Mining area fixed point data surveying and mapping acquisition monitoring device

Abstract

The invention relates to the technical field of mining area monitoring, in particular to a mining area fixed-point data surveying, mapping, collecting and monitoring device which comprises a main control module used for coordinating and controlling operation of the device. The system comprises a multi-source data acquisition module, a data fusion processing module, a change detection algorithm module, a communication module, a positioning module, a power supply module, a storage module, an environment sensor module and a mobile terminal interface module. By integrating the satellite remote sensing unit, the unmanned aerial vehicle aerial photography unit and the ground measurement unit, a sky-ground integrated monitoring system is constructed, the limitation of a single data source can be overcome, mining area global coverage and high-frequency data acquisition are realized, and the monitoring efficiency and timeliness are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mining area monitoring, and particularly to a device for surveying, collecting and monitoring fixed-point data in a mining area. Background Art

[0002] At present, the monitoring of ground object changes in mining areas mainly relies on manual inspections, single remote sensing image analysis or traditional surveying methods. Manual inspections are inefficient, limited by terrain complexity and weather conditions, and it is difficult to achieve high-frequency and full-coverage monitoring. Traditional remote sensing technologies are mostly based on single-source satellite data, restricted by problems such as long revisit cycles, insufficient resolution or cloud cover, resulting in poor timeliness and limited accuracy of monitoring results. Moreover, existing monitoring systems usually lack the ability to fuse multi-source data and cannot effectively integrate satellite, drone and ground measurement data, making it difficult to meet the accurate identification requirements of dynamic changes in mining areas.

[0003] Existing technologies rely on a single remote sensing data source, which is easily restricted by weather and sensor performance, leading to monitoring blind spots. At the same time, the detection of ground object changes mostly relies on manual visual interpretation, with low efficiency and strong subjectivity, making it difficult to adapt to large-scale mining area inspections. Moreover, the data of satellites, drones and ground equipment are processed independently, lacking a unified platform for multi-source data fusion and intelligent analysis, thus affecting the integrity and reliability of monitoring results. Traditional methods cannot achieve high-frequency data updates and rapid warnings, and it is difficult to detect illegal land occupation or environmental anomalies in a timely manner.

[0004] Therefore, in view of the above problems, the present invention proposes a device for surveying, collecting and monitoring fixed-point data in a mining area, which realizes the automatic and high-precision monitoring of ground object changes in the mining area by integrating a "sky-earth" multi-source data acquisition module, a deep learning change detection algorithm and an intelligent early warning system. Summary of the Invention

[0005] In order to overcome the problems of low automation degree and poor system coordination in the existing mining area monitoring technology, the present invention proposes a device for surveying, collecting and monitoring fixed-point data in a mining area.

[0006] The technical solution of the present invention is as follows: The device for surveying, collecting and monitoring fixed-point data in a mining area includes: A main control module for coordinating and controlling the operation of the device; A multi-source data acquisition module integrating a satellite remote sensing receiving unit, a drone aerial photography unit and a ground measurement unit to realize "sky-earth" integrated data acquisition; A data fusion processing module for screening, matching and fusing multi-source remote sensing data to generate high-precision spatially continuous data; A change detection algorithm module for detecting ground object changes in the collected data based on a deep learning model; A communication module for transmitting the collected and processed data to the cloud or the mine area big data management system; A positioning module using a high-precision GPS or Beidou positioning system; A power supply module for providing stable power supply for the device; A storage module for temporarily storing the collected raw data and processing results; An environmental sensor module for monitoring the temperature, humidity and air pressure parameters in the mine area; A mobile terminal interface module for supporting connection with the intelligent inspection APP to realize real-time data interaction and task scheduling.

[0007] Preferably, the satellite remote sensing receiving unit of the multi-source data acquisition module supports the data of the Gaojing-1 satellite with a resolution of 0.5 meters and is compatible with other commercial satellite data. By optimizing the star source combination scheme, this unit can ensure the monitoring continuity through redundant data when a single satellite is affected by weather, and at the same time support data acquisition with a 10-day revisit cycle to adapt to the ground object inspection tasks in the whole mine area or the well field range.

[0008] Preferably, the UAV aerial photography unit is equipped with a high-resolution optical camera and a lidar. After the satellite remote sensing preliminarily identifies the ground object changes, it can conduct centimeter-level refined aerial survey on key areas. This unit obtains the position, occupation area, shape, height, volume and construction progress of the target area through the mapping UAV, providing high-resolution auxiliary verification data for the satellite remote sensing detection results.

[0009] Preferably, the ground measurement unit includes a total station and portable measurement equipment for manual on-site review and precise mapping in the areas restricted by UAV aerial survey. This unit collects the precise spatial attributes of the changing ground objects through mobile measurement equipment and generates a review report in combination with image data, forming a closed-loop verification system for "sky-ground" integrated monitoring.

[0010] Preferably, the data fusion processing module adopts a fusion method combining "wavelet transform + local algorithm + principal component analysis". It performs frequency division processing on the spatio-temporal spectral information of multi-source images through wavelet transform, enhances details using the local variance algorithm, retains low-frequency features with local difference weighting, and then reduces spectral distortion through the PCA transformation, finally generating a high-resolution multi-spectral fusion image.

[0011] Preferably, the change detection algorithm module is based on the U-Net deep learning model, supports the ResNet 18 and ResNet 34 backbone networks, and realizes pixel-level classification through an encoder-decoder structure and skip connections. This module is trained using a pre-constructed 2052-group four-channel sample data set. When the area of the changed patch is greater than 200 pixels, the accuracy rate ≥ 0.7 and the recall rate ≥ 0.9, and it can identify and output the boundaries, types and change intensities of the ground object change patches.

[0012] Preferably, the communication module supports 4G / 5G networks and LoRa wireless transmission technology, adapts to the complex environment of the mining area through multi-mode communication, and uses LoRa to achieve long-distance and low-power data transmission in areas without public network coverage. In the base station coverage area, it switches to the 4G / 5G high-speed network. This module can transmit the change detection results, warning information, and original image data to the cloud or the big data management system in real time.

[0013] Preferably, the warning module automatically triggers hierarchical warnings by analyzing the change detection results in real time and combining preset rules. When detecting changes in illegal ground objects or abnormal environmental parameters, it immediately pushes warning information to the big data management system and synchronizes it to the mobile inspection APP to guide on-site personnel to quickly locate and verify, forming a closed-loop management process of "monitoring - warning - disposal".

[0014] Preferably, the device shell adopts an IP67-level dust and waterproof design, the internal circuit is treated by seismic reinforcement, the shell material is set as a lightweight and high-strength alloy, and the surface is coated with an anti-corrosion coating.

[0015] Preferably, the device is seamlessly docked with the integrated monitoring big data management system of the mining area through a standardized API interface, supports automatic uploading of ground object change patches, environmental parameters, and review evidence chain data. The system can send inspection tasks to the mobile device interface of the device in the opposite direction for intelligent management of task scheduling, data synchronization, and result feedback. At the same time, expansion interfaces are reserved to be compatible with future newly added monitoring devices or data sources.

[0016] Advantages of the present invention: 1. By integrating satellite remote sensing, UAV aerial photography, and ground measurement units, a "sky-ground" integrated monitoring system is constructed, which can overcome the limitations of single data sources, achieve full coverage of the mining area and high-frequency data collection, significantly improve the monitoring efficiency and timeliness, and avoid the blind spots and lag of manual inspections.

[0017] 2. Adopting a change detection algorithm based on the U-Net deep learning model, combined with the ResNet backbone network and a large number of training samples, can accurately identify the change characteristics of typical ground objects such as buildings, soil piles, and water bodies. When the area of the change patch is greater than 200 pixels, the accuracy is not less than 0.7 and the recall rate is not less than 0.9, greatly reducing the false detection and missed detection rates.

[0018] 3. Through the multi-source data fusion technology of "wavelet transform + local algorithm + principal component analysis", the matching problem between different sensors and remote sensing images with different resolutions is effectively solved, generating high-precision and spatially continuous comprehensive monitoring data, providing more comprehensive and reliable information support for mining area management.

[0019] 4. Through the intelligent early warning module and the mobile interface, it is possible to push the early warning information of ground object changes to the big data steward system and the inspection personnel terminal in real time, forming a rapid response mechanism to help management personnel discover and handle illegal land occupation and other behaviors in a timely manner, and improve the management efficiency of the mining area.

[0020] 5. Through automatic data collection, processing and analysis, the workload of manual inspection and data processing is greatly reduced. At the same time, through the standardized interface, it is seamlessly connected with the existing management system, reducing the technical threshold and operation and maintenance costs, and realizing the intelligent and intensive management of mining area monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It shows the schematic diagram of the system architecture of the present invention; Figure 2 It shows the schematic diagram of the working process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] Please refer to Figure 1 , the present invention provides an embodiment: a mining area fixed-point data surveying, collecting and monitoring device, including: A main control module for coordinating and controlling the operation of the device; A multi-source data collection module integrating a satellite remote sensing receiving unit, an unmanned aerial vehicle aerial photography unit and a ground measurement unit to realize "sky-earth" integrated data collection; A data fusion processing module for screening, matching and fusing multi-source remote sensing data to generate high-precision spatially continuous data; A change detection algorithm module for detecting ground object changes in the collected data based on a deep learning model; A communication module for transmitting the collected and processed data to the cloud or the mining area big data steward system; A positioning module using a high-precision GPS or Beidou positioning system; A power supply module for providing stable power supply for the device; A storage module for temporarily storing the collected original data and processing results; An environmental sensor module for monitoring the temperature, humidity and air pressure parameters of the mining area; The mobile interface module supports connection with the intelligent inspection APP to achieve real-time data interaction and task scheduling.

[0024] The satellite remote sensing receiving unit of the multi-source data acquisition module supports the data of the Gaojing-1 satellite with a resolution of 0.5 meters and is compatible with other commercial satellite data. By optimizing the star source combination scheme, this unit can ensure the monitoring continuity through redundant data when a single satellite is affected by weather. At the same time, it supports data acquisition with a 10-day revisit cycle, adapting to the ground feature inspection tasks in the whole mining area or the mine field.

[0025] The UAV aerial photography unit is equipped with a high-resolution optical camera and a lidar. After the satellite remote sensing preliminarily identifies the ground feature changes, it can perform centimeter-level refined aerial survey on key areas. This unit obtains the position, occupied area, shape, height, volume, construction progress and other attribute information of the target area through the mapping UAV, providing high-resolution auxiliary verification data for the satellite remote sensing detection results. It is especially suitable for the precise mapping of key ground features such as buildings and soil piles, thus making up for the deficiencies of satellite images in local details.

[0026] The ground measurement unit includes total stations and portable measurement devices, which are used for manual on-site review and precise mapping in areas where UAV aerial survey is restricted. This unit collects the precise spatial attributes of the changed ground features through mobile measurement devices and generates a review report in combination with image data, thus ensuring the reliability of satellite and UAV data and forming a closed-loop verification system for "sky-ground" integrated monitoring.

[0027] The data fusion and processing module adopts a fusion method combining "wavelet transform + local algorithm + principal component analysis". It processes the spatio-temporal spectral information of multi-source images through wavelet transform frequency division, enhances details using the local variance algorithm, retains low-frequency features with local difference weighting, and then reduces spectral distortion through the PCA transformation. Finally, it generates a high-resolution multi-spectral fusion image. This technology can effectively solve the imaging differences, geometric differences and scale differences of multi-modal remote sensing data, thus improving the spectral and spatial expression accuracy of ground feature targets in the mining area.

[0028] The change detection algorithm module is based on the U-Net deep learning model, supports the ResNet 18 and ResNet 34 backbone networks, and realizes pixel-level classification through an encoder-decoder structure and skip connections. This module is trained using a pre-built 2052-group four-channel sample data set (covering 5 types of ground features such as buildings, soil piles, and water bodies). When the area of the changed patch is greater than 200 pixels, the accuracy rate ≥ 0.7 and the recall rate ≥ 0.9, and it can identify and output the boundaries, types and change intensities of the ground feature change patches, thus providing algorithm support for the monitoring of illegal land occupation behaviors in the mining area.

[0029] The communication module supports 4G / 5G networks and LoRa wireless transmission technology. It adapts to the complex environment in the mining area through multi-mode communication. In areas without public network coverage, LoRa is used to achieve long-distance and low-power data transmission, while in the base station coverage area, it switches to the 4G / 5G high-speed network. This module can transmit the change detection results, warning information, and original image data to the cloud or the big data management system in real time, thus meeting the needs of remote monitoring and emergency response in the mining area.

[0030] The warning module automatically triggers hierarchical warnings by analyzing the change detection results in real time and combining preset rules. When detecting changes in illegal ground objects or abnormal environmental parameters, it immediately pushes warning information to the big data management system and synchronizes it to the mobile inspection APP to guide on-site personnel to quickly locate and verify, forming a closed-loop management process of "monitoring - warning - disposal".

[0031] The device shell adopts an IP67-level dust and waterproof design, and the internal circuit is treated by seismic reinforcement. It can work stably in the temperature range of -20°C to 60°C. The shell material is selected as a lightweight and high-strength alloy, and the surface is coated with an anti-corrosion coating, which can adapt to the harsh environments such as high dust, humidity, and mechanical impact in the mining area, thus ensuring the reliability of the equipment during long-term field operations.

[0032] The device is seamlessly connected to the integrated monitoring big data management system in the mining area through a standardized API interface, supporting the automatic upload of ground object change patches, environmental parameters, and review evidence chain data. The system can send inspection tasks to the mobile device interface of the device in a reverse direction to achieve intelligent management of task scheduling, data synchronization, and result feedback. At the same time, expansion interfaces are reserved to be compatible with future newly added monitoring devices or data sources, thus meeting the information management needs of the entire life cycle of the mining area.

[0033] Furthermore, the multi-source data acquisition module of the present invention adopts a "space-air-ground" integrated architecture, which consists of three parts: a satellite remote sensing receiving unit, an unmanned aerial vehicle aerial photography unit, and a ground measurement unit. The satellite remote sensing receiving unit takes the Gaojing-1 satellite as the core data source, is equipped with a dedicated receiving antenna and a high-performance data processing server, supports the real-time reception and preprocessing of 0.5-meter panchromatic and 2-meter multispectral images. This unit is built with a multi-satellite source compatible interface and can automatically switch to backup satellite data sources such as Beijing-3 and Jilin-1 to ensure data continuity under harsh weather conditions. The system sets an intelligent scheduling algorithm to automatically adjust the data acquisition cycle according to the monitoring requirements of different zones in the mining area.

[0034] The UAV aerial photography unit uses the DJI M300 RTK professional UAV platform, equipped with a Sony α7RIV full-frame camera and a RIEGL VUX-1LR lidar system to form a multi-sensor collaborative acquisition system. This unit is equipped with an independently developed intelligent flight path planning system, which can automatically generate centimeter-level accurate aerial survey routes according to the changed patches detected by satellites, support the synchronous acquisition of oblique photography and laser point clouds. The system is set with an emergency handling mechanism, including functions such as automatic return when the battery is low, autonomous cruise when the communication is interrupted, and emergency avoidance in abnormal weather, thus ensuring the operation safety in complex environments. The collected data is preprocessed in real time through an on-board edge computing node, including POS data fusion, rapid image stitching, and point cloud filtering, thus greatly improving the subsequent processing efficiency.

[0035] The ground measurement unit consists of an intelligent total station measurement system, GNSS mobile measurement equipment, and a portable 3D laser scanner, forming a multi-scale ground verification ability. This unit is equipped with a customized data acquisition terminal, integrating a Huace P5 GNSS receiver and a Leica TS16 total station, supporting RTK centimeter-level positioning and total station millimeter-level measurement. At the same time, a dedicated field data acquisition APP is developed to achieve rapid positioning of changed patches, attribute entry, and image association, support offline operation and automatic data synchronization. For complex areas that are difficult to cover by UAVs, a Faro Focus S350 3D laser scanner is equipped, which can quickly obtain a high-precision surface model, forming an effective complement to aerial data. The three sub-units are data-associated through a unified spatio-temporal reference. All collected data is automatically appended with WGS84 coordinates and UTC timestamps, thus ensuring the seamless fusion of multi-source data.

[0036] Furthermore, the data processing core module of the present invention adopts a multi-level pipeline architecture, mainly including two core components: a data fusion processing module and a change detection algorithm module. The data fusion processing module adopts a three-level progressive processing flow: in the first-level processing, based on the improved SIFT-GMS (Grid-based Motion Statistics) algorithm, automatic matching of multi-source images is realized. By constructing a Gaussian pyramid space and optimizing the feature point screening strategy, the matching efficiency is improved, and the matching accuracy can be better than 1 pixel on 0.5-meter resolution images. In the second-level processing, the adaptive wavelet transform fusion algorithm is applied, and the decomposition layers (3-5 layers) and fusion rules are automatically adjusted according to different ground object types. The local variance maximization criterion is adopted for high-frequency information, and the weighted average strategy is adopted for low-frequency information, effectively retaining the spectral features while enhancing the spatial details. In the third-level processing, the improved PCA (Principal Component Analysis) transform is introduced. Through band grouping and noise suppression processing, the spectral distortion is controlled within 5%. Finally, a 16-bit GeoTIFF format fusion result is generated, including complete metadata information and a quality assessment report.

[0037] The change detection algorithm module is designed based on a hybrid architecture of deep learning. The core adopts an improved U-Net++ network structure. The input layer supports the input of image slices of four channels (R, G, B, NIR) with 512×512 pixels. The sample diversity is expanded through a data augmentation strategy. The encoder part uses the pre-trained ResNet34 as the backbone network and incorporates the CBAM (Convolutional Block Attention Module) attention mechanism to enhance the ability to extract key features. The decoder part adopts a progressive upsampling strategy and combines dense skip connections to achieve multi-level feature fusion, effectively solving the problem of missed detection of small targets. The output layer uses a multi-task learning framework to synchronously output the land cover classification results (5 categories such as buildings, soil piles, water bodies, etc.) and the change intensity map, and introduces CRF (Conditional Random Field) post-processing to optimize the boundary accuracy. The model training adopts a transfer learning strategy.

[0038] In terms of system implementation, the core data processing module is deployed on a high-performance server equipped with NVIDIA Tesla V100 GPUs, encapsulated using Docker containerization, and supports horizontal expansion. At the same time, a dedicated task scheduling engine is developed to achieve automated management of the data processing process, including intelligent functions such as raw data quality inspection (cloud cover < 10%, imaging angle < 25°, etc.), dynamic adjustment of processing priorities (priority for emergency change areas), and automatic restart of abnormal tasks. For typical application scenarios, the system presets three processing modes: fast mode (complete the processing of a 1 km² area within 30 minutes), standard mode (complete the full process within 2 hours), and high-precision mode (complete sub-pixel analysis within 6 hours). Users can flexibly select according to actual needs. All processing results will be automatically connected to the mine area spatio-temporal database, thus establishing a complete data traceability chain to support the backtracking of processing parameters and the evaluation of the quality of results.

[0039] Please refer to Figure 2 , and further, the working process of the present invention will be described: The present invention adopts an intelligent task-driven mode and realizes fully automated operation through a three-level scheduling mechanism. At the data acquisition layer, the system automatically starts the satellite data retrieval program every day at dawn, and queries the transit plans and data availability status of satellites such as Gaojing-1 and Beijing-3 in real time through the API interface. For images that meet the quality requirements (cloud cover rate < 15%, side-sway angle < 20°), the download task is automatically triggered, and the newly acquired images are spatio-temporally associated with the historical database through the metadata parsing engine. At 8:00 am every Monday, the intelligent task planning module automatically generates a detailed inspection plan for the unmanned aerial vehicle (UAV) based on the hot spots of changes detected by the satellite. Considering factors such as airspace restrictions, weather conditions, and equipment status, the most efficient flight route is optimized, and the task package is sent to the designated UAV through the 4G network. On the 1st and 15th of each month, the system starts the ground verification cycle task, automatically divides the key verification areas according to the change detection confidence level, generates a mobile task book containing the navigation path, verification points, and standard operation procedures, and pushes it to the on-site operators.

[0040] At the data processing layer, the system adopts a pipeline operation mode to achieve efficient operation. The newly acquired satellite data first enters the preprocessing pipeline and sequentially completes standardized processing such as radiometric calibration, atmospheric correction, and orthorectification. Then, according to the regional priority, it enters the change detection queue. The UAV data is automatically uploaded to the processing server through a dedicated high-speed transmission channel, and a centimeter-level accurate three-dimensional model is generated through steps such as POS data solution, multi-view image matching, and point cloud classification. The ground measurement data is transmitted back in real time through a secure encryption channel and automatically performs spatial association and attribute matching with the remote sensing data at the corresponding location. When the multi-source data is ready, the intelligent fusion engine is automatically started. First, the spatio-temporal reference is unified, and then a multi-level fusion algorithm is executed, and finally a fusion product with a quality score is generated. The change detection module adopts a dynamic computing resource allocation strategy, enables the high-precision mode (512×512 slices) for key areas, and adopts the efficiency mode (1024×1024 slices) for ordinary areas, and ensures that the full-automatic analysis of a 100 km² area is completed within 8 hours.

[0041] At the achievement application layer, the system establishes a multi-level output system to meet different needs. Basic achievements automatically enter the mine area spatio-temporal database, which is organized and managed using the spatio-temporal cube model, supporting multi-dimensional retrieval by time, space, feature type, etc. A change monitoring briefing for the previous 24 hours is automatically generated daily and pushed synchronously through the WebGIS platform and mobile APP. Key changes such as newly added building area exceeding 50 m² and soil pile height exceeding 2 m are highlighted. The early warning information adopts a hierarchical push mechanism. Level 1 early warnings (such as suspected illegal buildings) trigger SMS notifications in real time, and level 2 early warnings (such as abnormal ecological indicators) are included in the daily summary report. A comprehensive analysis report for the previous month's monitoring is automatically generated before the 5th of each month, including standardized content such as change statistics, trend analysis, and thematic maps. All output achievements are digitally signed using blockchain technology to ensure data authenticity and traceability.

[0042] Furthermore, the response to cloudy weather is described as follows: At the data acquisition level, when it is detected that the cloud cover in the target area exceeds 30%, the standby data source switching protocol is automatically triggered: First, attempt to call the archived data of other passing satellites. If there is no suitable archive, start the emergency drone reflight program, and select the best flight window through real-time meteorological cloud map analysis to ensure data collection in key areas during weather gaps. For persistent cloudy weather, the system automatically switches to the radar satellite data source, uses SAR images of satellites such as Sentinel-1 for change detection, and conducts fusion analysis in combination with historical optical images to ensure the continuity of monitoring work.

[0043] At the data processing level, for thin cloud cover (cloud amount 30 - 50%), an improved dark channel prior algorithm is used to estimate and remove cloud interference through the atmospheric scattering model; for medium cloud amount (50 - 70%), cloud removal technology based on the generative adversarial network (GAN) is applied, using the Pix2PixHD model trained with a large number of cloud - cloudless image pairs to achieve intelligent reconstruction of ground object information under the cloud; for areas with thick cloud cover (>70%), start the spatio-temporal fusion engine, comprehensively use the STARFM (Spatial and Temporal Adaptive Reflectance Fusion Model) algorithm to fuse low-resolution cloudless images with historical high-resolution images to generate approximately real high-quality images. All cloud removal processes are accompanied by a confidence evaluation layer for subsequent analysis reference.

[0044] At the application decision-making level, the change detection algorithm automatically adjusts the sensitivity parameter, sets stricter verification conditions for the cloud edge area; marks the warning information with the label "affected by weather", and automatically associates recent cloud-free images for comparison and verification; a separate chapter "Cloud Impact Assessment" is established in the result report to detail the data quality status and possible monitoring blind spots. At the same time, the system intelligently adjusts the subsequent monitoring plan, gives priority to re-measuring the cloud-affected areas, and automatically starts a special review process after the weather improves to ensure the integrity and reliability of the monitoring results. This set of countermeasures enables the system to maintain an effective monitoring rate of over 85% in mining areas with an average annual cloud cover of 60%.

Claims

1. A mining area fixed-point data surveying, collecting and monitoring device, characterized in that, It includes: A main control module for coordinating and controlling the operation of the device; A multi-source data acquisition module integrating a satellite remote sensing receiving unit, an unmanned aerial vehicle (UAV) aerial photography unit, and a ground measurement unit to achieve "sky-ground" integrated data acquisition; A data fusion processing module for screening, matching, and fusing multi-source remote sensing data to generate high-precision spatially continuous data; A change detection algorithm module for detecting ground object changes in the collected data based on a deep learning model; A communication module for transmitting the collected and processed data to the cloud or the mine area big data steward system; A positioning module using a high-precision GPS or Beidou positioning system; A power supply module for providing stable power supply to the device; A storage module for temporarily storing the collected original data and processing results; An environmental sensor module for monitoring the temperature, humidity, and air pressure parameters in the mine area; A mobile terminal interface module supporting connection with the intelligent inspection APP to achieve real-time data interaction and task scheduling.

2. The mine area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The satellite remote sensing receiving unit of the multi-source data acquisition module supports the data of the Gaojing-1 satellite with a resolution of 0.5 meters and is compatible with other commercial satellite data. By optimizing the satellite source combination scheme, this unit can ensure the monitoring continuity through redundant data when a single satellite is affected by weather, and at the same time supports data acquisition with a 10-day revisit cycle, adapting to the ground object inspection tasks in the whole mine area or the minefield range.

3. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, wherein: The UAV aerial photography unit is equipped with a high-resolution optical camera and lidar. After the satellite remote sensing preliminarily identifies ground object changes, it can conduct centimeter-level refined aerial survey on key areas. This unit obtains the location, occupation area, shape, height, volume, and construction progress of the target area through a mapping UAV, providing high-resolution auxiliary verification data for the satellite remote sensing detection results.

4. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The ground measurement unit includes a total station and portable measurement equipment for manual on-site review and precise mapping in areas where UAV aerial survey is restricted. This unit collects the precise spatial attributes of the changed ground objects through mobile measurement equipment and generates a review report in combination with image data, forming a closed-loop verification system for "sky-ground" integrated monitoring.

5. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The data fusion processing module adopts a fusion method combining "wavelet transform + local algorithm + principal component analysis". It processes the spatio-temporal spectral information of multi-source images through wavelet transform frequency division, enhances details using the local variance algorithm, retains low-frequency features by local difference weighting, and then reduces spectral distortion through the transformation of PCA, finally generating a high-resolution multi-spectral fusion image.

6. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The change detection algorithm module is based on the U-Net deep learning model, supports the ResNet 18 and ResNet 34 backbone networks, and realizes pixel-level classification through an encoder-decoder structure and skip connections. This module is trained using a pre-constructed dataset of 2052 groups of four-channel samples. When the area of the changed patch is greater than 200 pixels, the accuracy rate ≥ 0.7 and the recall rate ≥ 0.9, and it can identify and output the boundaries, types, and change intensities of the ground object change patches.

7. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The communication module supports 4G / 5G networks and LoRa wireless transmission technology, adapts to the complex environment of the mining area through multi-mode communication. In areas without public network coverage, LoRa is used to achieve long-distance and low-power data transmission, while in the base station coverage area, it switches to the 4G / 5G high-speed network. This module can transmit the change detection results, early warning information and original image data back to the cloud or the big data steward system in real time.

8. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The early warning module automatically triggers hierarchical early warnings by analyzing the change detection results in real time and combining with preset rules. When detecting changes in illegal ground objects or abnormal environmental parameters, it immediately pushes early warning information to the big data steward system and synchronizes it to the mobile inspection APP to guide on-site personnel to quickly locate and verify, forming a closed-loop management process of "monitoring - early warning - disposal".

9. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The device shell adopts an IP67-level dust and waterproof design, and the internal circuit is processed by seismic reinforcement. The shell material is set as a lightweight and high-strength alloy, and the surface is coated with an anti-corrosion coating.

10. The mining area fixed-point data surveying, collecting and monitoring device according to claim 1, characterized in that: The device is seamlessly connected to the integrated monitoring big data steward system of the mining area through a standardized API interface, supports automatic uploading of ground object change patches, environmental parameters and review evidence chain data. The system can send inspection tasks downwards to the mobile device interface of the device for intelligent management of task scheduling, data synchronization and result feedback. At the same time, expansion interfaces are reserved to be compatible with future newly added monitoring devices or data sources.

Images