Full-scene scanning device and three-dimensional scanning modeling system

Abstract

The invention discloses a full-scene scanning device and a three-dimensional scanning modeling system, and the three-dimensional scanning modeling system comprises the following steps: S1, deployment and initialization; s2, data acquisition and transmission; s3, key technology processing; and S4, constructing and perfecting the model. The full-scene scanning device comprises a sliding frame, two air cylinders are fixedly connected to the sliding frame, the output ends of the two air cylinders are fixedly connected with supporting plates respectively, a mounting frame is rotationally connected between the two supporting plates, and remote communication relay equipment is fixedly mounted on the mounting frame. A panoramic scanning device body is rotationally connected to the upper portion of the mounting frame. Stable backhaul of scanning data is ensured by utilizing technologies such as dynamic relay and the like, and the signal problem caused by complex terrains and electromagnetic interference is effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of scanning modeling technology, and in particular to a full-scene scanning device and a three-dimensional scanning modeling system. Background Art

[0002] As the wave of digitalization sweeps across the world, many fields such as geographic information, architectural design, and industrial manufacturing are undergoing profound changes. The precise grasp of spatial information has become a key factor in promoting industry development and enhancing competitiveness.

[0003] The full-scene scanning device and three-dimensional scanning modeling system disclosed in this patent number "CN207037750U" obtain image data by scanning through the full-scene scanning device, and convert the image data into a three-dimensional point cloud, and convert the three-dimensional point cloud into a two-dimensional plane map, and first perform coarse alignment on the three-dimensional point cloud and the two-dimensional plane map, and then use the ICP algorithm to perform fine alignment and fusion to obtain a three-dimensional model, which can achieve fast and accurate modeling.

[0004] However, in actual use, in cities, buildings are dense, the environment is changeable, and electromagnetic interference is strong. Scanning is easily obstructed, resulting in data loss, which affects the accuracy of the model. At the same time, data transmission is unstable, making it difficult to meet real-time modeling needs. In mountainous areas, the terrain is undulating, the vegetation is dense, the signal is attenuated, the collected data is incomplete, and relay transmission is difficult to guarantee, which affects the construction of high-precision models and restricts the development of related work. In industrial plants, there are large mechanical vibrations and electromagnetic interference, the quality of collected data is low, and production has high requirements for real-time data transmission. Once the transmission is stuck, the model cannot serve actual production.

[0005] Based on this, the present application proposes a full-scene scanning device and a three-dimensional scanning modeling system. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a full-scene scanning device and a three-dimensional scanning modeling system.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A three-dimensional scanning modeling system comprises the following steps:

[0009] S1. Deployment and initialization: Place the panoramic scanning device at the beginning of the target area, adjust its angle and the position of the drone, start all equipment to complete initialization and establish communication connection;

[0010] S2. Data collection and transmission: The panoramic scanning device and the drone scan the target area from the ground and air respectively to obtain point cloud data and image data. The drone transmits the data to the panoramic scanning device, which is then sent to the data processing center by the remote communication relay equipment.

[0011] S3. Key technical processing:

[0012] Model optimization and update: by collecting data again, the new data and the old model are roughly aligned to find the corresponding relationship, and then finely aligned and adjusted until convergence, and the new building or renovation information is updated;

[0013] Dynamic relay selection: When the signal is poor, the relay device evaluates the relay nodes based on the terrain, node distribution and real-time signal data, and selects the optimal path to ensure stable transmission;

[0014] Interference suppression modulation: The relay equipment analyzes and filters out interference signals, and automatically adjusts modulation parameters according to the signal quality after interference suppression to ensure accurate and efficient data transmission;

[0015] S4. Model construction and improvement: The data center integrates data to build a model. After the quality is improved through optimization algorithms, it is verified and evaluated based on actual applications. The model is updated and maintained as needed in the application to ensure its timeliness and reliability.

[0016] The present invention also provides a full-scene scanning device, including a sliding frame, two cylinders are fixedly connected to the sliding frame, two cylinder output ends are respectively fixedly connected to support plates, a mounting frame is rotatably connected between the two support plates, a remote communication relay device is fixedly installed on the mounting frame, a panoramic scanning device body is rotatably connected above the mounting frame, a drone is provided on the panoramic scanning device body, and a narrowband communication module is fixedly installed on the drone.

[0017] Preferably, two rollers are rotatably connected to both sides of the sliding frame.

[0018] Preferably, a stop slot is provided on the panoramic scanning device body, and two operating rods are fixedly mounted on the panoramic scanning device body.

[0019] Preferably, the drone is parked on a parking slot, and the drone establishes a data communication connection with the panoramic scanning device body through a narrowband communication module.

[0020] Preferably, a laser emitter and an image sensor are disposed in the panoramic scanning device body.

[0021] The present invention has the following beneficial effects:

[0022] 1. Accurately construct three-dimensional models by using model optimization and updating technology, effectively align and integrate newly collected data with existing models, meticulously display architectural details, and assist in the development of work in related urban fields; with the help of optimized data collection and transmission technology, ensure efficient data flow to meet the needs of real-time modeling; rely on a flexible model update mechanism to timely integrate new information according to urban changes, so that the model is always in line with reality and provide strong support for urban management and other decision-making.

[0023] 2. By utilizing technologies such as dynamic relay, the scanning data can be stably transmitted back and the signal problems caused by complex terrain and electromagnetic interference can be effectively overcome. The high-precision model constructed can clearly present the landform and resource information, greatly assisting resource exploration work, while realizing remote control and real-time monitoring, reducing the operational risks of personnel and improving the safety and efficiency of the overall operation.

[0024] 3. Create accurate models through multi-source data fusion to provide accurate guidance for equipment maintenance and factory layout optimization; have strong anti-interference capabilities to ensure reliable and real-time data transmission, facilitate real-time control of factory dynamics, and achieve accurate maintenance of equipment based on the model, extend equipment service life, and enhance the overall competitiveness of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a schematic diagram of the overall structure of a full-scene scanning device proposed by the present invention;

[0026] Figure 2 This is a schematic diagram of the connection structure of an operating lever, a stop slot and a body of a panoramic scanning device of a full-scene scanning device proposed by the present invention;

[0027] Figure 3 This is a structural schematic diagram of a remote communication relay device for a full-scene scanning device proposed by the present invention;

[0028] Figure 4 This is a schematic diagram of the connection structure of the sliding frame, rollers, cylinders and other components of a full-scene scanning device proposed by the present invention;

[0029] Figure 5 This is a schematic diagram of the connection structure between a narrowband communication module of a full-scene scanning device and a drone proposed by the present invention;

[0030] Figure 6 A diagram showing part of the codes used for model optimization and updating in Embodiment 1 of a three-dimensional scanning modeling system proposed by the present invention;

[0031] Figure 7 A diagram showing part of the codes used for dynamic relay selection in Embodiment 2 of a three-dimensional scanning modeling system proposed by the present invention;

[0032] Figure 8This is a diagram showing part of the codes used for interference suppression in Embodiment 3 of a three-dimensional scanning modeling system proposed by the present invention.

[0033] In the figure: 1 panoramic scanning device body, 2 drone, 3 mounting frame, 4 support plate, 5 cylinder, 6 roller, 7 sliding frame, 8 stop slot, 9 operating lever, 10 remote communication relay equipment, 11 narrowband communication module. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0035] A three-dimensional scanning modeling system comprises the following steps:

[0036] S1. Deployment and initialization:

[0037] Before starting the full-scene scanning operation, the operator needs to select the starting point for the panoramic scanning device according to the topography, range size and expected scanning focus of the target area. Use a high-precision angle measuring instrument to carefully adjust its horizontal and vertical angles to ensure accurate coverage of the scanning field of view. At the same time, control UAV 2 to take off to the predetermined airspace. The height of this airspace is set according to the height characteristics of the target area. It is necessary to avoid obstacles and ensure the clarity of image acquisition. Generally, the urban area is 100-300 meters, and the mountainous area is flexibly adjusted to 500-1000 meters according to the height of the mountain. After the hardware is in place, start the panoramic scanning device, UAV 2 and various supporting sensors, communication modules and other equipment, perform self-test and initialization, establish a stable communication connection between each other, and lay a solid foundation for subsequent data collection.

[0038] S2. Data collection and transmission:

[0039] This stage is the key link for obtaining the original material. The panoramic scanning device uses multi-angle laser radar, high-definition camera and other equipment on the ground to perform a 360-degree full-range scan of the surrounding environment at a fixed frequency, capturing detailed point cloud data and clear images of building facades, ground facilities, etc. In the air, UAV 2, with its flexible maneuverability, is equipped with a special telephoto lens camera and a lightweight laser radar to supplement the scanning blind area of ​​the panoramic scanning device, the top of high-rise buildings and other areas. High-speed short-range wireless communication technology, such as Wi-Fi6 or a digital radio station with a dedicated frequency band, is used between UAV 2 and the panoramic scanning device to ensure real-time data transmission and avoid data loss or dislocation due to delays. After the panoramic scanning device summarizes the data collected by itself and transmitted by UAV 2, it sends the massive data to the data processing center through the remote communication relay device 10, using the 4G / 5G network or satellite communication link (remote mountainous areas). The relay device has intelligent frequency selection and signal enhancement functions, and automatically switches to the optimal communication frequency band according to the electromagnetic interference conditions of the environment to ensure the smoothness of data transmission.

[0040] S3. Key technical processing:

[0041] Model optimization and update: When a new building is completed or an existing building is renovated in the target area, the model optimization and update process needs to be started. UAV 2 and the panoramic scanning device are dispatched again to collect data in the changed area. After the newly collected data is imported into the data processing center, a rough alignment is performed first. By calculating the offset between the centroid of the new data point cloud and the centroid of the corresponding area of ​​the old model, a preliminary translation alignment is performed to quickly find the approximate correspondence. Then the iterative closest point algorithm (ICP) is used for precise alignment, and the coordinates of the new data are repeatedly adjusted until the average distance error with the old model converges to a minimum value, accurately integrating the new building or renovation information, so that the model keeps pace with the times.

[0042] Dynamic relay selection: In complex terrain or areas where signals are susceptible to interference, such as mountain canyons and urban high-rise canyon areas, the risk of data transmission being blocked increases greatly. At this time, the relay device starts the dynamic relay selection mechanism. It collects surrounding terrain data in real time, builds a digital terrain model, and combines the location distribution of deployed relay nodes and real-time signal strength, signal-to-noise ratio and other data to perform a comprehensive evaluation using a preset cost function. The cost function considers factors such as signal attenuation, transmission distance, and interference intensity, and scores each relay node through calculation, selecting the optimal relay path with the lowest score to ensure stable and reliable data transmission and avoid data collection failures due to signal interruptions.

[0043] Interference suppression modulation: The relay device has a built-in intelligent interference analysis module, which performs real-time spectrum analysis on the received signal and accurately identifies various interference signals, such as electromagnetic noise in industrial plants and co-frequency interference in cities. Adaptive filtering algorithms, such as the least mean square algorithm (LMS), are used to filter out interference components and restore pure signals. Subsequently, the modulation parameters are automatically adjusted according to the signal quality after interference suppression. If the signal quality is good, the modulation order is increased, such as switching from QPSK to 64QAM, to improve data transmission efficiency; if the signal quality is poor, the order is reduced to ensure transmission accuracy, so as to achieve accurate and efficient data transmission.

[0044] S4. Model construction and improvement:

[0045] After receiving massive amounts of data from the front end, the data processing center uses advanced point cloud fusion algorithms to deeply fuse point cloud data and image data collected from different perspectives and different devices to construct an initial three-dimensional model. Then, optimization algorithms such as the Laplace smoothing algorithm and the Poisson reconstruction algorithm are used to smooth the model surface, fill holes, and improve model quality. Afterwards, based on actual application scenarios, such as urban planning, geological exploration, and industrial production management, professionals are invited to conduct verification and evaluation, and score the model from multiple dimensions such as accuracy, completeness, and practicality. In the subsequent actual application process, continue to pay attention to changes in the target area, and update and maintain the model regularly or in real time as needed.

[0046] The present invention also provides a full-scene scanning device, including a sliding frame 7, two rollers 6 are rotatably connected on both sides of the sliding frame, and the rollers 6 can make the device move conveniently on a plane, and facilitate rapid position adjustment to meet different operation starting point requirements. Two cylinders 5 are fixedly connected to the sliding frame 7, and the output ends of the two cylinders are fixedly connected to support plates 4. The height position of the support plates 4 can be adjusted by the extension and contraction of the cylinders 5, thereby flexibly changing the height of the upper equipment from the ground to adapt to different terrains or scanning height requirements.

[0047] A mounting frame 3 is rotatably connected between the two support plates 4. The mounting frame 3 is used to carry key equipment, and a remote communication relay device 10 is fixedly installed on it. The remote communication relay device 10 undertakes the task of transferring long-distance data transmission, relaying the collected data to the data processing center, and ensuring the continuity of data transmission, especially overcoming problems such as signal attenuation in complex environments. A panoramic scanning device body 1 is rotatably connected above the mounting frame 3. The panoramic scanning device body 1 is the core data acquisition unit, and a laser transmitter and an image sensor are arranged inside. The laser transmitter obtains point cloud data of the target area by emitting a laser beam to accurately outline the contour of the object, and the image sensor is responsible for collecting clear image information. The two work together to capture details of the surrounding environment from the ground in all directions.

[0048] The panoramic scanning device body 1 is provided with a drone 2. The drone 2, as a supplementary force for data collection in the air, can break through the field of view limitations of ground scanning and collect data from the top of high-rise buildings, deep valleys and other areas. A narrowband communication module 11 is fixedly installed on the drone 2. The module is used for the drone 2 to establish a data communication connection with the panoramic scanning device body 1, and transmit data quickly and stably in a short distance in a narrowband communication manner to avoid data loss.

[0049] The panoramic scanning device body 1 is provided with a parking slot 8, which is specially designed for the drone 2 and is the parking position of the drone 2 to ensure the storage and protection of the drone 2 during non-operating periods. Two operating rods 9 are fixedly installed on the panoramic scanning device body 1. The operator pushes, pulls or rotates the entire device by holding the operating rods 9 to achieve manual control, which is convenient for precise force in the aspects of starting the equipment and fine-tuning the position. The drone 2 is parked on the parking slot 8, and the drone 2 establishes a data communication connection with the panoramic scanning device body 1 through the narrowband communication module 11.

[0050] Example 1: Urban building area

[0051] Step 1: Deployment and initialization

[0052] Transport the panoramic scanning device body 1 to a suitable starting point, adjust its position by the roller 6 on the sliding frame 7, adjust the angle and height of the panoramic scanning device body 1 with the operating lever 9, check the status of the drone 2 in the parking slot 8, and after ensuring that the equipment is normal, turn on the panoramic scanning device body 1 and the drone 2, initialize the relevant components of the two, and at the same time, establish a communication connection between the drone 2 and the panoramic scanning device body 1 to prepare for data collection.

[0053] Step 2: Data Collection and Transmission

[0054] The panoramic scanning device body 1 scans the building from the ground to obtain three-dimensional point cloud data and image data. The drone 2 takes off to perform additional scanning on the high places and hard-to-reach areas of the building, and also obtains point cloud data and image data. These data are transmitted to the panoramic scanning device body 1 through the narrowband communication module 11, and then sent to the data processing center through the remote communication relay device 10. In the data processing center, feature points are extracted from the collected image data, and the three-dimensional coordinates of the feature points are calculated based on the pixel coordinates of the feature points in different images and the relationship between the camera's internal and external parameters, using the triangulation principle combined with the laser scanning depth information, thereby constructing an initial three-dimensional model. For example, if the pixel coordinates of the feature points in two images and the camera parameters are known, their three-dimensional coordinates can be calculated through a specific formula, and then the initial model can be constructed.

[0055] Step 3: Key technical processing

[0056] Model optimization and update: If there are changes in urban buildings, the device is started again to scan and obtain new data. The new data is roughly aligned with the existing model, and a matching algorithm based on geometric features is used to minimize a specific objective function (such as where p i ′ is a new data point, p i is the old model point, T c is a rough registration transformation matrix) and an iterative optimization algorithm (such as gradient descent) is used to solve T c Then, we perform precise registration, constantly find the closest point of the new point cloud in the old model, and use the singular value decomposition method to solve the transformation matrix T that minimizes the distance between the two. f , repeat until the convergence condition is met (such as the change in the transformation matrix is ​​less than a threshold). For new buildings or renovated parts, the geometry and texture information is constructed based on the new data and added to the old model to obtain an updated model.

[0057] Step 4: Model construction and improvement

[0058] The root mean square error ( p mi is the coordinate of the model point, p ti is the actual building measurement coordinate, N is the number of verification points) and compares the updated model with the actual building measurement data. If the RMSE is less than the set threshold, the model accuracy meets the standard and can be used in urban planning, navigation and other fields.

[0059] Example 2: Complex environment in mountainous areas

[0060] Step 1: Deployment and initialization

[0061] The scanning device is brought to a suitable location in the mountainous area, and the power of the panoramic scanning device body (1) and the drone (1) is turned on to initialize them. At the same time, the remote communication relay device (10) searches for signals and establishes a preliminary communication link with surrounding communication base stations or satellites.

[0062] Step 2: Data Collection and Transmission

[0063] The drone 2 flies along a preset path, and the panoramic scanning device body 1 also scans the surroundings. The two collect three-dimensional point cloud data and image data of the mountain area, transmit them to the panoramic scanning device body 1, and then send them to the data processing center via the remote communication relay device 10. During the process, the remote communication relay device 10 monitors the signal strength, signal-to-noise ratio and other parameters between each device and external communication node in real time.

[0064] Step 3: Key technical processing

[0065] Dynamic relay selection: If the signal quality decreases or there is a risk of interruption, the remote communication relay device (10) starts the dynamic relay selection algorithm. According to the mountain terrain, communication node distribution and real-time monitoring signal data, a cost function (such as α, β, γ are weight coefficients, s n 、snr n The link signal strength and signal-to-noise ratio of relay node n are used to evaluate the advantages and disadvantages of the optional relay nodes, and C(n) is selected to make the smallest relay node to ensure stable data transmission.

[0066] Interference suppression and adaptive modulation: When electromagnetic interference in mountainous areas affects communication, the remote communication relay device 10 uses a built-in interference detection module to analyze the interference components, and uses a deep learning-based interference suppression algorithm to identify and filter out interference signals (such as the trained model D to achieve S′=D(S,J), S′ is the signal after interference suppression, S is the original signal, and J is the interference component). Then, according to the signal-to-noise ratio after interference suppression, according to the set rules (such as Use m i 、r i Modulation coding mode, otherwise use m i -1, r i -1) Automatically adjust the parameters of the adaptive modulation module to ensure data transmission accuracy and efficiency.

[0067] Step 4: Model construction and improvement

[0068] The data processing center fuses the data collected by the drone 2 and the panoramic scanning device 1, processes the data using an optimization algorithm, and constructs a high-precision three-dimensional model of the mountain area. The quality of the model is evaluated by calculating the integrity index and accuracy index of the model. The integrity index can be calculated based on the proportion of the actual area covered by the model, and the accuracy index can be calculated by comparing with the actual measurement data similar to the urban building model.

[0069] Example 3: Industrial Plant

[0070] Step 1: Deployment and initialization

[0071] A scanning plan is formulated according to the layout and scale of the factory area, the location of the scanning device and the flight path of the drone 2 are determined, the device is moved to the starting point, the angle of the panoramic scanning device body 1 is adjusted, the panoramic scanning device body 1 and the drone 2 are started, and preparations are made to collect the three-dimensional point cloud data and image data of the factory area.

[0072] Step 2: Data Collection and Transmission

[0073] The drone 2 and the panoramic scanning device body 1 collect factory data. The drone 2 data is transmitted to the panoramic scanning device body 1 via the narrowband communication module 11 and then sent by the remote communication relay device 10. During the process, the remote communication relay device 10 monitors the signal strength, interference and load conditions of each communication node in the factory communication environment in real time.

[0074] Step 3: Key technical processing

[0075] Dynamic relay selection: When data transmission is affected, the telecommunication relay device 10 initiates a dynamic relay selection algorithm.

[0076] By calculating the cost function (such as α, β, γ are weight coefficients, s n is the signal strength, i n is the interference intensity, l n The relay node is evaluated based on the load situation and the best relay path is selected. If it is found that the communication base station load is high and the transmission delay increases, it will switch to a relay node with light load and good signal to ensure fast and stable data transmission.

[0077] Interference suppression and model building: In response to electromagnetic interference, the remote communication relay device 10 uses interference suppression technology to pre-process the signal (such as using a filtering algorithm F to achieve S′=F(S,I nosise ), S′ is the signal after interference suppression, S is the original signal, I noise In the data processing center, the collected data is fused to build a high-precision 3D model of the industrial plant using a 3D modeling algorithm based on multi-source data fusion, and a point cloud registration algorithm (such as by minimizing are the points of two point cloud data sets, T is the registration transformation matrix, and an iterative optimization algorithm is used to solve T) to improve the model quality by denoising, registering, and splicing data.

[0078] Step 4: Model construction and improvement

[0079] The constructed 3D model is used for equipment maintenance, layout optimization, safety monitoring, etc. of industrial plants. The model is updated and maintained regularly, and the above scanning and modeling steps are repeated according to the actual changes in the plant to ensure that the model always reflects the actual status of the plant, providing reliable 3D information for all aspects of industrial production, improving plant management efficiency and safety, and reducing operating costs.

[0080] It should be noted that, in the comparative examples, comparative example 1 relies only on ground scanning equipment, lacks other equipment assistance, and data collection is not comprehensive, resulting in a relatively small amount of transmitted data but serious packet loss, poor model construction accuracy, and a large error with the actual scene; in comparative example 2, there is no interference suppression measure, and electromagnetic interference has a significant impact on communication quality, which increases the packet loss rate and transmission delay, thereby affecting the model accuracy, and the error also increases accordingly. The test data parameters of each group of embodiments and comparative examples are shown in Table 1:

[0081] Table 1: Comparison of test parameters of various embodiments and comparative examples

[0082]

[0083] In summary, the three-dimensional scanning modeling system proposed by the present invention, from the perspective of data transmission, uses optimization technology in embodiments one, two, and three to control the packet loss rate to ≤5%, ≤8%, and ≤6%, respectively, which can better ensure data integrity. Comparative Example 1 only uses ground scanning, with a packet loss rate of ≥18%; Comparative Example 2 has no interference suppression, ≥12%, and a large amount of data is lost. At the same time, in terms of transmission efficiency, Example 1 is ≤100ms, ≤150ms, and ≤120ms, which basically meet the real-time requirements of the corresponding scene. Due to poor equipment and transmission, the delay of Example 1 is ≥280ms; Comparative Example 2 is interfered with, ≥220ms, and it is difficult to meet real-time decision-making. From the perspective of model accuracy, the error of the urban building model in Example 1 is ≤5cm, the error of the mountain model in Example 2 is ≤8cm, and the error of the industrial plant model in Example 3 is ≤6cm, which can accurately reflect the scene. For comparative example 1, the error is ≥18cm, and the data collection is one-sided; for comparative example 2, the error is ≥10cm, the data is inaccurate due to interference, and the model accuracy is poor. Therefore, compared with the two comparative examples, the present invention has outstanding advantages and adaptability in different scenarios, and can provide strong support for practical work.

[0084] Specifically, in each embodiment, key code examples and corresponding analysis are written in Python language.

[0085] Furthermore, if Figure 6 As shown, in Example 1, a rough alignment is performed first. By calculating the centroid of the new and old point clouds, a translation vector is obtained to align the new point cloud, laying the foundation for the follow-up. Then, the ICP algorithm is used for fine alignment, and KDTree is used to find corresponding points. After multiple iterations of optimizing the transformation matrix, the position of the new point cloud is continuously adjusted until the average distance is less than the set tolerance to achieve precise alignment. Finally, the new point cloud after fine alignment is merged with the old model to complete the update. The overall process is clear, but in actual applications, details such as texture updates need to be improved, and more complex situations need to be considered to ensure that the model update can accurately reflect changes in urban buildings and help the development of related applications.

[0086] Furthermore, if Figure 7 As shown, in Example 2, by first defining the signal strength, signal-to-noise ratio, distance parameter and weight coefficient of each relay node, the weight can be adjusted as needed. Then, through a custom cost function, these factors are comprehensively considered to calculate the cost of each node. In the function, weak signal, low signal-to-noise ratio and long distance will increase the cost. Finally, np.argmin is used to find the index with the lowest cost to determine the best relay node. In practical applications, links can be configured in this way to ensure stable data transmission in mountainous areas, but it is necessary to combine the real environment and refine the parameters and functions to meet complex communication conditions.

[0087] Furthermore, if Figure 8As shown, in Example 3, the interference suppression part uses a mean filter function to process the simulated interference signal. Although it is simple, it can smooth the signal to a certain extent. In the adaptive modulation part, the signal-to-noise ratio threshold and name corresponding to different modulation methods are set, and the signal-to-noise ratio is calculated by simulation, and the appropriate modulation method is selected by comparing with the threshold, so as to balance the transmission efficiency and accuracy. In practice, it is necessary to improve the filtering algorithm, accurately calculate the signal-to-noise ratio, and coordinate with the overall system to better cope with the complex electromagnetic environment of the factory.

[0088] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A three-dimensional scanning modeling system, characterized in that: The following steps are involved: S1. Deployment and initialization: Place the panoramic scanning device at the beginning of the target area, adjust its angle and the position of the drone, start all equipment to complete initialization and establish communication connection; S2. Data collection and transmission: The panoramic scanning device and the drone scan the target area from the ground and air respectively to obtain point cloud data and image data. The drone transmits the data to the panoramic scanning device, which is then sent to the data processing center by the remote communication relay equipment. S3. Key technical processing: Model optimization and update: by collecting data again, the new data and the old model are roughly aligned to find the corresponding relationship, and then finely aligned and adjusted until convergence, and the new building or renovation information is updated; Dynamic relay selection: When the signal is poor, the relay device evaluates the relay nodes based on the terrain, node distribution and real-time signal data, and selects the optimal path to ensure stable transmission; Interference suppression modulation: The relay equipment analyzes and filters out interference signals, and automatically adjusts modulation parameters according to the signal quality after interference suppression to ensure accurate and efficient data transmission; S4. Model construction and improvement: The data center integrates data to build a model. After the quality is improved through optimization algorithms, it is verified and evaluated based on actual applications. The model is updated and maintained as needed in the application to ensure its timeliness and reliability.

2. A full-scene scanning device, characterized in that: The invention comprises a sliding frame (7), two cylinders (5) are fixedly connected to the sliding frame (7), the output ends of the two cylinders (5) are respectively fixedly connected to support plates (4), a mounting frame (3) is rotatably connected between the two support plates (4), a remote communication relay device (10) is fixedly installed on the mounting frame (3), a panoramic scanning device body (1) is rotatably connected above the mounting frame (3), a drone (2) is provided on the panoramic scanning device body (1), and a narrowband communication module (11) is fixedly installed on the drone (2).

3. A full scene scanning device according to claim 2, characterized in that: The two sides of the sliding frame (7) are rotatably connected to two rollers (6).

4. The full-scene scanning device according to claim 1, characterized in that: A stop slot (8) is provided on the panoramic scanning device body (1), and two operating rods (9) are fixedly mounted on the panoramic scanning device body (1).

5. A full scene scanning device according to claim 4, characterized in that: The drone (2) is parked on the parking slot (8), and the drone (2) establishes a data communication connection with the panoramic scanning device body (1) via a narrowband communication module (11).

6. The full scene scanning device according to claim 1, characterized in that: A laser emitter and an image sensor are arranged in the panoramic scanning device body (1).

Images