An elevator inspection method and system based on augmented reality technology

Abstract

This invention relates to the field of augmented reality (AR) technology, and in particular provides an elevator inspection method and system based on AR technology. The method includes receiving a beam and analyzing its phase hysteresis; converting each phase hysteresis value into spatial coordinate values ​​according to the spatial arrangement order of directional reflection units to obtain a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values; synchronizing the longitudinal cross-sectional profile dot matrix with real-time spatial attitude parameters output by an orientation sensing component moving with the inspection device to obtain a virtual ruler grid; projecting the virtual ruler grid onto the surface of the elevator component under inspection; acquiring the offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection; comparing the offset vector point-by-point with a preset allowable offset threshold to obtain a quantified value of the deviation degree. This invention achieves automated, high-precision, and visualized inspection of elevator component morphology, providing reliable data support for elevator safety status assessment.

Description

Technical Field

[0001] This invention relates to the field of augmented reality technology, and in particular to an elevator inspection method and system based on augmented reality technology. Background Technology

[0002] Elevators, as indispensable vertical transportation equipment in modern buildings, directly impact the safety of passengers' lives and property. Elevator inspection and maintenance are crucial for ensuring their safe operation. Accurate inspection of elevator components, such as guide rails, car, and landing doors, to verify their dimensions, installation positions, and surface contours against design specifications is a fundamental basis for assessing their mechanical performance and operational stability. Traditional elevator inspection relies primarily on manual measurement using physical tools such as calipers, levels, and laser rangefinders. This method suffers from inefficiency, subjectivity, difficulty in achieving comprehensive and rapid scanning, and limited accuracy for complex curved surfaces or concealed areas. With the development of computer vision and augmented reality technologies, inspection methods using projected structured light or laser meshes for 3D shape reconstruction have emerged. However, their application in the dynamic, lighting-varying, and structurally complex enclosed environment of elevator shafts still faces multiple challenges in balancing real-time performance, robustness, accuracy, and engineering practicality.

[0003] In existing technologies, advanced measurement methods related to elevator inspection mainly include the following categories: 1. Laser scanning-based 3D reconstruction technology: This technology scans the elevator shaft or component surface using lidar or line laser scanners to acquire point cloud data and reconstruct a 3D model, which is then compared with the design model. This technology has high accuracy, but the equipment is expensive, data processing is complex, real-time performance is poor, and it is sensitive to the on-site environment, such as vibration and dust, making it difficult to achieve rapid and dynamic measurements in elevator operation inspections. 2. Structured light projection-based visual measurement technology: This technology projects an coded structured light pattern onto the surface being measured, captures the deformed pattern with a camera, and calculates the 3D morphology of the surface. It is widely used in industrial inspection, but it usually requires a stable projection and imaging environment, is weakly resistant to interference factors such as complex lighting, uneven reflection, and component movement that may exist in the elevator shaft, and the system calibration is complex, making it difficult to adapt to dynamic changes in the viewing angle when the inspection device moves in the shaft. 3. Positioning and Measurement Technology Based on Inertial Navigation and Visual Fusion: Combining an inertial measurement unit (IMU) and visual odometry, this technology enables autonomous positioning and scene reconstruction of the inspection device within the elevator shaft. This method focuses on global positioning and map building, but has limited ability to accurately and rapidly quantify microscopic surface deviations of components, and accumulated errors may affect the accuracy of long-term measurements. 4. Augmented Reality Measurement Technology Based on Pre-set Targets: Specific targets, such as QR codes or reflective markers, are pre-positioned within the elevator shaft. The correspondence between the world coordinate system and the image coordinate system is established by recognizing these targets, and virtual information is then overlaid for measurement. This method relies on the density and recognition stability of the targets. Deployment and maintenance costs are high in large, structurally complex elevator shafts, and the targets may be obstructed or damaged, affecting system reliability.

[0004] In summary, existing technologies have the following drawbacks when applied to elevator inspection scenarios: Insufficient environmental adaptability: Elevator shaft environments are typically subject to interference from factors such as changing lighting, dust, oil stains, and mechanical vibrations. Existing visual or laser measurement technologies are easily affected by these factors, leading to high data noise, difficulty in feature extraction, and decreased measurement stability. Real-time performance and efficiency bottlenecks: Many high-precision measurement methods, such as 3D laser scanning, consume significant time for data acquisition and processing, making it difficult to meet the real-time requirements of rapid, point-by-point, or area-by-area screening of numerous components in elevator inspections, resulting in low inspection efficiency. Limited dynamic measurement capabilities: When the inspection device moves within the shaft or elevator components are in slight motion, existing technologies often struggle to achieve high-precision dynamic tracking and real-time measurement, easily introducing motion fuzziness or registration errors. Weak Capability for Complex Surfaces and Region Segmentation: For elevator components with complex surfaces or composed of multiple geometrically closed regions, such as embossed landing doors and multi-segmented guide rails, existing methods lack effective technical means to accurately quantify local morphological deviations and perform regional clustering and analysis based on the actual geometric boundaries of the components. Most methods only provide overall deviation statistics, making it difficult to pinpoint the out-of-tolerance situation in specific areas, which is not conducive to targeted maintenance. Low System Integration and Practicality: Existing solutions often require complex external equipment, cumbersome calibration procedures, or specific environmental setups. The system's portability, ease of use, and deployment convenience in real-world inspection scenarios are poor, increasing operational difficulty and cost. Weak Correlation Between Measurement Results and Engineering Semantics: Measurement results are mostly presented in the form of point cloud deviations or overall errors, failing to effectively correlate with the actual functional areas of elevator components, such as the working surface of the guide rails and the sealing surface of the landing doors. This hinders inspectors from intuitively understanding the engineering significance of the deviations and making maintenance decisions. Summary of the Invention

[0005] To achieve the above objectives, the present invention adopts the following technical solution: The elevator inspection method and system proposed in this invention, based on augmented reality technology, aims to overcome the shortcomings of the prior art by integrating a series of technical features such as directional reflection unit beam scanning, real-time spatial attitude synchronization, virtual scale grid dynamic binding, offset vector field gradient analysis, and out-of-tolerance node regional clustering based on graph theory and geometric inclusion determination. This enables efficient, accurate, regionalized, and engineering semantically clear quantitative inspection of surface morphology deviations of elevator components.

[0006] One aspect of the present invention provides an elevator inspection method based on augmented reality technology, comprising the following steps: A continuous scanning beam is emitted to multiple directional reflection units arranged longitudinally at intervals on the inner wall of the elevator shaft. The beams returned by each directional reflection unit are received and their phase lag is analyzed. The phase lag is converted into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units, and a longitudinal cross-sectional profile lattice composed of spatial coordinate values ​​is obtained. The longitudinal section contour dot matrix is ​​spatiotemporally synchronized with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are then mapped to the observation coordinate system of the inspection device through coordinate transformation to obtain a set of virtual scale meshes bound to the current observation direction of the inspection device. A virtual ruler grid is projected onto the surface of the elevator component under inspection through an optical imaging component. The offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection is collected. The offset vector is compared point by point with a preset allowable offset threshold to obtain a quantitative value of the degree of deviation of each area on the surface of the elevator component under inspection from the design shape.

[0007] In one aspect, the present invention provides an elevator inspection system based on augmented reality technology, comprising: The hysteresis conversion module is used to emit continuous scanning beams to multiple directional reflection units arranged longitudinally on the inner wall of the elevator shaft, receive the beams returned by each directional reflection unit and analyze their phase hysteresis, and convert each phase hysteresis into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units to obtain a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values. The spatiotemporal synchronization module is used to spatiotemporally synchronize the longitudinal section contour dot matrix with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are mapped to the observation coordinate system of the inspection device through coordinate transformation to obtain a set of virtual scale grids bound to the current observation direction of the inspection device. The offset vector comparison module is used to project a virtual ruler grid onto the surface of the elevator component under inspection through an optical imaging component, collect the offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection, and compare the offset vector with a preset allowable offset threshold point by point to obtain the quantitative value of the degree of deviation of each area on the surface of the elevator component under inspection from the design shape.

[0008] This invention achieves high-precision 3D point matrix reconstruction of the longitudinal cross-sectional profile of elevator shafts by analyzing and converting the beam phase hysteresis returned by the directional reflection unit, providing an accurate spatial reference for inspection. The profile point matrix is ​​spatiotemporally synchronized with real-time spatial attitude parameters and mapped to the observation coordinate system of the inspection device through coordinate transformation, ensuring real-time matching between the virtual ruler grid and the observation direction of the inspection device, thus solving the coordinate drift problem caused by device movement during inspection. The virtual ruler grid is projected onto the component surface through an optical imaging component, and the offset vector between the projected image and the actual profile is collected, enabling real-time visual comparison between the component surface morphology and the design morphology. Then, quantified deviation data is output through threshold comparison. From profile scanning and coordinate synchronization to projection comparison, a closed-loop inspection process is formed, enabling the identification and quantification of elevator component morphological deviations without manual intervention, improving inspection efficiency and consistency. Based on beam scanning and optical projection, non-contact measurement is achieved, avoiding physical impact on the component surface, while phase analysis and coordinate mapping ensure sub-millimeter measurement accuracy. Attached Figure Description

[0009] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the elevator inspection method based on augmented reality technology provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the elevator inspection method based on augmented reality technology provided in Embodiment 1 of the present invention; Figure 3 This is a process diagram of obtaining a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values, as provided in Embodiment 2 of the present invention. Figure 4 This is a process diagram of obtaining a set of virtual scale grids bound to the current observation direction of the inspection device, as provided in Embodiment 5 of the present invention; Figure 5 This is a process diagram illustrating the quantitative values ​​of the deviation of each region on the surface of the inspected elevator component from the design shape, as provided in Embodiment 9 of the present invention. Figure 6 This is a block diagram of the elevator inspection system based on augmented reality technology provided in Embodiment 15 of the present invention; Figure 7 A block diagram of the electronic device provided by the present invention; Figure 8 A block diagram of a computer-readable storage medium provided for this invention.

[0010] Reference numerals in the attached diagram: 1. Lag conversion module; 2. Spatiotemporal synchronization module; 3. Offset vector comparison module; 4. Central processing unit / microprocessor / main control chip; 5. Storage medium; 6. Data bus; 7. Input / output bus / external bus / device bus; 8. Display; 9. Input / output device; 10. Computer-readable instructions; 11. Non-transitory computer-readable storage medium. Detailed Implementation

[0011] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0012] Hereinafter, the terms "first," "second," etc., are used for descriptive convenience 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," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0013] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed mechanical connection, a detachable mechanical connection, or an integral part; or, "connection" can be a direct connection or an indirect connection through an intermediate medium. Furthermore, unless otherwise explicitly specified and limited, the term "coupling" should be interpreted broadly. For example, "coupling" can be a direct electrical connection, such as physical contact and electrical conduction between two components; it can also be understood as an electrical connection between different components in a circuit structure through physical lines capable of transmitting electrical signals, such as copper foil or wires on a printed circuit board (PCB), to transmit electrical signals; or, "coupling" can be an indirect electrical connection between two components through an intermediate medium; or, "coupling" can be an electrical connection between two components in a non-contact manner, such as an electrical connection between two components using capacitive coupling to transmit electrical signals.

[0014] In this embodiment of the invention, directional terms such as "up," "down," "left," and "right" may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.

[0015] A specific application scenario of this invention: During the elevator installation and acceptance phase of a high-rise building, inspectors use an inspection device integrating a continuous scanning beam transceiver unit, an orientation sensing component, and an optical imaging component to enter the elevator shaft. The inspection device moves longitudinally from bottom to top along a pre-set guide rail within the shaft.

[0016] Before the inspection began, directional reflection units were fixedly installed every 1.5 meters along the longitudinal direction on the inner wall of the elevator shaft, for a total of 20 units, forming a spatial reference point array with longitudinal intervals. After the inspection device was started, its beam transceiver unit sequentially emitted continuous scanning beams to each directional reflection unit and received the beams returned by each unit. The device resolved the phase lag corresponding to each directional reflection unit, and used the relationship given in Example 1 to constrain and verify the difference between adjacent phase lag values, identifying and correcting an abnormal phase lag value caused by construction obstruction within the shaft. Subsequently, the device converted the corrected phase lag values ​​into spatial coordinate values ​​with the device as the origin according to the spatial arrangement order of the directional reflection units, obtaining a longitudinal cross-sectional profile point array composed of 20 spatial coordinate values.

[0017] During the device's movement, the orientation sensing component outputs the roll angle, pitch angle, yaw angle, and displacement integral at each moment in real time. Following the steps of Examples 2 to 4, the device correlates each spatial coordinate value with the spatial attitude parameters at the same moment according to the receiving sequence. By comparing the difference in spatial coordinate values ​​corresponding to adjacent directional reflector units with the difference in displacement integral output by the orientation sensing component, it identifies the speed fluctuation caused by slight unevenness of the guide rail when the device passes through the interval between the 8th and 12th directional reflector units. Based on this, the device divides this interval into independent segments and calculates the cumulative longitudinal displacement compensation, which is then superimposed onto the longitudinal coordinate components of the corresponding spatial coordinate values ​​to obtain a longitudinal cross-sectional profile lattice after dynamic error compensation. Subsequently, the device connects the compensated coordinate points located on the same horizontal projection plane to form a spatial reference network with 20 directional reflector units as nodes.

[0018] The device continues to rise to the elevator guide rail to be inspected. The operator holds the tablet-shaped inspection device and aligns the optical imaging component with the guide rail surface. Following the steps of Examples 5 to 8, the device spatiotemporally synchronizes the spatial reference grid with the attitude angle output by the orientation sensing component at the current moment, and maps it into an observation coordinate system with the optical center of the optical imaging component as the origin and the current observation direction as the Z-axis. Based on the observation field of view of the optical imaging component, the horizontal 60 degrees, and the spatial distance between each horizontal line and the optical center, the device calculates the theoretical projection length of each horizontal line on the image plane, and divides it proportionally according to a preset grid density of 20 division points per meter to generate a horizontal division ratio parameter sequence. Subsequently, the device connects each horizontal line to the adjacent horizontal line according to the division points to form a grid skeleton, ultimately forming a set of virtual ruler grids bound to the current observation direction. The display position of the virtual ruler grid on the tablet screen precisely corresponds to the actual spatial position of the guide rail.

[0019] The device projects a virtual scale grid onto the guide rail surface in green visible light using an optical imaging component, while simultaneously acquiring a real-time image containing both the grid projection and the actual contour of the guide rail. The operator can visually observe a significant offset between the projected grid lines and the guide rail edge within a 2-meter section in the middle of the guide rail. Following the steps of Examples 9 to 14, the device extracts the theoretical imaging position of each grid node and the imaging trajectory of the actual guide rail contour edge from the projected image, calculating the image plane offset vector at each grid node. By analyzing the gradient changes of the offset vector, the device identifies a continuous offset region in the middle of the guide rail. The device compares the offset vector components of each grid node point-by-point with a preset allowable offset threshold, such as 3 millimeters, marking nodes exceeding the threshold as out-of-tolerance nodes and grouping them into connected groups based on the grid topology adjacency. Subsequently, the device determines the spatial inclusion relationship between the connected groups and the closed region formed by the guide rail contour edge, segments the connected groups that span different regions, and finally clusters the continuous out-of-tolerance nodes in the middle of the guide rail into an independent region. The independent region contains 36 grid nodes, of which 32 nodes have offset vectors that exceed the threshold. The device calculates the deviation degree quantization value of the independent region as 88.9%, and displays this result along with the offset position and offset vector field distribution on the flat panel's operating interface.

[0020] Based on the quantitative results and visual annotations on the screen, inspectors quickly locate the deformed sections of the guide rail and record the deviation values ​​as the basis for acceptance. The entire inspection process does not require any physical rulers or contact measuring tools on the guide rail surface; a single longitudinal movement can complete the entire closed-loop inspection process, from establishing the wellbore reference and dynamically calibrating the device to quantitatively outputting the surface deviation of the components.

[0021] Example 1: As Figure 1As shown, this embodiment of the invention provides an elevator inspection method based on augmented reality technology, comprising the following steps: Step S100: Continuous scanning beams are emitted to multiple directional reflection units arranged longitudinally on the inner wall of the elevator shaft, the beams returned by each directional reflection unit are received and their phase lag is analyzed, and each phase lag is converted into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units to obtain a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values. Let the wavelength of the continuous scanning beam be... The longitudinal spacing between two adjacent directional reflection units on the inner wall of the shaft is The testing device in the first The phase lag between the returned beam and the transmitted beam received at each directional reflection unit is: Then the following relation is constructed: The relationship links the wavelength of the continuous scanning beam, the longitudinal spacing of the directional reflecting units, and the phase lag difference between adjacent reflecting units; the propagation path increment of the continuous scanning beam between two adjacent directional reflecting units is equal to the longitudinal spacing. The product of the beam propagation direction cosine; when the beam is emitted perpendicular to the inner wall of the shaft, the propagation path increment is the longitudinal spacing. At this time, the difference in phase lag of the returned beam from adjacent reflecting units equals the propagation path increment With beam number The product of the product and the relation establishes a closed condition based on the longitudinal interval through the mathematical expression of physical constraints. By using the relation to constrain the continuous phase lag sequence acquired by the testing device during longitudinal movement, abnormal phase lag values ​​caused by occasional obstruction or multiple reflections in the beam propagation path can be identified and eliminated. When the difference between adjacent phase lag values ​​does not satisfy the relation, it is determined that the measurement value of the corresponding directional reflection unit has a deviation, and the phase lag is corrected according to the longitudinal interval and wavelength to ensure that the spatial coordinate values ​​converted from each phase lag value conform to the preset interval distribution of the directional reflection unit in the longitudinal dimension, providing self-consistent basic data for the longitudinal cross-sectional profile lattice. Step S200: The longitudinal section contour dot matrix is ​​spatiotemporally synchronized with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are mapped to the observation coordinate system of the inspection device through coordinate transformation to obtain a set of virtual scale meshes bound to the current observation direction of the inspection device. Step S300: Project the virtual ruler grid onto the surface of the elevator component under inspection through the optical imaging component, collect the offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection, compare the offset vector with the preset allowable offset threshold point by point, and obtain the quantitative value of the deviation of each area of ​​the surface of the elevator component under inspection from the design shape.

[0022] The longitudinal spacing refers to the arrangement of multiple directional reflective units at predetermined distances along the vertical direction on the inner wall of the elevator shaft. This establishes a spatial reference benchmark aligned with the elevator's direction of travel, enabling the inspection device to continuously acquire discrete position markers on the shaft's inner wall during longitudinal movement, providing spatial anchor points for constructing the longitudinal cross-sectional profile. The optical elements of the directional reflective units, positioned at a specified height on the inner wall of the elevator shaft, possess the characteristic of reflecting the incident beam back along the incident direction. This forms a stable beam reflection array in the longitudinal direction of the shaft, allowing the inspection device to obtain precise distances to each reflective unit by measuring the round-trip time difference of the beam during movement. Phase lag refers to the phase difference between the emitted continuous scanning beam and the beam returned by the directional reflective units. This phase difference reflects the length of the beam propagation path and is used to eliminate integer ambiguity in distance measurements between the inspection device and the reflective units, achieving millimeter-level spatial distance resolution. The spatial coordinate values, derived from the phase lag and using the inspection device as a reference point, represent three-dimensional coordinate data. This data quantifies the actual spatial positions of each directional reflective unit within the shaft, providing the basic data units for forming the profile array. The longitudinal section profile point array is a discrete set of points composed of multiple spatial coordinate values ​​arranged longitudinally on the inner wall of the elevator shaft according to the longitudinal arrangement of directional reflection units. It characterizes the actual geometric shape of the elevator shaft in its longitudinal vertical section, serving as the raw spatial data for fusion with the attitude of the inspection device. The inspection device is a carrier that moves longitudinally along the guide rails within the elevator shaft. It integrates functional units such as beam transceiver, attitude sensing, coordinate transformation, and image projection, providing a unified physical platform and spatial reference for each step. The orientation sensing component, mounted on the inspection device, is an inertial measurement unit that outputs the attitude angles and displacement changes of the inspection device in three-dimensional space in real time, providing dynamic correction parameters to eliminate the influence of the inspection device's own motion on the measurement data. The real-time spatial attitude parameters are the roll angle, pitch angle, yaw angle, and triaxial acceleration integral displacement of the inspection device output by the orientation sensing component at any given time. These parameters are used to transform the longitudinal section profile point array from a measurement coordinate system centered on the device to an absolute coordinate system referenced to the shaft. The virtual scale grid is a spatial virtual graphic generated by transforming the longitudinal cross-sectional contour point matrix and binding it to the current observation direction of the inspection device. It serves as a geometric comparison benchmark within the field of view of the optical imaging component, precisely corresponding to the spatial position of the inspected elevator component. This allows inspectors to obtain a measurement reference consistent with the observation angle without manual calibration. The optical imaging component, a transmission imaging unit mounted on the inspection device, projects the virtual scale grid onto the surface of the inspected elevator component in visible light, while simultaneously acquiring the actual image of the component's surface, achieving spatial overlap and visual superposition between the virtual benchmark and the physical component. The inspected elevator component refers to physical components in the elevator system, such as guide rails, supports, car wall panels, or landing sills, whose geometry requires inspection. These components serve as the carrier for the virtual scale grid projection and the object for extracting the offset vector.The offset vector refers to the spatial deviation vector between the theoretical grid line position and the actual contour edge of the component surface after the virtual scale grid is projected onto the surface of the elevator component under inspection. It is used to quantify the deformation or installation error of each position on the component surface relative to the design shape, and to provide numerical basis for point-by-point comparison.

[0023] The principles described in the above embodiments are referenced in the appendix. Figure 2 This embodiment achieves high-precision three-dimensional point matrix reconstruction of the longitudinal cross-sectional profile of the elevator shaft by analyzing and converting the beam phase hysteresis returned by the directional reflection unit, providing an accurate spatial reference for inspection. The profile point matrix is ​​spatiotemporally synchronized with real-time spatial attitude parameters and mapped to the observation coordinate system of the inspection device through coordinate transformation, ensuring real-time matching between the virtual ruler grid and the observation direction of the inspection device, thus solving the coordinate drift problem caused by device movement during inspection. The virtual ruler grid is projected onto the component surface through an optical imaging component, and the offset vector between the projected image and the actual profile is collected, enabling real-time visual comparison between the component surface morphology and the design morphology. Then, quantified deviation data is output through threshold comparison. From profile scanning and coordinate synchronization to projection comparison, a closed-loop inspection process is formed, enabling the identification and quantification of elevator component morphological deviations without manual intervention, improving inspection efficiency and consistency. Based on beam scanning and optical projection, non-contact measurement is achieved, avoiding physical impact on the component surface, while phase analysis and coordinate mapping ensure sub-millimeter measurement accuracy.

[0024] In summary, this embodiment achieves automation, high precision, and visualization of elevator component morphological inspection, providing reliable data support for elevator safety status assessment.

[0025] Example 2: As Figure 3 As shown, based on Example 1, the process of obtaining the longitudinal cross-sectional profile point matrix composed of spatial coordinate values ​​in step S100 of this embodiment of the invention specifically includes the following steps: Step S101: Convert each phase hysteresis value into spatial coordinate values ​​with the test device as the origin according to the longitudinal interval arrangement of the directional reflection unit on the inner wall of the well. Then, associate each spatial coordinate value with the spatial attitude parameters output by the orientation sensing component at the same time according to the receiving time sequence to obtain the longitudinal cross-sectional profile dot matrix carrying the attitude label. Step S102: Extract the spatial coordinate difference between adjacent directional reflection units from the longitudinal cross-sectional contour dot matrix carrying attitude tags, compare the spatial coordinate difference with the displacement integral of the orientation sensing component in the corresponding time period, and correct the cumulative error introduced by the longitudinal movement of the inspection device in each spatial coordinate value according to the comparison result to obtain the longitudinal cross-sectional contour dot matrix after dynamic error compensation. Step S103: Connect the coordinate points in the longitudinal section contour point array that are located on the same horizontal projection plane after dynamic error compensation to form a spatial reference net with the arrangement position of the directional reflection unit as the node. Project the spatial reference net onto the surface of the elevator component under inspection through the optical imaging component to obtain a virtual comparison benchmark superimposed on the actual contour of the component surface.

[0026] In the above embodiments, this embodiment converts the phase lag into spatial coordinate values ​​with the inspection device as the origin according to the longitudinal interval sequence of the directional reflection units on the inner wall of the well. Each spatial coordinate value is then associated with the spatial attitude parameters output by the orientation sensing component according to the receiving time sequence, achieving spatial coordinate modeling of the inner wall contour of the well. Simultaneously, an attitude label is attached to each coordinate point. This ensures that the contour point array not only contains positional information but also carries the device attitude data at the time of acquisition, providing a basis for spatial relationships and attitude association in subsequent processing. The spatial coordinate difference between adjacent directional reflection units is extracted from the longitudinal cross-sectional contour point array carrying attitude labels, and this difference is compared with the displacement integral of the orientation sensing component within the corresponding time period. Through comparison, the accumulated error introduced by the longitudinal movement of the inspection device can be identified and corrected, thereby generating a longitudinal cross-sectional contour point array with dynamic error compensation. This improves the spatial accuracy of the contour point array, reduces data drift caused by device movement, and ensures the accuracy of the contour representation. By connecting the coordinate points of the longitudinal cross-sectional profile point array that are located on the same horizontal projection plane after dynamic error compensation, a spatial reference net is formed with the arrangement position of the directional reflection unit as the node. The spatial reference net is projected onto the surface of the elevator component under inspection through an optical imaging component, realizing the superposition of the virtual comparison benchmark and the actual component surface profile. This provides an intuitive visual comparison reference, making it easy for inspection personnel to directly observe the deviation between the component surface profile and the theoretical design, thereby supporting the rapid evaluation and verification of the installation quality or deformation state of the elevator component.

[0027] Example 3: Based on Example 2, the process of correcting the cumulative error introduced by the longitudinal movement of the inspection device in step S102 of this embodiment of the invention according to the comparison results specifically includes the following steps: Step S1021: Divide the spatial coordinate values ​​of each point in the longitudinal section contour dot matrix carrying the attitude tag into several adjacent point pairs according to the longitudinal arrangement order of the directional reflection unit. Extract the difference of spatial coordinate values ​​corresponding to each adjacent point pair and the difference of displacement integral output by the orientation sensing component at the corresponding two moments. Use the ratio between the difference of spatial coordinate values ​​and the difference of displacement integral as the instantaneous velocity deviation coefficient of the longitudinal movement of the inspection device within the adjacent point pair interval to obtain the instantaneous velocity deviation coefficient sequence corresponding to each longitudinal interval. Step S1022: Perform difference processing on the instantaneous velocity deviation coefficients corresponding to two adjacent intervals in the instantaneous velocity deviation coefficient sequence. Based on the interval boundary where the absolute value of the difference exceeds the preset fluctuation threshold, divide the longitudinal section profile point matrix into longitudinal segments with different velocity deviation characteristics. Accumulate all instantaneous velocity deviation coefficients in each segment to obtain the longitudinal displacement cumulative compensation amount corresponding to each segment. Step S1023: The cumulative longitudinal displacement compensation for each segment is superimposed onto the longitudinal coordinate components of the corresponding spatial coordinate values ​​in the longitudinal cross-sectional contour point array carrying the attitude label, according to the distribution position of the directional reflection unit in each segment. The compensated spatial coordinate values ​​are then rearranged in the original order to obtain the longitudinal cross-sectional contour point array after dynamic error compensation.

[0028] In the above embodiments, this embodiment divides the spatial coordinate values ​​in the longitudinal cross-sectional contour dot matrix carrying attitude tags into adjacent point pairs according to the longitudinal arrangement order of the directional reflection units, and extracts the difference in spatial coordinate values ​​corresponding to each adjacent point pair and the difference in displacement integral output by the orientation sensing component at two corresponding moments. The ratio of the two is calculated as the instantaneous velocity deviation coefficient of the longitudinal movement of the inspection device within the adjacent point pair interval. This realizes the quantification of the instantaneous velocity deviation of the inspection device during longitudinal movement, establishes the correspondence between spatial coordinate changes and actual device displacement, and provides serialized data based on velocity deviation for error analysis. The instantaneous velocity deviation coefficients of adjacent intervals in the instantaneous velocity deviation coefficient sequence are interpolated. Based on the interval boundary where the absolute value of the difference exceeds the preset fluctuation threshold, the longitudinal cross-sectional profile dot matrix is ​​divided into longitudinal segments with different velocity deviation characteristics. Subsequently, all instantaneous velocity deviation coefficients within each segment are accumulated to obtain the cumulative longitudinal displacement compensation amount corresponding to each segment. This allows for the identification of segments where the velocity changes significantly during the longitudinal movement of the device, achieving segmented segmentation of error characteristics. The overall displacement compensation amount of each characteristic segment is obtained through cumulative calculation, providing a segmented compensation basis for error correction. The cumulative longitudinal displacement compensation amount corresponding to each segment is then superimposed in reverse onto the longitudinal coordinate components of the corresponding spatial coordinate values ​​in the longitudinal cross-sectional profile dot matrix carrying attitude tags, according to the distribution position of the directional reflection units within each segment. By recombining the compensated spatial coordinate values ​​in their original order, a longitudinal cross-sectional profile dot matrix with dynamic error compensation is obtained. This achieves directional correction of the longitudinal coordinate components, effectively eliminating the cumulative error caused by inconsistent longitudinal movement speed of the inspection device or drift of the measurement system, thereby improving the spatial accuracy and consistency of the profile dot matrix in the longitudinal dimension.

[0029] Example 4: Based on Example 3, the process of dividing the longitudinal cross-sectional profile dot matrix into longitudinal segments with different velocity deviation characteristics in step S0122 of this embodiment of the invention specifically includes the following steps: Step S10221: Arrange the coefficients in the instantaneous velocity deviation coefficient sequence according to the longitudinal order of their corresponding directional reflection units, extract the difference between two adjacent instantaneous velocity deviation coefficients, compare each difference with the preset fluctuation threshold item by item, and mark the positions of adjacent coefficients whose absolute difference exceeds the preset fluctuation threshold to obtain the instantaneous velocity deviation coefficient sequence marked with the segmentation position. Step S10222: In the instantaneous velocity deviation coefficient sequence marked with segmentation positions, all instantaneous velocity deviation coefficients located between two adjacent segmentation positions are collected into a coefficient subset. The arithmetic center value of each coefficient in each coefficient subset is extracted as the velocity deviation feature value of the longitudinal segment corresponding to the coefficient subset, and the velocity deviation feature value corresponding to each longitudinal segment is obtained. Step S10223: Associate the velocity deviation characteristic values ​​corresponding to each longitudinal segment with the spatial coordinate value set of the corresponding segment in the longitudinal section contour point array carrying attitude labels, according to the longitudinal distribution range of the directional reflection unit in each segment; use the velocity deviation characteristic values ​​of each segment as the overall compensation benchmark for the longitudinal coordinate components of the spatial coordinate values ​​in the segment, and obtain the longitudinal section contour point array division result with segment velocity deviation labels.

[0030] In the above embodiments, this embodiment achieves automatic identification of the longitudinal position where the velocity deviation changes significantly, providing an accurate basis for segmentation and ensuring that the segmentation boundary is located at the point where the velocity characteristic actually changes abruptly. Transforming the continuous velocity deviation coefficient sequence into a set of segments with representative feature values ​​simplifies the data dimension of the processing, while preserving the overall trend and central level of velocity deviation within each longitudinal segment, providing stable and statistically significant feature parameters for compensation calculation. By using the velocity deviation feature values ​​of each segment as the overall compensation benchmark for the longitudinal coordinate components of the spatial coordinate values ​​within the segment, the segmentation of the longitudinal cross-sectional contour point matrix is ​​realized, and a unified velocity deviation label is assigned to each segment. A mapping relationship between velocity deviation features and the spatial coordinate set is established, providing a direct structured data foundation for segmented error compensation based on a unified benchmark.

[0031] Example 5: Figure 4 As shown, based on Example 1, the process of obtaining a set of virtual scale grids bound to the current observation direction of the inspection device in step S200 of this embodiment of the invention specifically includes the following steps: Step S201: Align the beam receiving time of each spatial coordinate value in the longitudinal section contour dot matrix with the attitude angle output time in the real-time spatial attitude parameters output by the azimuth sensing component; using the beam receiving time corresponding to each directional reflection unit as a reference, extract the roll angle, pitch angle and yaw angle at the same time from the real-time spatial attitude parameters, bind each spatial coordinate value with its corresponding attitude angle, and obtain the longitudinal section contour dot matrix carrying absolute attitude labels. Step S202: Extract the position components of the spatial coordinate values ​​of each directional reflection unit in the absolute coordinate system from the longitudinal cross-sectional contour dot matrix carrying absolute attitude tags; according to the origin and axis of the observation coordinate system corresponding to the current observation direction of the optical imaging component in the inspection device, subtract the position component of the origin of the observation coordinate system from the position components of each spatial coordinate value; then perform projection decomposition according to the unit vector of each axis of the observation coordinate system to obtain the set of projected coordinate values ​​of each directional reflection unit in the observation coordinate system. Step S203: Connect the projected coordinate values ​​of the directional reflection units located in the same longitudinal height range in the set of projected coordinate values ​​to form horizontal lines distributed longitudinally along the elevator shaft; divide each horizontal line at equal intervals according to the observation field of view of the optical imaging component, and collect the position points obtained by interpolation of adjacent horizontal lines at the division nodes into grid nodes; the grid nodes form a virtual scale grid that is bound to the current observation direction of the inspection device and is used to compare with the surface contour of the inspected elevator component.

[0032] In the above embodiments, this embodiment aligns the time of beam reception of the directional reflection unit corresponding to each spatial coordinate value in the longitudinal cross-section contour dot matrix with the time of attitude angle output in the real-time spatial attitude parameters output by the azimuth sensing component, and extracts the roll angle, pitch angle and yaw angle at the same moment based on each beam reception time, and binds each spatial coordinate value with the corresponding attitude angle, thereby realizing the precise correlation between spatial coordinate values ​​and the absolute attitude of the device; ensuring that each coordinate point in the contour dot matrix carries the attitude information of the device in the absolute coordinate system at the moment of acquisition, providing an accurate attitude reference for subsequent coordinate transformation. The positional components of each spatial coordinate value in the absolute coordinate system are extracted from the longitudinal cross-sectional contour point array carrying absolute attitude labels. Based on the origin and axis of the observation coordinate system corresponding to the current observation direction of the optical imaging component, the positional components are translated and decomposed by coordinate system translation and projection. After subtracting the positional component of the origin of the observation coordinate system, the projection is performed according to the unit vector of each axis of the observation coordinate system to obtain the set of projected coordinate values ​​of each directional reflection unit in the observation coordinate system. The contour data in the absolute coordinate system is transformed into a local coordinate system based on the current observation direction of the inspection device, so that all coordinate values ​​are unified under a reference system consistent with the imaging observation angle, thus preparing the data for generating a visual comparison benchmark. The projected coordinate values ​​of directional reflection units located in the same longitudinal height range within the projected coordinate value set are connected to form horizontal lines distributed longitudinally along the elevator shaft. Subsequently, these horizontal lines are equidistantly divided according to the observation field of view of the optical imaging component, and grid nodes are obtained by interpolation between adjacent horizontal lines at the division nodes. These grid nodes constitute a virtual scale grid bound to the current observation direction of the inspection device. The grid forms a regular spatial reference structure in the observation coordinate system, and its node distribution matches the imaging field of view. It can be directly superimposed on the optical image of the surface of the inspected elevator component, providing a precise, stable, and real-time synchronized spatial size comparison benchmark for visual inspection.

[0033] Example 6: Based on Example 5, the process of aggregating the position points obtained by interpolation of adjacent horizontal lines at the segmented node into a grid node in step S203 of this embodiment of the invention specifically includes the following steps: Step S2031: Based on the horizontal angular range of the observation field of view of the optical imaging component and the spatial distance between the location of each horizontal line and the optical center of the optical imaging component, calculate the theoretical projection length of each horizontal line on the image plane corresponding to the observation field of view; divide the theoretical projection length into equal proportions according to the preset grid density value to obtain the horizontal segmentation ratio parameter sequence corresponding to each horizontal line. Step S2032: Connect the division points of each horizontal line according to the division point determined by its corresponding horizontal division ratio parameter sequence to the division points of the two adjacent horizontal lines with the same division ratio parameter position; take the corresponding division points on the adjacent horizontal lines as the endpoints and generate connecting line segments along the vertical direction to obtain the grid skeleton composed of the division points on all horizontal lines and the vertical connecting line segments. Step S2033: Group all the dividing points in the grid skeleton according to their order on the horizontal connecting lines and the vertical layer sequence. Use the closed area enclosed by the line segments between adjacent dividing points on each horizontal connecting line and the vertical connecting line segments between adjacent horizontal connecting lines as the grid unit boundary. Collect the dividing points at the intersection of all grid unit boundaries as the grid nodes of the virtual scale grid. The virtual scale grid bound to the current observation direction of the inspection device is formed by all grid nodes.

[0034] In the above embodiments, this embodiment maps horizontal lines in three-dimensional space to the imaging plane and generates a horizontal segmentation scheme matching the imaging resolution according to a preset grid density, ensuring that the generated grid has a uniform spatial distribution density on the image plane that meets the requirements of visual measurement. Spatial associations of horizontal segmentation points in the vertical dimension are established, and discrete horizontal segmentation points are integrated into a spatial grid framework with a three-dimensional topological structure through vertical connections, providing the basic geometric structure for the construction of the complete grid. The transformation from a grid skeleton to a regular grid structure is completed, the geometric boundary of each grid unit is defined, and the spatial position of the grid nodes is clarified. The final generated set of grid nodes constitutes a two-dimensional reference grid that is regularly arranged in the observation coordinate system and aligned with the imaging field of view, which can be directly used for accurate spatial position comparison and size measurement with the target surface contour.

[0035] Example 7: Based on Example 6, the process of calculating the theoretical projection length of each horizontal line on the image plane corresponding to the observation field of view in step S2031 of this embodiment of the invention specifically includes the following steps: Step S20311: Extract the spatial coordinate values ​​of the directional reflection units corresponding to the two ends of each horizontal line from the longitudinal cross-sectional contour dot matrix carrying the absolute attitude label. Combine the attitude angle output by the orientation sensing component at the same moment to determine the spatial connection direction vector of the two ends of each horizontal line in the absolute coordinate system. Compare the spatial connection direction vector with the optical axis direction of the optical imaging component to obtain the tilt deflection of each horizontal line relative to the image plane of the optical imaging component. Step S20312: Couple the spatial distance between the location of each horizontal line and the optical center of the optical imaging component with the tilt deflection corresponding to each horizontal line. Based on the angle range of the observation field of view of the optical imaging component in the horizontal direction, establish a projection mapping relationship with the spatial coordinates of the two ends of each horizontal line as the boundary and the image plane as the projection target, and obtain the projection start point and projection end point of each horizontal line on the image plane. Step S20313: Take the length of the line connecting the projection start point and projection end point of each horizontal line on the image plane as the theoretical projection length, and divide the theoretical projection length proportionally according to the preset grid density value; take the distance ratio of each segmented position relative to the projection start point as the horizontal segmentation ratio parameter, and collect the horizontal segmentation ratio parameters corresponding to all horizontal lines to form a sequence of horizontal segmentation ratio parameters corresponding to each horizontal line.

[0036] In the above embodiments, this embodiment establishes the geometric relationship between the three-dimensional spatial direction of the horizontal connecting line and the optical axis of the imaging system, quantifies the projection distortion factors caused by the spatial attitude of the horizontal connecting line, and provides accurate tilt correction parameters for projection calculation. It realizes the perspective projection conversion from three-dimensional spatial line segments to the two-dimensional image plane, while considering the combined effects of distance attenuation and tilt deflection, ensuring that the projection mapping conforms to the geometric model of optical imaging, thereby obtaining the accurate endpoint positions of the projection line segments on the image plane. The calculated projection length is converted into operable meshing parameters. Proportional meshing ensures the uniformity of the mesh on the image plane, and the generated segmentation ratio parameter sequence provides a standardized basis for reconstructing mesh nodes with consistent visual density in three-dimensional space.

[0037] Example 8: Based on Example 7, the process of obtaining the projection start point and projection end point of each horizontal connecting line on the image plane in step S20312 of this embodiment of the invention specifically includes the following steps: Step S203121: Transform the spatial coordinates of the directional reflection units corresponding to the two ends of each horizontal line with the roll angle, pitch angle and yaw angle output by the orientation sensing component at the same time; construct an imaging coordinate system with the optical center of the optical imaging component as the origin and the optical axis direction as the Z-axis, and transform the two ends of each horizontal line from the absolute coordinate system to the imaging coordinate system to obtain the position vector of each two ends of the horizontal line in the imaging coordinate system; Step S203122: Based on the horizontal angular range of the observation field of view of the optical imaging component, determine the position parameters of the image plane perpendicular to the optical axis in the imaging coordinate system. Using the position vectors of the two endpoints of each horizontal line as the projection boundary, draw parallel projection lines along the optical axis through the two endpoints to the image plane. Take the intersection of each projection line with the image plane as the projection point of the two endpoints on the image plane. Step S203123: Mark the projection points of the two ends of each horizontal line onto the image plane as the projection start point and projection end point of the horizontal line respectively. According to the spatial arrangement order of the position vectors of the two ends of each horizontal line in the imaging coordinate system, pair the projection start point and projection end point in sequence to obtain the coordinate marks of the projection start point and projection end point of each horizontal line on the image plane.

[0038] In the above embodiments, this embodiment transforms the spatial coordinate values ​​of the directional reflection units corresponding to the two endpoints of each horizontal line with the roll angle, pitch angle, and yaw angle output by the orientation sensing component at the same time. An imaging coordinate system is constructed with the optical center of the optical imaging component as the origin and the optical axis direction as the Z-axis. The two endpoints of each horizontal line are transformed from the absolute coordinate system to the imaging coordinate system to obtain the position vectors of the two endpoints of each horizontal line in the imaging coordinate system. The three-dimensional spatial points in the absolute coordinate system are uniformly transformed to the local coordinate system with the imaging system as the reference, so that the projection calculation can be directly based on the geometric model of the imaging system, eliminating the interference of the absolute attitude on the projection mapping and ensuring that the reference of the projection transformation is strictly aligned with the optical center and optical axis direction of the imaging system. Based on the horizontal angular range of the observation field of view of the optical imaging component, the position parameters of the image plane perpendicular to the optical axis in the imaging coordinate system are determined. Using the position vectors of the two endpoints of each horizontal line as the projection boundary, parallel projection lines along the optical axis are drawn from the two endpoints to the image plane. The intersection of each projection line with the image plane is taken as the projection point of the two endpoints on the image plane. This achieves orthogonal projection mapping from the three-dimensional imaging coordinate system to the two-dimensional image plane. The parallel projection lines along the optical axis ensure that the projection process conforms to the visual geometry of the imaging system. At the same time, the accurate position of the image plane is determined by combining the observation field of view parameters, thus obtaining the precise projection position of the two endpoints of each horizontal line on the image plane. The projection points of the two endpoints of each horizontal line on the image plane are respectively marked as the projection start point and projection end point of the horizontal line. According to the spatial arrangement order of the position vectors of the two endpoints of each horizontal line in the imaging coordinate system, the projection start point and projection end point are paired in sequence to obtain the coordinate marks of the projection start point and projection end point of each horizontal line on the image plane. The logical organization and pairing of the projection points are completed, ensuring that the projection line segment of each horizontal line on the image plane has a clear and consistent sequence of start and end points, providing structurally complete and clearly ordered projection coordinate data for calculating the projection length and dividing the grid.

[0039] Example 9: As Figure 5 As shown, based on Example 1, the process of obtaining the quantified value of the deviation of each area of ​​the surface of the inspected elevator component from the design shape in step S300 of this embodiment of the invention specifically includes the following steps: Step S301: After projecting the virtual ruler grid onto the surface of the elevator component under inspection through the optical imaging component, the theoretical imaging position of each grid node of the virtual ruler grid on the image plane is extracted from the projection image based on the projection image acquired by the optical imaging component; at the same time, the imaging trajectory of the actual contour edge of the surface of the elevator component under inspection on the image plane is extracted from the same projection image; the theoretical imaging position and the imaging trajectory are spatially correlated to obtain the set of image plane offset vectors between the theoretical position of the virtual ruler grid at each grid node and the actual contour of the surface of the elevator component under inspection. Step S302: Group the offset vectors in the image plane offset vector set into a grid according to the vertical and horizontal order of the virtual scale grid. The offset vectors corresponding to adjacent grid nodes in the same vertical order form a horizontal offset vector sequence, and the offset vectors corresponding to adjacent vertical orders in the same horizontal order form a vertical offset vector sequence. Cross-compare the horizontal and vertical offset vector sequences to extract the gradient change of the offset vector at each grid node relative to the offset vectors of adjacent nodes, and obtain the offset vector field carrying local gradient features. Step S303 compares the offset vectors at each grid node in the offset vector field carrying local gradient features with the preset allowable offset threshold point by point. The components of the offset vectors at each grid node that exceed the allowable offset threshold are clustered according to the surface area of ​​the inspected elevator component to which they belong. The ratio of the number of grid nodes exceeding the threshold in each clustered area to the total number of grid nodes in the area is used as the quantification value of the deviation of the area, thus obtaining the quantification value of the deviation of each area of ​​the surface of the inspected elevator component relative to the design shape.

[0040] In the above embodiments, this embodiment establishes a direct spatial correspondence between the theoretical design location and the actual imaging contour. It quantifies the local deformation or positional deviation at each grid node using offset vectors, providing a basic data set based on the image plane coordinate system for deviation analysis. Discrete offset vectors are organized into a vector field with a topological structure. By calculating the gradient changes in the horizontal and vertical directions, the spatial distribution pattern and trend of the deviation are revealed, thus expanding simple node deviations into vector field information reflecting local deformation characteristics and spatial continuity. This achieves the transformation from vector field analysis to regional quantitative evaluation. Abnormal deviation regions exceeding the allowable range are identified through threshold screening and spatial clustering. Using the node ratio as a quantitative indicator, the overall deviation degree of each surface region relative to the design shape is finally output, providing a clear regional quantitative basis for quality assessment and processing.

[0041] Example 10: Based on Example 9, the process of clustering the components of the offset vector at each grid node that exceed the allowable offset threshold according to their respective surface areas of the inspected elevator component in step S303 of this embodiment of the invention specifically includes the following steps: Step S3031: From the offset vector field carrying local gradient features, filter the values ​​of each component of the offset vector at each grid node; mark the grid node whose component value exceeds the allowable offset threshold as an out-of-tolerance node, and extract the vertical sequence index and horizontal sequence index corresponding to each out-of-tolerance node in the virtual scale grid; establish an out-of-tolerance node index set using the two-dimensional index pair composed of the vertical sequence index and the horizontal sequence index as elements. Step S3032: Arrange the two-dimensional index pairs in the out-of-tolerance node index set in ascending order of vertical hierarchical index and horizontal sequential index; traverse adjacent index pairs in turn, and when the difference between the vertical hierarchical index and the difference between the horizontal sequential index of two adjacent index pairs are not greater than the preset connectivity step size; group the out-of-tolerance nodes corresponding to the two adjacent index pairs into the same connectivity group, and after traversal, obtain an out-of-tolerance node connectivity group set composed of several connectivity groups; Step S3033: Determine the spatial inclusion relationship between the grid node positions corresponding to the out-of-tolerance nodes in each connected group of the out-of-tolerance node connected group set and the imaging trajectory of the actual contour edge of the surface of the inspected elevator component on the image plane; use the closed area enclosed by the imaging trajectory as the boundary to divide the connected groups that cross different closed areas into subgroups; use each subgroup after division as the out-of-tolerance node clustering result divided according to the geometric area of ​​the surface of the inspected elevator component.

[0042] In the above embodiments, this embodiment can gradually identify and aggregate spatially continuous out-of-tolerance distributions that are associated with the actual geometric regions of elevator components from discrete grid node out-of-tolerance data. First, an ordered set of out-of-tolerance nodes is established through index filtering and sorting. Then, adjacent nodes are grouped based on connectivity rules to initially form spatially coherent out-of-tolerance regions. Finally, geometric region boundary segmentation ensures that each clustering result strictly corresponds to a specific surface region of the elevator component. This achieves a complete transformation from numerical out-of-tolerance detection to spatial regionalization, providing a structured data foundation for deviation analysis and processing of specific component regions.

[0043] Example 11: Based on Example 10, the process of determining the spatial inclusion relationship between step S3033 of this embodiment and the imaging trajectory of the actual contour edge of the inspected elevator component on the image plane specifically includes the following steps: Step S30331: Map the positions of the grid nodes corresponding to the out-of-tolerance nodes in each connected group of the out-of-tolerance node connected group set from the node index of the virtual scale grid to the pixel coordinates on the image plane, and extract the closed contour boundary contained in the imaging trajectory of the actual contour edge of the inspected elevator component surface on the image plane; discretize the closed contour boundary into a sequence of boundary points arranged in order, and divide each sequence of boundary points into several independent closed regions according to the geometric connectivity of the imaging trajectory; Step S30332: Starting from the pixel coordinates corresponding to each out-of-tolerance node in each connected group, draw a ray in a preset fixed direction in the image plane, and calculate the intersection points of the ray with the adjacent boundary points in the boundary point sequence of each closed region in turn. Count the number of intersection points between the ray and the boundary line of each closed region. Determine whether the out-of-tolerance node is located inside or outside the closed region based on the parity of the number of intersection points, and obtain the attribution label of each out-of-tolerance node and each closed region. Step S30333: Group out-of-tolerance nodes with the same closed region attribution label within the same connected group into subgroups. When there are multiple different attribution labels within a connected group, divide the connected group that spans different closed regions into several subgroups according to the actual position of the closed region boundary on the image plane, so that all out-of-tolerance nodes in each subgroup are located within the same closed region. Use each subgroup after division as the clustering result of out-of-tolerance nodes divided according to the geometric region of the surface of the inspected elevator component.

[0044] In the above embodiments, this embodiment realizes the transformation from grouping out-of-tolerance nodes based on grid connectivity to precise region division based on the actual geometric contour of the component; through coordinate mapping and contour discretization, a digital representation of the geometric region is established; through the parity determination of ray intersections, a stable and reliable spatial attribution discrimination method is provided; finally, through attribution label consistency and boundary segmentation, it is ensured that each cluster subset strictly corresponds to a specific geometric region on the surface of the component; thus enabling out-of-tolerance analysis to be implemented in specific component regions, providing an accurate spatial basis for tracing, evaluating and processing deviations in different regions.

[0045] Example 12: Based on Example 11, the process of dividing a connected group spanning different closed regions into several sub-groups according to the boundary lines in step S30333 of this embodiment of the invention specifically includes the following steps: Step S303331: Select connected groups containing multiple different closed region attribution labels from the set of out-of-tolerance nodes as the groups to be segmented, and extract the pixel coordinates of all out-of-tolerance nodes in the group to be segmented and their corresponding attribution labels; at the same time, extract the polygon vertex set formed by the boundary point sequence of each closed region to obtain the correspondence between the group to be segmented and the boundary polygons of each closed region. Step S303332: Using the pixel coordinates of the out-of-tolerance nodes in the group to be segmented as graph nodes and the topological adjacency relationship between adjacent grid nodes in the virtual scale grid as graph edges, construct the graph structure corresponding to the group to be segmented; convert the boundary polygons of each closed region into closed regions, traverse each graph edge in the graph structure, and when the two graph nodes connected by the graph edge belong to different closed regions in the belonging label, mark the graph edge as a cross-region connection edge; Step S303333: Remove all graph edges marked as cross-regional connection edges from the graph structure corresponding to the group to be segmented, so that the original graph structure is decomposed into several disconnected subgraphs; group the out-of-range nodes corresponding to the graph nodes in each subgraph into a subgroup, so that all out-of-range nodes in each subgroup are located in the same closed region, and obtain the set of segmented subgroups.

[0046] In the above embodiments, this embodiment realizes the process of accurately dividing the out-of-tolerance node connectivity group spanning multiple closed regions into several sub-groups based on geometric boundaries; by constructing a graph structure and identifying cross-regional connection edges, the spatial affiliation is transformed into connection attributes in graph theory; by removing cross-regional connection edges, the graph structure is decomposed, thereby maintaining the topological adjacency of nodes within the group while strictly completing the grouping and segmentation according to the geometric region boundaries; ensuring that each sub-group corresponds one-to-one with a specific closed region on the component surface, so that the out-of-tolerance clustering results retain both local connectivity and conform to the actual geometric partitioning, providing clear and consistent data support for regional deviation analysis.

[0047] Example 13: Based on Example 12, the process of partially constructing the graph structure corresponding to the group to be segmented in step S303332 of this embodiment of the invention specifically includes the following steps: Step S3033321: Extract the vertical hierarchical index and horizontal sequence index of each out-of-tolerance node in the virtual scale grid from the group to be segmented. Based on the vertical hierarchical index and horizontal sequence index, match each out-of-tolerance node with the preset grid node index system in the virtual scale grid to obtain the node positioning parameters of each out-of-tolerance node in the virtual scale grid topology. Step S3033322: Based on the topological adjacency relationship between adjacent grid nodes in the virtual scale grid, filter out the out-of-tolerance node pairs with direct adjacency relationships from the node positioning parameters; take the pixel coordinates corresponding to the two out-of-tolerance nodes in each out-of-tolerance node pair as the two vertices of the graph structure, and take the topological adjacency relationship corresponding to the out-of-tolerance node pair as the graph edge to generate a graph edge set. Step S3033323: Take the pixel coordinates of all out-of-tolerance nodes in the group to be segmented as all vertices of the graph structure, and take all the graph edges in the graph edge set as connections between vertices to form the graph structure corresponding to the group to be segmented.

[0048] In the above embodiments, this embodiment realizes the process of constructing an out-of-tolerance node graph structure based on a virtual scale grid topology structure; the topological position of the node in the grid is determined by index mapping, and the node pairs with direct connection relationship are selected by using the grid's preset adjacency relationship. Then, the node position and connection relationship are mapped to the vertices and edges of the graph, respectively, and finally a complete graph model is formed; the spatial adjacency characteristics of out-of-tolerance nodes in the original grid are preserved, so that the graph structure-based operation can accurately reflect the actual topological relationship between nodes, and provide a structured mathematical model basis for the identification and segmentation of cross-regional connection edges.

[0049] Example 14: Based on Example 13, the process of filtering out pairs of out-of-tolerance nodes with direct adjacency from the node positioning parameters in step S3033322 of this embodiment of the invention specifically includes the following steps: Step S30333221: Based on the topological adjacency relationship between adjacent grid nodes in the virtual scale grid, the judgment condition for direct adjacency is set as follows: the absolute difference of the vertical sequence index of two grid nodes is equal to the preset neighborhood step size and the horizontal sequence index is equal, or the vertical sequence index is equal and the absolute difference of the horizontal sequence index is equal to the preset neighborhood step size. The judgment condition is used as the adjacency relationship benchmark for screening out-of-tolerance node pairs. Step S30333222: Extract the vertical hierarchical index and horizontal sequence index of each out-of-tolerance node from the node location parameters. Sort all out-of-tolerance nodes according to the vertical hierarchical index as the primary key and the horizontal sequence index as the secondary key. Using the sorted sequence as the traversal order, take each out-of-tolerance node as the current node in turn. Based on the adjacency relationship benchmark, at the position where the vertical hierarchical index of the current node is increased or decreased by a preset neighborhood step and the horizontal sequence index is equal, and at the position where the vertical hierarchical index is equal and the horizontal sequence index is increased or decreased by a preset neighborhood step, search for the node location parameters of another out-of-tolerance node. Pair the current node with the search result with the searched out-of-tolerance node to obtain a candidate adjacency pair set. Step S30333223: Merge the same out-of-tolerance node pairs that appear repeatedly in the candidate adjacency pair set, and store them uniquely with the smaller of the two node indices in the out-of-tolerance node pair as the preorder and the larger of the two node indices as the postorder, to obtain the set of out-of-tolerance node pairs with direct adjacency relationships.

[0050] In the above embodiments, this embodiment systematically filters out out-of-tolerance node pairs with direct adjacency relationships from node location parameters; by defining strict adjacency judgment conditions, the geometric basis for node pair filtering is clarified; by traversing and retrieving based on index sorting, the discovery of node pairs that meet the conditions is completed efficiently; by merging and unique storage, a set of adjacency relationships without redundancy is obtained; the adjacency relationships in the grid topology are accurately converted into connection data in the form of node pairs, providing a complete adjacency relationship foundation for constructing the graph edge set, ensuring that the graph structure can truly reflect the original connection topology of out-of-tolerance nodes in the virtual scale grid.

[0051] Example 15: As Figure 6 As shown, based on Embodiments 1-14, the elevator inspection system based on augmented reality technology provided in this embodiment of the invention includes: The hysteresis conversion module 1 is used to emit a continuous scanning beam to multiple directional reflection units arranged longitudinally on the inner wall of the elevator shaft, receive the beams returned by each directional reflection unit and analyze their phase hysteresis, and convert each phase hysteresis into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units to obtain a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values. The spatiotemporal synchronization module 2 is used to spatiotemporally synchronize the longitudinal section contour dot matrix with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are mapped to the observation coordinate system of the inspection device through coordinate transformation to obtain a set of virtual scale grids bound to the current observation direction of the inspection device. The offset vector comparison module 3 is used to project the virtual ruler grid onto the surface of the elevator component under inspection through the optical imaging component, collect the offset vector between the projection image formed by the virtual ruler grid on the surface of the component and the actual contour of the elevator component under inspection, and compare the offset vector with the preset allowable offset threshold point by point to obtain the quantitative value of the degree of deviation of each area of ​​the surface of the elevator component under inspection relative to the design shape.

[0052] In the above embodiments, this embodiment constructs a high-precision three-dimensional contour lattice of the longitudinal section of the elevator shaft through beam phase lag analysis and coordinate transformation of the directional reflection unit, providing a reliable spatial reference benchmark for inspection; the contour lattice is spatiotemporally synchronized with real-time spatial attitude parameters, and mapped to the observation coordinate system of the inspection device through coordinate transformation, realizing real-time dynamic binding between the virtual scale grid and the observation direction of the inspection device, eliminating the measurement benchmark drift caused by changes in device pose; the virtual scale grid is projected onto the surface of the component using an optical imaging component, and the offset vector between the projected image and the actual contour is collected to realize the surface morphology of the component and the design. Real-time visual comparison and deviation extraction of the design shape; point-by-point comparison of the collected offset vector with the preset allowable offset threshold, outputting the quantitative deviation of each area on the surface of the inspected elevator component relative to the design shape, realizing an objective quantitative assessment of the deviation; based on beam scanning, coordinate fusion and optical projection technology, a complete non-contact automatic inspection process is formed, which can complete high-precision shape detection without physical contact with the inspected component, ensuring the safety and repeatability of the inspection process; the modules are connected through data flow to realize closed-loop detection from contour reconstruction, coordinate synchronization to deviation comparison, improving the overall efficiency and consistency of elevator component shape inspection.

[0053] In summary, this embodiment ultimately constructs an augmented reality system capable of automating, visualizing, and performing high-precision inspections of elevator component morphology, providing reliable technical support for elevator safety status assessment.

[0054] Figure 7 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0055] The electronic device may include a central processing unit / microprocessor / main control chip 4; and a storage medium 5 coupled to the central processing unit / microprocessor / main control chip 4 and storing computer-executable instructions therein for performing the steps of various methods of embodiments of the present invention when executed by the processor.

[0056] The central processing unit / microprocessor / main control chip 4 may include, but is not limited to, one or more processors or microprocessors.

[0057] Storage medium 5 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disk, floppy disk, solid-state drive, removable disk, CDROM, DVDROM, Blu-ray disc, etc.).

[0058] In addition, the electronic device may include (but is not limited to) a data bus 6, an input / output bus / external bus / device bus 7, a display 8, and input / output devices 9 (e.g., keyboard, mouse, speaker, etc.).

[0059] The central processing unit / microprocessor / main control chip 4 can communicate with external devices (8, 9, etc.) via wired or wireless networks (not shown) through the input / output bus / external bus / device bus 7.

[0060] The storage medium 5 may also store at least one computer-executable instruction for performing the steps of various functions and / or methods in the embodiments described herein when the central processing unit / microprocessor / main control chip 4 is running.

[0061] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0062] Figure 8 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0063] like Figure 8 As shown, the non-transitory computer-readable storage medium 11 stores instructions, such as computer-readable instructions 10. When the computer-readable instructions 10 are executed by a processor, the various methods described above can be performed. The non-transitory computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-transitory non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium 11 can be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions 10 stored on the non-transitory computer-readable storage medium 11, the various methods described above can be performed.

[0064] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0065] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0066] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0067] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods of the various embodiments of this invention through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0068] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An elevator inspection method based on augmented reality technology, characterized in that, Includes the following steps: The virtual ruler grid is projected onto the surface of the elevator component under inspection through an optical imaging component. The offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection is collected. The offset vector is compared point by point with the preset allowable offset threshold to obtain the quantitative value of the degree of deviation of each area of ​​the surface of the elevator component under inspection from the design shape. The process of obtaining quantitative values ​​of the deviation of each area on the surface of the inspected elevator component from its design shape includes the following steps: After projecting the virtual ruler grid onto the surface of the elevator component under inspection through an optical imaging component, the theoretical imaging position of each grid node of the virtual ruler grid on the image plane is extracted from the projection image using the projection image acquired by the optical imaging component as a reference. At the same time, the imaging trajectory of the actual contour edge of the surface of the elevator component under inspection on the image plane is extracted from the same projection image. The theoretical imaging position and the imaging trajectory are spatially correlated to obtain the set of image plane offset vectors between the theoretical position of the virtual ruler grid at each grid node and the actual contour of the surface of the elevator component under inspection. The offset vectors in the image plane offset vector set are meshed and grouped according to the vertical and horizontal order of the virtual scale grid. The offset vectors corresponding to adjacent grid nodes in the same vertical order form a horizontal offset vector sequence, and the offset vectors corresponding to adjacent vertical orders in the same horizontal order form a vertical offset vector sequence. The horizontal offset vector sequence and the vertical offset vector sequence are cross-compared to extract the gradient change of the offset vector at each grid node relative to the offset vectors of adjacent nodes, thus obtaining an offset vector field carrying local gradient characteristics. The offset vectors at each grid node in the offset vector field carrying local gradient features are compared point by point with the preset allowable offset threshold. The components of the offset vectors at each grid node that exceed the allowable offset threshold are clustered according to the surface area of ​​the elevator component under inspection to which they belong. The ratio of the number of grid nodes exceeding the threshold in each region after clustering to the total number of grid nodes in the region is used as the quantitative value of the deviation of the region, thus obtaining the quantitative value of the deviation of each region on the surface of the inspected elevator component relative to the design shape.

2. The elevator inspection method based on augmented reality technology as described in claim 1, characterized in that, The process of clustering components in the offset vector at each grid node that exceed the allowable offset threshold according to their respective surface regions of the inspected elevator component includes the following steps: From the offset vector field carrying local gradient features, filter the values ​​of each component of the offset vector at each grid node; mark the grid node whose component value exceeds the allowable offset threshold as an out-of-tolerance node, and extract the vertical sequence index and horizontal sequence index corresponding to each out-of-tolerance node in the virtual scale grid; use the two-dimensional index pair composed of the vertical sequence index and the horizontal sequence index as elements to establish an out-of-tolerance node index set. Arrange the two-dimensional index pairs in the out-of-tolerance node index set in ascending order of vertical hierarchical index and horizontal sequential index; traverse adjacent index pairs in turn, and when the difference between the vertical hierarchical index and the difference between the horizontal sequential index of two adjacent index pairs is not greater than the preset connectivity step size; group the out-of-tolerance nodes corresponding to the two adjacent index pairs into the same connectivity group. After traversal, a set of out-of-tolerance node connectivity groups consisting of several connectivity groups is obtained. The spatial inclusion relationship between the grid node positions corresponding to the out-of-tolerance nodes in each connected group of the out-of-tolerance node connected group set and the imaging trajectory of the actual contour edge of the surface of the inspected elevator component on the image plane is determined. Using the closed region enclosed by the imaging trajectory as the boundary, the connected groups that cross different closed regions are divided into subgroups; the subgroups are used as the clustering results of out-of-tolerance nodes based on the geometric region of the surface of the inspected elevator component.

3. The elevator inspection method based on augmented reality technology as described in claim 2, characterized in that, The process of determining the spatial inclusion relationship between the actual contour edge of the inspected elevator component and the imaging trajectory on the image plane includes the following steps: The positions of the grid nodes corresponding to the out-of-tolerance nodes in each connected group of the out-of-tolerance node connected group set are mapped from the node index of the virtual scale grid to the pixel coordinates on the image plane. The closed contour boundary contained in the imaging trajectory of the actual contour edge of the inspected elevator component surface on the image plane is extracted. The closed contour boundary is discretized into a sequence of boundary points arranged in order, and each sequence of boundary points is divided into several independent closed regions according to the geometric connectivity of the imaging trajectory. Starting from the pixel coordinates of each out-of-tolerance node in each connected group, a ray is drawn in a preset fixed direction in the image plane. The intersection points of the ray with the adjacent boundary points in the boundary point sequence of each closed region are calculated in turn, and the number of intersection points of the ray with the boundary line of each closed region is counted. The parity of the number of intersections determines whether the out-of-tolerance node is located inside or outside the closed region, and the attribution label of each out-of-tolerance node to each closed region is obtained. Out-of-tolerance nodes with the same closed region attribution label within the same connected group are grouped into subgroups. When there are multiple different attribution labels within a connected group, the connected group spanning different closed regions is divided into several subgroups according to the actual position of the closed region boundary on the image plane, so that all out-of-tolerance nodes in each subgroup are located within the same closed region. The subgroups after division are used as the clustering results of out-of-tolerance nodes divided according to the geometric region of the surface of the inspected elevator component.

4. The elevator inspection method based on augmented reality technology as described in claim 3, characterized in that, The process of dividing a connected group that spans different closed regions into several subgroups along the boundary lines includes the following steps: From the set of connected groups with out-of-tolerance nodes, select connected groups containing multiple different closed region attribution labels as the groups to be segmented, and extract the pixel coordinates of all out-of-tolerance nodes in the groups to be segmented and their corresponding attribution labels. Simultaneously, extract the polygon vertex set formed by the boundary point sequence of each closed region to obtain the correspondence between the group to be segmented and the boundary polygon of each closed region; Using the pixel coordinates of the out-of-tolerance nodes in the group to be segmented as graph nodes and the topological adjacency relationship between adjacent grid nodes in the virtual scale grid as graph edges, construct the graph structure corresponding to the group to be segmented; convert the boundary polygons of each closed region into closed regions, traverse each graph edge in the graph structure, and mark the graph edge as a cross-region connection edge when the two graph nodes connected by the graph edge belong to different closed regions in the belonging label. Remove all graph edges marked as cross-region connection edges from the graph structure corresponding to the group to be segmented, so that the original graph structure is decomposed into several unconnected subgraphs; The out-of-tolerance nodes corresponding to the graph nodes in each subgraph are grouped into a subgroup, so that all out-of-tolerance nodes in each subgroup are located in the same closed region, resulting in the segmented subgroup set.

5. The elevator inspection method based on augmented reality technology as described in claim 4, characterized in that, The process of constructing the graph structure corresponding to the groups to be segmented includes the following steps: Extract the vertical hierarchical index and horizontal sequence index of each out-of-tolerance node in the virtual scale grid from the group to be segmented. Based on the vertical hierarchical index and horizontal sequence index, match each out-of-tolerance node with the preset grid node index system in the virtual scale grid to obtain the node positioning parameters of each out-of-tolerance node in the virtual scale grid topology. Based on the topological adjacency relationship between adjacent grid nodes in the virtual scale grid, out-of-tolerance node pairs with direct adjacency relationships are selected from the node positioning parameters; Using the pixel coordinates of the two out-of-tolerance nodes in each out-of-tolerance node pair as the two vertices of the graph structure, and the topological adjacency relationship of the out-of-tolerance node pair as the graph edge, a graph edge set is generated. The pixel coordinates of all out-of-tolerance nodes in the group to be segmented are used as all vertices of the graph structure, and all edges in the graph edge set are used as connections between vertices to form the graph structure corresponding to the group to be segmented.

6. The elevator inspection method based on augmented reality technology as described in claim 5, characterized in that, The process of filtering out out-of-tolerance node pairs with direct adjacency from node location parameters includes the following steps: Based on the topological adjacency relationship between adjacent grid nodes in the virtual scale grid, the judgment condition for direct adjacency is set as follows: the absolute difference of the vertical sequence index of two grid nodes is equal to the preset neighborhood step size and the horizontal sequence index is equal, or the vertical sequence index is equal and the absolute difference of the horizontal sequence index is equal to the preset neighborhood step size. The judgment condition is used as the adjacency relationship benchmark for screening out-of-tolerance node pairs. Extract the vertical hierarchical index and horizontal sequence index of each out-of-tolerance node from the node location parameters. Sort all out-of-tolerance nodes according to the vertical hierarchical index as the primary key and the horizontal sequence index as the secondary key. Use the sorted sequence as the traversal order. Take each out-of-tolerance node as the current node in turn. Based on the adjacency relationship benchmark, search for the existence of node location parameters of another out-of-tolerance node at the position where the vertical hierarchical index of the current node is increased or decreased by a preset neighborhood step size and the horizontal sequence index is equal, and at the position where the vertical hierarchical index is equal and the horizontal sequence index is increased or decreased by a preset neighborhood step size. Pair the current node with the search result with the searched out-of-tolerance node to obtain a set of candidate adjacency pairs. Merge duplicate pairs of the same out-of-tolerance node in the candidate adjacency pair set, and store them uniquely with the smaller of the two node indices in the out-of-tolerance pair as the preorder and the larger as the postorder, to obtain a set of out-of-tolerance node pairs with direct adjacency relationships.

7. The elevator inspection method based on augmented reality technology as described in claim 1, characterized in that, A continuous scanning beam is emitted to multiple directional reflection units arranged longitudinally at intervals on the inner wall of the elevator shaft. The beams returned by each directional reflection unit are received and their phase lag is analyzed. The phase lag is converted into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units, and a longitudinal cross-sectional profile lattice composed of spatial coordinate values ​​is obtained.

8. The elevator inspection method based on augmented reality technology as described in claim 7, characterized in that, The longitudinal section contour dot matrix is ​​spatiotemporally synchronized with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are then mapped to the observation coordinate system of the inspection device through coordinate transformation, resulting in a set of virtual scale meshes bound to the current observation direction of the inspection device.

9. An elevator inspection system based on augmented reality technology, used to implement the elevator inspection method based on augmented reality technology as described in any one of claims 1 to 8, characterized in that, include: The hysteresis conversion module is used to emit continuous scanning beams to multiple directional reflection units arranged longitudinally on the inner wall of the elevator shaft, receive the beams returned by each directional reflection unit and analyze their phase hysteresis, and convert each phase hysteresis into spatial coordinate values ​​according to the spatial arrangement order of the directional reflection units to obtain a longitudinal cross-sectional profile dot matrix composed of spatial coordinate values. The spatiotemporal synchronization module is used to spatiotemporally synchronize the longitudinal section contour dot matrix with the real-time spatial attitude parameters output by the orientation sensing component that moves with the inspection device. The synchronized real-time spatial attitude parameters are mapped to the observation coordinate system of the inspection device through coordinate transformation to obtain a set of virtual scale grids bound to the current observation direction of the inspection device. The offset vector comparison module is used to project a virtual ruler grid onto the surface of the elevator component under inspection through an optical imaging component, collect the offset vector between the projected image of the virtual ruler grid on the component surface and the actual contour of the elevator component under inspection, and compare the offset vector with a preset allowable offset threshold point by point to obtain the quantitative value of the degree of deviation of each area on the surface of the elevator component under inspection from the design shape.

Images