibration contours are generated around the potential hazard in the secondary image to obtain one or more derived calibration maps with calibration trajectories. The derived calibration map is sent to the AR glasses as a hazard calibration map for imaging.

[0011] Preferably, the camera captures images of the potential hazard area, and after selecting the corresponding risk level on the laser pointer, it uses the laser pointer to draw trajectories to mark the potential hazard targets in the potential hazard area, specifically including: When the operator observes a potential hazard area, the camera is pointed towards the hazard area to capture an image; At the same time, the operator judges the risk level of the potential hazard area, and presses the corresponding risk level button on the laser pointer according to the risk level. When the corresponding risk level button is activated, the operator presses the calibration button on the laser pointer, and the laser pointer emits a laser beam, controlling the laser to hit the potential hazard area and making the light spot draw a trajectory around the periphery of the potential hazard target for trajectory calibration.

[0012] Preferably, the risk data collection method for assessing the safety risks of residential assets further includes: After the intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, it generates a folder with a unique number according to the corresponding hazard target and stores all hazard identification maps corresponding to the corresponding hazard target into the corresponding folder.

[0013] Secondly, a risk data collection system for assessing the safety risks of residential assets is provided, which applies the aforementioned risk data collection method for assessing the safety risks of residential assets, and includes: AR glasses, a laser pointer, and a smart device, wherein: The AR glasses are equipped with a camera and are designed for operators to wear. The camera is used to capture images of the hazard area, while the laser pointer is used to select the corresponding risk level and mark the hazard targets in the hazard area by drawing their trajectories. The camera is used to send the image of the hazard area with the light spot trajectory as the captured image to the smart device. The AR glasses or laser pointer are used to input the operator's characteristic description of the potential hazard target and convert it into text to obtain a potential hazard text sample, and then send the potential hazard text sample to the smart device. The intelligent device is used to generate one or more hazard calibration maps with calibration trajectories based on the light spot trajectory in the captured image, and send all hazard calibration maps to AR glasses for imaging, and the operator selects the corresponding hazard calibration map as the preferred calibration map; The intelligent device is also used to associate the risk level, the preferred calibration map, and the hidden danger text sample, and record them together as the assessment result of the hidden danger target into the housing asset safety risk assessment template.

[0014] Preferably, the laser pointer is equipped with a risk level button group and a calibration button, wherein: The risk level button group includes multiple different risk level buttons; Once the operator determines the risk level of the potential hazard area, the corresponding risk level button is pressed to activate it. When the corresponding risk level button is activated, the calibration button is pressed to allow the laser pointer to emit laser light.

[0015] Preferably, the risk level button group includes one or more of the following: a general risk level button, a medium risk level button, a severe risk level button, and a high risk level button.

[0016] Unlike existing technologies, the present invention has at least the following beneficial effects: The AR glasses capture images of potential hazard areas using a camera, and use a laser pointer to mark the hazard targets. The smart device processes the recorded images based on the light spot trajectory to obtain a preferred calibration map corresponding to the hazard target. The smart device associates the risk level, the preferred calibration map, and the hazard text sample, and records them together as the assessment result of the hazard target into the building asset safety risk assessment template. Through the linkage of AR glasses, laser pointer, and smart device, the efficiency and convenience of operators in collecting and assessing hazard targets are greatly improved. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments of the present invention will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is a flowchart of a risk data collection method for assessing the safety risks of housing assets, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of AR glasses in a risk data collection method for assessing the safety risks of housing assets provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of one of the hidden danger targets in the assessment table of a risk data collection method for assessing the safety risks of housing assets provided in an embodiment of the present invention; Figure 4This is a flowchart illustrating the acquisition of hazard identification maps in a risk data collection method for assessing the safety risks of housing assets, provided by an embodiment of the present invention. Figure 5 This is a flowchart illustrating the acquisition of hazard identification maps in another risk data collection method for assessing the safety risks of housing assets, provided by an embodiment of the present invention. Figure 6 This is a derived calibration diagram in a risk data collection method for assessing the safety risks of housing assets provided in an embodiment of the present invention; Figure 7 This is an application diagram of a risk data collection method for assessing the safety risks of housing assets provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] Unless the context otherwise requires, throughout the specification and claims, the term "comprising" is interpreted as openly inclusive, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples; that is, although they may be incorporated into embodiments or examples using the above terms for reasons such as order and position, it does not limit them to be incorporated in combination by a single embodiment or example.

[0021] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of this disclosure, unless otherwise stated, "a plurality of" means two or more. Furthermore, for example, the description may use the prefix "A" or "B" to describe the same type of nouns as two independent entities. In this case, the corresponding features defined with "A" and "B" are used only to distinguish between similar entities and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features.

[0022] In the description of this invention, the expression “A and / or B” (where A and B are used to formally represent specific features) will be used. The corresponding expression includes the following three combinations: only A, only B, and a combination of A and B.

[0023] As used in this invention, “about,” “approximately,” or “approximately” includes the stated value and the average value within an acceptable range of deviation from a particular value, wherein the acceptable range of deviation is determined by a person skilled in the art taking into account the measurement under discussion and the error associated with the measurement of the particular quantity (i.e., the limitations of the measurement system).

[0024] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0025] Example 1: This embodiment provides a risk data collection method for assessing the safety risks of residential assets, such as... Figure 1 As shown, the method flow includes the following.

[0026] In step 101, the operator wears AR glasses with an integrated camera.

[0027] like Figure 2 As shown, the camera is located in the middle area of ​​the AR glasses. After the operator wears the AR glasses, the operator faces the designated area and looks directly through the AR glasses. The camera can then be directly facing the corresponding area and record or photograph that area.

[0028] In step 102, the camera captures images of the hazard area, and after selecting the corresponding risk level on the laser pointer, it marks the hazard target in the hazard area by drawing a trajectory with the laser pointer. The camera then sends the image of the hazard area with the light spot trajectory as the captured image to the smart device.

[0029] In this embodiment, the potential hazard area is the location of the area with a calibrated trajectory. The camera captures images of the potential hazard area by either recording video or shooting frame by frame. The AR glasses, laser pointer, and smart device form a set of interconnected system devices. The AR glasses and laser pointer can wirelessly connect to the smart device via Bluetooth, hotspot, or other communication methods. Both the AR glasses and laser pointer are used to interact with external objects and record photos, transmitting the data obtained after interaction to the smart device. The smart device optimizes and stores the received data or transmits it to the AR glasses for imaging. In this embodiment, the camera can record the potential hazard area by using the shutter button on the laser pointer or by having the operator pause the view for a preset duration. The preset duration is set by those skilled in the art based on actual conditions; using the shutter button to start recording is relatively more stable. In this embodiment, the smart device can be a smartphone or other professional smart data acquisition device.

[0030] Considering that simply photographing the potential hazard area using a camera typically results in a large image that is difficult to accurately identify, this embodiment requires the operator to observe and photograph the potential hazard area using AR glasses while simultaneously using a handheld laser pointer to mark the potential hazard targets within the area. The laser pointer has multiple buttons, including multiple risk level buttons. After the operator visually determines the risk level of the potential hazard, they first press the button corresponding to that risk level on the laser pointer and send the corresponding risk level to the smart device. Activating the corresponding risk level allows for automatic association of the captured image with the corresponding risk level. Then, the operator presses the light source switch on the laser pointer, and controls the laser pointer to draw a trajectory with a light spot around the potential hazard target to mark it. Simultaneously, the camera recognizes the corresponding light spot trajectory and maps it onto the image of the potential hazard area captured by the camera, thus obtaining an image of the potential hazard area with the light spot trajectory. This image is then sent to the smart device.

[0031] In step 103, the operator's description of the hazard target is entered into text using AR glasses or a laser pointer and converted into text to obtain a hazard text sample.

[0032] In this embodiment, when a potential hazard is collected, it is necessary not only to select the corresponding risk level and input the corresponding image, but also to record the operator's description of the hazard on-site. This makes the recording more timely and accurate, allowing for a more intuitive understanding of the risk situation of the hazard during subsequent hazard assessments. The AR glasses or laser pointer are equipped with a microphone recording component for recording the operator's voice. The operator can typically describe the hazard based on its location, type, and shape. The voice content is then converted into text. The intelligent device can more accurately identify the location and type of the hazard in the captured image based on the description of the hazard in the text sample, thereby generating a more precise marking contour around the hazard in the captured image. In this embodiment, the type of hazard may include cracks, weathering, deformation, and insect infestation. The marking contour is used to circle the hazard on the image, marking the hazard object so that users can more easily find the location and size of the hazard object when viewing the image.

[0033] In step 104, the intelligent device generates one or more hazard calibration maps with calibration trajectories based on the light spot trajectory in the captured image, and sends all hazard calibration maps to the AR glasses for imaging. The operator selects the corresponding hazard calibration map as the preferred calibration map.

[0034] In this embodiment, the light spot trajectory is used to define the location range of the potential hazard area. After the intelligent device acquires the captured image, it can directly generate a calibration contour based on the light spot trajectory, thereby obtaining an image that marks the potential hazard target. This image serves as the potential hazard calibration map, which is the image that encloses the potential hazard target. However, the problem is that during normal operation, it is difficult for the operator to accurately draw a circle around the potential hazard target. There may be situations where the light spot contour is not suitable or has a low matching degree with the potential hazard target. Therefore, in this embodiment, after the captured image is sent to the intelligent device, the intelligent device can first identify the potential hazard target based on the range defined by the light spot trajectory, identifying the type, location, and contour of the potential hazard target. If the operator has recorded voice and converted it into a potential hazard text sample, the location and shape of the potential hazard target can be identified more accurately based on the potential hazard text sample. On the basis of the above, the intelligent device generates multiple different calibration contours on the periphery of the potential hazard target in the captured image based on the identified potential hazard target, thereby obtaining multiple more visual potential hazard calibration maps containing the potential hazard target.

[0035] Since the assessment form of the housing asset safety risk assessment template can usually display one or more hazard identification maps for each hazard target, the smart device sends all hazard identification maps to the AR glasses. The AR glasses then display all the generated hazard identification maps to the operator, who selects one or more suitable hazard identification maps as preferred identification maps for subsequent entry into the assessment form.

[0036] In step 105, the risk level, the preferred calibration map, and the hidden danger text sample are associated and entered into the housing asset safety risk assessment template as the assessment result of the hidden danger target.

[0037] In this embodiment, the building asset safety risk assessment template includes an assessment table for recording various potential hazards. Each hazard is described by three indicators: a hazard text sample, a preferred calibration map, and a risk level. For example, when an operator observes an inverted V-shaped crack on the exterior wall of a building, they select a medium risk level on a laser pointer, capture an image of the location of the inverted V-shaped crack on the exterior wall using a camera, and simultaneously mark the outer edge of the crack with a laser pointer. The operator then inputs a voice description of "an inverted V-shaped crack on the exterior wall of the building," obtaining the corresponding hazard text sample. The intelligent device generates multiple hazard calibration maps based on the calibrated captured image. The operator selects the appropriate map from all the hazard calibration maps as the preferred calibration map, and associates the medium risk level, the preferred calibration map, and the hazard text sample "an inverted V-shaped crack on the exterior wall of the building" together as the assessment result for "an inverted V-shaped crack on the exterior wall of the building," which is then entered into the assessment table. Figure 3 The image shown is a schematic diagram illustrating an example of the assessment of one of the potential hazards in the assessment table.

[0038] Example 2: In practical applications, various methods can be used to generate hazard identification maps. This embodiment provides a low-cost operation method. In this embodiment, there is no need to deploy an AI model on the smart device. The identification contour is directly generated on the captured image based on the light spot outline drawn by the operator using a laser pointer to mark the hazard target.

[0039] Unlike the aforementioned Embodiment 1, the intelligent device generates one or more hazard calibration maps with calibration trajectories based on the light spot trajectory in the captured image. Specifically, the intelligent device generates a calibration contour consistent with the light spot trajectory on the captured image according to the outline of the light spot trajectory and the position of the light spot trajectory in the captured image, thereby obtaining the hazard calibration map.

[0040] In this embodiment, after the smart device acquires the captured image, it directly generates a calibration contour on the captured image along the outline of the light spot around the hazard target, thus obtaining the hazard calibration map, which includes the hazard target. This satisfies the requirement for illustrating the hazard target. Furthermore, the smart device and AR glasses do not need to be equipped with a trained AI model, resulting in lower costs. However, this method requires high operator skills. The operator needs to ensure that the shape of the light spot outline drawn by the laser pointer matches the outline of the hazard target as closely as possible each time the operator performs calibration. The operator has to perform relatively precise trajectory drawing each time, resulting in a relatively poor user experience.

[0041] Example 3: Considering that directly generating a calibration contour on the captured image based on the light spot contour to obtain a hazard calibration map, as in Example 2, results in poor matching between the outer contour of the hazard calibration map and the hazard target, or an unattractive outer contour, and also places high demands on the operator's trajectory drawing skills, leading to a poor user experience, this embodiment provides the following design to optimize these issues: Figure 4 As shown, unlike the aforementioned Embodiment 2, the method flow of this embodiment includes the following steps.

[0042] In step 201, the intelligent device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image, and obtains the position and outline edge of the potential hazard target in the captured image.

[0043] That is, in this embodiment, the potential hazard target is first identified by intelligent identification of the light spot trajectory.

[0044] In step 202, based on the position of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edges of the hazard target in the captured image, to obtain one or more hazard calibration maps with calibration trajectories.

[0045] In this embodiment, the smart device has a built-in AI model. When a captured image is obtained from the camera, the AI ​​model identifies the type, location, and edge contour of the potential hazard within the light spot outline based on the range circled by the light spot outline in the captured image. After identifying the corresponding potential hazard, the AI ​​model generates multiple different calibration contours based on the edge contour of the potential hazard. The calibration contour can be consistent with the light spot outline during calibration, the edge contour of the potential hazard, or the outer contour of the object where the potential hazard is located. It can also be a circular or square frame containing the entire potential hazard. The calibration contour can be used to represent the shape of the potential hazard, the shape of the object where the potential hazard is located, or for aesthetic purposes and ease of subsequent observation. Multiple potential hazard calibration images are obtained and sent to the AR glasses for imaging and viewing by the operator. The operator can select the appropriate potential hazard calibration image as the preferred calibration image through gestures or voice.

[0046] To illustrate the above solution more intuitively, this embodiment uses the following example: The operator observes a semi-circular notch in a square glass window. Using the camera on the AR glasses, the entire square glass window is captured, and a laser pointer is used to mark the periphery of the semi-circular notch, encircling it to obtain a captured image. This captured image is sent to a smart device. The smart device identifies the shape, location, and type of the semi-circular notch based on the light spot contour. Once the identified hazard is the semi-circular notch in the square glass window, the smart device generates multiple marking contours around the notch, including those along the outer contour of the square glass window, along the light spot contour, along the outer contour of the semi-circular notch, and a circular contour surrounding the notch. Each marking contour is generated as a hazard marking map on the captured image, resulting in multiple hazard marking maps with these contours. All hazard marking maps are sent to the AR glasses for imaging, and the operator selects one as the preferred marking map.

[0047] In summary, in this embodiment, the AI ​​model identifies potential hazards by using the light spot contours marked by the operator, thereby identifying the type, contour, and location of the hazard. This generates multiple more suitable marking contours for the hazard, resulting in multiple hazard marking maps. The operator then selects a suitable map from all the hazard marking maps as the preferred map based on their needs. This ensures that the selected preferred map more easily meets the schematic requirements, and the operator does not need to precisely draw the light spot contours to generate a suitable hazard marking map, thus improving the user experience.

[0048] Example 4: Based on the above embodiments, in order to further improve the recognition accuracy of AI models in smart devices or AR glasses for potential hazards in captured images, the following design is also involved: the smart device generates one or more hazard calibration maps with calibration trajectories based on the light spot trajectories and hazard text samples in the captured image, such as... Figure 5 As shown, the method flow includes the following steps.

[0049] In step 301, the characteristics of the hidden danger target are obtained through intelligent analysis based on the text sample of the hidden danger.

[0050] That is, in this embodiment, target recognition is not only performed by smart devices, but also requires the combination of the operator's textual samples of potential hazards to determine the characteristics of the potential hazard targets.

[0051] In step 302, the intelligent device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image and the characteristics of the potential hazard target, and obtains the position and outline edge of the potential hazard target in the captured image.

[0052] In step 303, based on the position of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edge of the hazard target in the captured image, to obtain one or more hazard calibration maps with calibration trajectories.

[0053] In this embodiment, the smart device has a built-in AI model. When it acquires a captured image and a hazard text sample from the AR glasses, the AI ​​model can more quickly and accurately identify the type, location, and edge contour of the hazard target within the light spot outline based on the range circled by the light spot contour in the captured image and the feature description of the hazard target in the hazard text sample. After identifying the corresponding hazard target, the AI ​​model will generate multiple different calibration contours based on the edge contour of the hazard target. The calibration contour can be consistent with the light spot contour during calibration, the edge contour of the hazard target, or the outer contour of the object where the hazard target is located. It can also be a circular or square frame containing the entire hazard target. The calibration contour can be used to represent the shape of the hazard target, the shape of the object where the hazard target is located, or for aesthetic purposes and ease of subsequent observation. Multiple hazard calibration images are obtained and sent to the AR glasses for imaging and viewing by the operator. The operator can select the appropriate hazard calibration image as the preferred calibration image through gestures or voice.

[0054] To illustrate the above solution more intuitively, this embodiment uses the following example: The operator observed a semi-circular notch in a square glass window. They then voice-recorded "There is a semi-circular notch in the glass window," and the camera on the AR glasses captured the entire square window. A laser pointer was used to mark the periphery of the semi-circular notch, encircling it and creating a captured image. This image was sent to a smart device. The smart device, based on the light spot outline and the hazard text sample, quickly and accurately identified the shape, location, and type of the semi-circular notch within the light spot outline. Once the hazard target was identified as the semi-circular notch in the square glass window, the smart device generated multiple marking contours around the notch, including those along the outer contour of the square glass window, along the light spot outline, along the outer contour of the semi-circular notch, and a circular contour surrounding the notch. Each marking contour was generated as a hazard marking image on the captured image, resulting in multiple hazard marking images with these contours. All hazard marking images were sent to the AR glasses for imaging, and the operator selected one as the preferred marking image.

[0055] In summary, in this embodiment, the AI ​​model identifies the hazard target by using the light spot outline and hazard text samples marked by the operator. Compared with Embodiment 3, it identifies the type, outline, and location of the hazard target faster and more accurately, thereby generating multiple marking outlines that better match the hazard target, thus generating multiple hazard marking maps. The operator can then select a suitable one from all the hazard marking maps as the preferred marking map according to their own needs. This makes the selected preferred marking map more likely to meet the illustration requirements, and the operator does not need to draw precise light spot outlines to generate a suitable hazard marking map, which improves the user experience.

[0056] Example 5: Furthermore, based on the above embodiments, this embodiment considers another application scenario: when an operator assesses and calibrates a potential hazard from one angle and records the captured image, the operator may deem the observation and recording of the hazard from the current angle or distance unclear or incomplete. Therefore, the operator may choose to re-capture the hazard from another angle or distance. However, if the operator still needs to recalibrate using a laser pointer when capturing images of the hazard from other angles or distances, the entire operation process becomes relatively cumbersome. Therefore, this embodiment also involves the following design: Figure 6 As shown, the risk data collection method for assessing the safety risks of housing assets also includes the following steps.

[0057] In step 401, after the camera acquires the captured image, the smart device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image.

[0058] In this embodiment, the smart device has a built-in AI model. When it acquires a captured image from the AR glasses, the AI ​​model identifies the type, location, and edge contour of the potential target within the light spot outline based on the range circled by the light spot outline on the captured image. Through the above identification, the smart device can know the specific object that needs to be identified and calibrated in the currently captured image.

[0059] In step 402, the camera captures one or more secondary images of the potential hazard target from different distances and / or different angles, and sends the secondary images to the smart device.

[0060] The operator adjusts their position, allowing the camera to capture one or more secondary images of the potential hazard from different distances and / or angles. During this process, the operator does not need to calibrate the hazard; the intelligent device determines the location and outline of the hazard in the secondary images based on the previously calibrated hazard information.

[0061] In step 403, the intelligent device identifies the location and outline edge of the potential hazard target in the secondary image.

[0062] The different distances and / or different angles refer to the comparison with the distance and shooting angle when the image was captured previously; the difference between the secondary image and the previous captured image is that the shooting distance and shooting angle are different, so the smart device will identify the same potential object in the secondary image based on the previously identified potential object.

[0063] In step 404, based on the position and contour edge of the hazard target in the secondary image, a variety of calibration contours are generated around the hazard target in the secondary image to obtain one or more derived calibration maps with calibration trajectories.

[0064] In this embodiment, the following feasible interaction logic is provided for steps 402 to 404: After the corresponding hidden danger object is marked by the laser pointer and the intelligent device identifies the hidden danger object in the captured image according to the light spot trajectory, for the secondary image subsequently received by the intelligent device, no matter how much the distance or angle of the secondary image changes compared to the captured image, the intelligent device continues to search for and identify the hidden danger object in the secondary image, and automatically generates a marking contour around the hidden danger target in the secondary image to obtain a derived marking map, until the laser pointer is activated again to mark a new hidden danger object.

[0065] It is worth mentioning that if the potential object is not identified in the secondary image, the secondary image is marked as a problem image and sent to the AR glasses for imaging in the future. The angle difference and distance difference between the problem image and the original captured image are marked on the corresponding problem image to inform the operator that the potential object could not be identified when shooting at that angle and / or distance, allowing the operator to choose whether to reshoot and identify it at that angle and / or distance.

[0066] In step 405, the derived calibration map is sent as a hazard calibration map to the AR glasses for imaging.

[0067] In this embodiment, since the intelligent device can identify the type, location, and shape of the potential hazard after the initial image capture and calibration, when the operator re-captures the same hazard from different angles and distances, the AI ​​model will directly identify the hazard in the secondary image without requiring the operator to use a laser pointer for calibration. Simultaneously, the AI ​​model will directly generate multiple calibration contours around the hazard in the secondary image to obtain the derived calibration map. All generated derived calibration maps are also sent to the AR glasses for imaging, allowing the operator to select the appropriate one from the hazard calibration maps corresponding to different angles or distances.

[0068] To illustrate the above scenarios and methods more intuitively, the following example will be used: The operator discovered a crack in a wall under an eave. They captured an image of the crack from a distance of ten meters, simultaneously marking it with a laser pointer. The intelligent device identified the crack by analyzing the laser beam trajectory within the captured image. The device then generated multiple marker contours around the target hazard in the image, resulting in several initial hazard identification maps. However, the operator also found that at the ten-meter distance, the upper part of the crack was obscured by the eaves, preventing a complete image capture. Therefore, the operator moved closer to five meters and adjusted the camera's position... The viewing angle is raised by 30 degrees to capture a secondary image of the complete wall crack. Since the smart device has already identified the wall crack, it directly identifies and locates the wall crack in the secondary image upon receiving it. Different calibration contours are generated around the wall crack in the secondary image, resulting in multiple derived calibration maps. This eliminates the need for the operator to recalibrate using a laser pointer at the same viewing angle and distance. All derived calibration maps are sent to the AR glasses as the second batch of hazard calibration maps for imaging. The operator then selects a suitable preferred calibration map from the first and second batch of hazard calibration maps.

[0069] Example 6: Based on the above embodiments, this embodiment considers that when assessing potential hazards through image capture, operators need to directly provide the risk level of the hazard on-site and subsequently enter it into the assessment form. Therefore, this embodiment provides a risk level entry logic as follows: The camera captures images of the hazard area, and after selecting the corresponding risk level on the laser pointer, the laser pointer is used to draw a trajectory to mark the hazard target in the hazard area, such as... Figure 7 As shown, the method flow includes the following.

[0070] In step 501, when the operator observes the potential hazard area, the camera is pointed towards the potential hazard area to capture an image.

[0071] In step 502, the operator simultaneously determines the risk level of the potential hazard area, presses the corresponding risk level button on the laser pointer according to the risk level, and presses the calibration button on the laser pointer when the corresponding risk level button is activated. The laser pointer emits a laser, controls the laser to hit the potential hazard area, and makes the light spot draw a trajectory around the periphery of the potential hazard target for trajectory calibration.

[0072] In this embodiment, after the operator observes the potential hazard, they can press the corresponding risk level button on the laser pointer to activate the corresponding risk level. At this time, the corresponding risk level button on the laser pointer will be continuously lit. In this state, the calibration button on the laser pointer will be pressed to calibrate the potential hazard. The subsequent hazard calibration map and preferred calibration map of the potential hazard object are all associated with the corresponding risk level.

[0073] Furthermore, in this embodiment, considering that multiple hazard identification images will be generated when optimizing the captured image, and it is necessary to select one of all hazard identification images as the preferred identification image, but considering that these hazard identification images need to be re-observed and evaluated when reviewing the hazard targets later, it is necessary to store all the generated hazard identification images. Therefore, this embodiment also involves the following design: the risk data collection method for building asset safety risk assessment further includes: after the intelligent device generates one or more hazard identification images with identification trajectories based on the light spot trajectory in the captured image, a folder with a unique number is generated according to the corresponding hazard target, and all hazard identification images corresponding to the corresponding hazard target are stored in the corresponding folder.

[0074] Furthermore, in this embodiment, in the corresponding evaluation table, a hyperlink is generated below each preferred calibration map corresponding to a hazard target. The hyperlink corresponds to the folder used to store the hazard calibration map for that hazard target. Through the hyperlink, one can directly jump to the corresponding folder to browse the hazard calibration map in the folder, or select a suitable hazard calibration map to generate.

[0075] Example 6: This embodiment, based on the above embodiments, provides a risk data acquisition system for building asset safety risk assessment, using the aforementioned risk data acquisition method for building asset safety risk assessment. The system includes: AR glasses, a laser pointer, and a smart device. The AR glasses integrate a camera and are worn by the operator. The camera captures images of potential hazard areas, while the laser pointer selects the corresponding risk level and marks the hazard targets within the hazard area by drawing trajectories. The camera sends the image of the hazard area with the light spot trajectory as a captured image to the smart device. The AR glasses or laser pointer inputs the operator's characteristic description of the hazard target and converts it into text, obtaining a hazard text sample, which is then sent to the smart device. The smart device generates one or more hazard calibration maps with calibration trajectories based on the light spot trajectory in the captured image and sends all hazard calibration maps to the AR glasses for imaging. The operator selects the corresponding hazard calibration map as the preferred calibration map. The smart device also associates the risk level, the preferred calibration map, and the hazard text sample, recording them together as the assessment result of the hazard target in the building asset safety risk assessment template.

[0076] In this embodiment, the smart device can be a smartphone or other professional smart data acquisition device. The laser pointer can have a red laser emitter, a set of function buttons, a wireless transceiver module, a microprocessor, and a power module. The function buttons include: a power-on button, a risk level button group, a calibration button, and a wireless signal connection button. The laser pointer or the AR glasses integrate a microphone recording component for recording the user's voice content.

[0077] In this embodiment, the risk level button group and the calibration button are designed as follows: the risk level button group includes multiple different risk level buttons; when the operator judges the risk level of the potential hazard area, the corresponding risk level button is pressed to activate it; when the corresponding risk level button is activated, the calibration button is pressed to allow the laser pointer to emit laser light.

[0078] The risk level button group includes one or more of the following: general risk level button, medium risk level button, severe risk level button, and high risk level button.

[0079] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A risk data collection method for assessing the safety risks of residential assets, characterized in that, include: Operators wear AR glasses with integrated cameras; The camera captures images of the potential hazard area. Simultaneously, after selecting the corresponding risk level on the laser pointer, the camera uses the laser pointer to draw a trajectory to mark the potential hazard target in the potential hazard area. The camera then sends the image of the potential hazard area with the light spot trajectory as the captured image to the smart device. By using AR glasses or a laser pointer, operators can input their descriptions of potential hazards and convert them into text to obtain hazard text samples. The intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, and sends all hazard identification maps to the AR glasses for imaging. The operator selects the corresponding hazard identification map as the preferred identification map. The risk level, the preferred calibration map, and the text sample of the hidden danger are associated and entered together as the assessment result of the hidden danger target into the housing asset safety risk assessment template.

2. The risk data collection method for assessing the safety risks of housing assets according to claim 1, characterized in that, The intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectories in the captured image, specifically including: The intelligent device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image, and obtains the position and outline edge of the potential hazard target in the captured image; Based on the location of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edges of the hazard target in the captured image, resulting in one or more hazard calibration maps with calibration trajectories.

3. The risk data collection method for assessing the safety risks of housing assets according to claim 1, characterized in that, The intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectories in the captured image, specifically including: The characteristics of the potential hazard target are obtained through intelligent analysis based on the text sample of the hazard; The intelligent device intelligently identifies the potential hazard based on the light spot trajectory in the captured image and the characteristics of the potential hazard, and obtains the position and outline edge of the potential hazard in the captured image; Based on the location of the hazard target in the captured image, multiple calibration contours are generated around the hazard target in the captured image, and / or calibration contours are generated along the contour edges of the hazard target in the captured image, resulting in one or more hazard calibration maps with calibration trajectories.

4. The risk data collection method for assessing the safety risks of housing assets according to claim 1, characterized in that, The intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectories in the captured image, specifically including: The intelligent device generates a calibration contour on the captured image that is consistent with the light spot trajectory according to the contour of the light spot trajectory and the position of the light spot trajectory in the captured image, thereby obtaining the hazard calibration map.

5. The risk data collection method for assessing the safety risks of housing assets according to claim 1, characterized in that, The risk data collection method for assessing the safety risks of housing assets also includes: Once the camera acquires the captured image, the smart device intelligently identifies the potential hazard target based on the light spot trajectory in the captured image; The camera captures one or more secondary images of the potential hazard from different distances and / or different angles, and sends the secondary images to the smart device. The intelligent device identifies the location and outline of the potential hazard in the secondary image; Based on the location and contour edge of the potential hazard in the secondary image, multiple calibration contours are generated around the potential hazard in the secondary image to obtain one or more derived calibration maps with calibration trajectories. The derived calibration map is sent to the AR glasses as a hazard calibration map for imaging.

6. The risk data collection method for assessing the safety risks of residential assets according to claim 1, characterized in that, The camera captures images of the potential hazard area, and after selecting the corresponding risk level on the laser pointer, it uses the laser pointer to draw trajectories to mark the potential hazard targets within the area. Specifically, this includes: When the operator observes a potential hazard area, the camera is pointed towards the hazard area to capture an image; At the same time, the operator judges the risk level of the potential hazard area, and presses the corresponding risk level button on the laser pointer according to the risk level. When the corresponding risk level button is activated, the operator presses the calibration button on the laser pointer, and the laser pointer emits a laser beam, controlling the laser to hit the potential hazard area and making the light spot draw a trajectory around the periphery of the potential hazard target for trajectory calibration.

7. The risk data collection method for assessing the safety risks of residential assets according to claim 1, characterized in that, The risk data collection method for assessing the safety risks of housing assets also includes: After the intelligent device generates one or more hazard identification maps with calibration trajectories based on the light spot trajectory in the captured image, it generates a folder with a unique number according to the corresponding hazard target and stores all hazard identification maps corresponding to the corresponding hazard target into the corresponding folder.

8. A risk data acquisition system for assessing the safety risks of residential assets, used in applying the risk data acquisition method for assessing the safety risks of residential assets as described in any one of claims 1-7, characterized in that, include: AR glasses, laser pointers, and smart devices, among which: The AR glasses are equipped with a camera and are designed for operators to wear. The camera is used to capture images of the hazard area, while the laser pointer is used to select the corresponding risk level and mark the hazard targets in the hazard area by drawing their trajectories. The camera is used to send the image of the hazard area with the light spot trajectory as the captured image to the smart device. The AR glasses or laser pointer are used to input the operator's characteristic description of the potential hazard target and convert it into text to obtain a potential hazard text sample, and then send the potential hazard text sample to the smart device. The intelligent device is used to generate one or more hazard calibration maps with calibration trajectories based on the light spot trajectory in the captured image, and send all hazard calibration maps to AR glasses for imaging, and the operator selects the corresponding hazard calibration map as the preferred calibration map; The intelligent device is also used to associate the risk level, the preferred calibration map, and the hidden danger text sample, and record them together as the assessment result of the hidden danger target into the housing asset safety risk assessment template.

9. The risk data acquisition system for assessing the safety risks of residential assets according to claim 8, characterized in that, The laser pointer is equipped with a risk level button group and a calibration button, wherein: The risk level button group includes multiple different risk level buttons; Once the operator determines the risk level of the potential hazard area, the corresponding risk level button is pressed to activate it. When the corresponding risk level button is activated, the calibration button is pressed to allow the laser pointer to emit laser light.

10. The risk data acquisition system for assessing the safety risks of residential assets according to claim 9, characterized in that, The risk level button group includes one or more of the following: general risk level button, medium risk level button, severe risk level button, and high risk level button.

Images