A comprehensive monitoring method and system for road construction

Abstract

The present application relates to a comprehensive monitoring method and system for road construction, the comprehensive monitoring method includes: obtaining construction plan information, periodically obtaining image information of each construction area of ​​the current paving and compaction process as an image information set according to the construction plan information; obtaining a preliminary inspection information set; obtaining image monitoring results of each construction area according to the image information set and the preliminary inspection information set; taking one or more construction areas that meet the first construction quality requirement as intermediate construction areas, and judging whether each intermediate construction area meets the second construction quality requirement within a preset time period from the current moment; taking one or more areas that do not meet the second construction quality requirement within a preset time period from the current moment as target construction areas, and generating monitoring prompt information for indicating the scope of the target construction area and sending it to the on-site user equipment. The system and method of the present application ensure that the adjustment of the construction plan information can take into account the comprehensive monitoring results.

Description

Technical Field

[0001] The present application belongs to the fields of computer vision technology and industrial data processing technology, and in particular, relates to a comprehensive monitoring method and system for road construction. Background Art

[0002] Road construction technology includes processes such as stone baking, asphalt mixing, and paving and compaction. The monitoring of the construction process is mainly focused on the monitoring of paving and roadbed compaction. When quality problems are found during monitoring, the cause is found to determine whether the previous process has an impact. In other words, there is no in-depth consideration of the quality standards before paving operations.

[0003] For example, taking the road construction comprehensive monitoring method, system, equipment and program product disclosed in application number CN202410430402.9 as an example, a road construction BIM model is constructed and road construction progress plan information is obtained, so that road construction personnel can perform road construction operations according to the road construction progress plan information; during the road construction process, the construction monitoring data of the construction area is collected in real time, and then the construction monitoring data is associated with the road construction BIM model to obtain the associated BIM model. Although it can perform comprehensive monitoring of road construction according to the associated BIM model, the mentioned road construction BIM model is only used to realize the three-dimensional representation of data such as road terrain, road structure and road infrastructure structure, that is, its focus is on the site of the road to be constructed to ensure that the road construction BIM model reflects the actual situation of the road construction site, and does not care about the impact of the previous process (stone baking or asphalt mixing, etc.) on the road construction.

[0004] Based on this, the present application provides a comprehensive monitoring method and system for road construction to improve related technologies. Summary of the invention

[0005] In view of the above deficiencies in the prior art, the purpose of this application is to provide a comprehensive monitoring method and system for road construction, which can avoid the impact of incomplete monitoring when only focusing on a single process through comprehensive consideration of the monitoring results of each process.

[0006] In a first aspect of the present application, a comprehensive monitoring method for road construction is proposed, the comprehensive monitoring method comprising:

[0007] Acquire construction plan information, and periodically acquire image information of each construction area of ​​the current paving and compaction process as an image information set according to the construction plan information; the construction plan information includes paving and compaction plan information for indicating the construction arrangement of the paving and compaction process;

[0008] Acquire a preliminary inspection information set, where the preliminary inspection information set is used to indicate monitoring results corresponding to the asphalt mixture in each of the construction areas;

[0009] Acquire image monitoring results of each of the construction areas according to the image information set and the initial inspection information set, wherein the image monitoring results are used to indicate whether the corresponding construction area meets the first construction quality requirement;

[0010] One or more construction areas that meet the first construction quality requirement are used as intermediate construction areas, and it is determined whether each of the intermediate construction areas meets the second construction quality requirement within a preset time period from the current moment; the second construction quality requirement is higher than the first construction quality requirement, and the preset time period includes multiple cycles for acquiring image information;

[0011] One or more areas that do not meet the second construction quality requirement within a preset period from the current moment are taken as target construction areas, and monitoring prompt information indicating the scope of the target construction area is generated and sent to the on-site user equipment.

[0012] In some optional embodiments, the image information set includes first preliminary inspection information and second preliminary inspection information corresponding to the asphalt mixture of each of the construction areas, the second preliminary inspection information is used to indicate the mixing monitoring result of the mixture in the mixing process, and the first preliminary inspection information is used to indicate the stone baking monitoring result of the stone in the mixture in the stone baking process;

[0013] The obtaining of image monitoring results of each of the construction areas according to the image information set and the initial inspection information set includes:

[0014] Determine, according to the first preliminary inspection information in the preliminary inspection information set, a target acquisition mode for each of the construction areas, the target acquisition mode including a first acquisition mode and a second acquisition mode; when the construction area corresponds to the first acquisition mode, the construction area is regarded as a first type of area; when the construction area corresponds to the second acquisition mode, the construction area is regarded as a second type of area;

[0015] For each first-type area, according to its corresponding image information and construction plan information, push the image information of the first-type area to the on-site user equipment through three-dimensional modeling, and receive the monitoring score of the first-type area by the technician using the on-site user equipment as the image monitoring result;

[0016] For each second type area, the corresponding image information and construction plan information are input into a preset image monitoring scoring model to obtain a predicted score as an image monitoring result.

[0017] In some optional embodiments, the method of determining whether the construction area meets the first construction quality requirement includes:

[0018] If the monitoring score or the predicted score is within the first preset score range, it is considered that the corresponding construction area meets the first construction quality requirement; otherwise, it is considered that the corresponding construction area does not meet the first construction quality requirement;

[0019] The method further includes: taking one or more construction areas that do not meet the first construction quality requirement as problematic construction areas.

[0020] In some optional embodiments, within a preset period from the current moment, for any intermediate construction area, a method for determining whether the second construction quality requirement is met includes:

[0021] Acquire multiple verification models within a score range where the monitoring score or the predicted score of the intermediate construction area is located, each of the verification models including multiple verification curves and a reference score corresponding to each verification curve;

[0022] Acquire actual curves of multiple monitoring scores or predicted scores within a preset time period starting from the current moment, and use the reference score corresponding to the verification curve with the highest similarity to the actual curve in each of the verification models as the second construction quality score;

[0023] If the second construction quality score is within a second preset score range, it is considered that the corresponding intermediate construction area meets the second construction quality requirement; otherwise, it is considered that the corresponding intermediate construction area does not meet the second construction quality requirement.

[0024] In some optional embodiments, the construction plan information also includes stone baking plan information, and the first initial inspection information is obtained in the following manner:

[0025] Obtaining a stone image information set according to the stone baking plan information, the stone image information set including a plurality of stone image information, each of the stone image information including identification data of a corresponding sampling point and stone image data of stones of the same batch at the sampling point; each sampling point corresponds to a combustion furnace in the stone baking process;

[0026] According to each piece of stone image information in the stone image information set, determining whether the batch of stones corresponding to each sampling point meets the first stone baking requirement;

[0027] The stone image information that does not meet the first stone baking requirement is deleted from the stone image information set, and the batches of stones corresponding to the sampling points that meet the first stone baking requirement are used as aggregates for the asphalt mixing plant; according to the stone image information set after the deletion processing, it is determined whether the batches of stones corresponding to the remaining sampling points meet the second stone baking requirement;

[0028] The stone image data in the stone image information of the sampling point that meets the second stone baking requirement is set to a null value, and the identification data of the sampling point that meets the second stone baking requirement and the stone image data with the null value are associated and used as the first initial inspection data in the first initial inspection information;

[0029] The stone image information of the sampling points that do not meet the second stone baking requirement and the corresponding identification data are associated and used as the first initial inspection data in the first initial inspection information.

[0030] In some optional embodiments, determining the target acquisition mode for each of the construction areas according to the first preliminary inspection information in the preliminary inspection information set includes:

[0031] For each construction area, judging whether the asphalt mixture in the construction area meets the preset conditions of the first acquisition mode according to the first initial inspection information of multiple batches of stone materials in the corresponding asphalt mixture, and if so, the target acquisition mode of the construction area is the first acquisition mode, otherwise, it is the second acquisition mode;

[0032] The preset conditions of the first acquisition mode are:

[0033] Obtaining each first preliminary inspection data in the first preliminary inspection information of the stone in the corresponding asphalt mixture;

[0034] When the stone image data of the first initial inspection data is null, statistics are started and the number of statistics is increased by one;

[0035] Detecting whether the proportion of the statistical number of times in the number of all first preliminary inspection data obtained is greater than a preset proportion;

[0036] When the proportion of the statistical number of times in the number of all first initial inspection data obtained is greater than a preset proportion, the statistical number of times is cleared and it is considered that the preset condition of the first acquisition mode is met.

[0037] In some optional embodiments, the construction plan information includes mixing plan information, the mixing plan information includes the ratio parameters of aggregate, powder and asphalt in each pot of mixed material during mixing, and process parameters, the process parameters include mixing temperature parameters and grading parameters;

[0038] The method for obtaining the second initial inspection information includes:

[0039] According to the mixing plan information, the mixing operation is performed using the mixing equipment of the asphalt mixing plant, and the mixing information collection equipment is used to obtain the mixing collection information during the mixing process of each pot of mixed material and form a mixing collection information set;

[0040] According to the mixing collection information, computer vision inspection technology is used to determine whether the asphalt mixture formed by each pot of mixture is qualified; the mixing collection information of the asphalt mixture that is judged to be qualified is used as the second initial inspection information.

[0041] In some optional embodiments, the method of judging whether the batch of stones corresponding to each sampling point meets the first stone baking requirement includes:

[0042] The stone image data in each stone image information are respectively input into the stone baking scoring model to obtain the corresponding baking score, and the obtained baking score is used as the stone baking score of the batch of stone corresponding to the stone image data; when the stone baking score is within the first baking scoring standard, it is considered whether the batch of stone corresponding to the sampling point meets the first stone baking requirement; otherwise, it is considered that it does not meet the first stone baking requirement, and the corresponding stone image information is deleted from the stone image information set; wherein, the stone baking score is used to indicate the combustion degree of the stone, and when the stone baking score is not within the first baking scoring standard, it means that the corresponding combustion furnace has insufficient oil.

[0043] In some optional embodiments, the method of judging whether the batch of stones corresponding to each sampling point meets the second stone baking requirement includes:

[0044] The stone image information set after the deletion process is used as the second stone image information set, and the stone image information within the preset time period of the same identification data in the second stone image information set is arranged and classified into one group according to the acquisition time; if there is only one stone image information with the same identification data within the preset time period, the preset time period is extended until at least N stone image information is obtained, where N is a positive integer greater than 4;

[0045] Input the same set of stone image information into the stability calculation model to obtain the stability scores corresponding to the stones with the same identification data;

[0046] When the stability score is within the second baking score standard, it is considered that the multiple batches of stone corresponding to the sampling point meet the second stone baking requirement; otherwise, it is considered that they do not meet the second stone baking requirement.

[0047] In a second aspect of the present application, a comprehensive monitoring system for road construction is proposed for implementing the comprehensive monitoring method provided in the first aspect, wherein the comprehensive monitoring system comprises a drone, a cloud server, and a stone baking controller, a mixing controller and a central controller that are communicatively connected to the cloud server;

[0048] The stone baking controller is arranged in the stone baking process, and is used to generate the first preliminary inspection information and transmit it to the cloud server; the mixing controller is arranged in the mixing process, and is used to obtain the first preliminary inspection information from the cloud server and generate the second preliminary inspection information, and transmit the second preliminary inspection information to the cloud server; the central controller is arranged in the paving and compacting process, and is used to obtain the construction plan information and the preliminary inspection information set consisting of the first preliminary inspection information and the second preliminary inspection information from the cloud server, and generate the acquisition control information according to the construction plan information and send it to the drone, so as to control the drone to periodically obtain the image information of each construction area of ​​the current paving and compaction process as the image information set.

[0049] In some optional embodiments, the stone baking process further includes a stone information collection device connected to the stone baking controller, which is used to collect images of the baked stones of each combustion furnace;

[0050] The mixing process also includes a mixing information collection device connected to the mixing controller, which is used to obtain mixing information of each pot of mixed material in the mixing equipment during the mixing process.

[0051] The beneficial effects of the present application are as follows: The present application proposes a comprehensive monitoring method and system for road construction, which uses a large amount of construction data collected in real time from multiple processes as industrial big data, and through hierarchical monitoring of two layers (first construction quality and second construction quality), as well as the introduction of the initial inspection information set when judging the first construction quality, and the second construction quality judgment method of analyzing the information changing trend over time after the first construction quality monitoring is qualified, it is ensured that the adjustment of the construction plan information can take into account the monitoring results of multiple processes, avoiding the influence of incomplete monitoring when only focusing on a single process, thereby affecting the timely adjustment of the construction plan information. Compared with the related art that only considers the monitoring of a single type of process and the industrial data processing of a single type of process, the production information of the relevant raw materials will be considered only after problems occur in the key process, that is, the monitoring and monitoring information of other construction processes are not taken seriously. The technical solution provided by the present application ensures that the construction quality meets the predetermined standards by continuously monitoring the construction process. Periodically acquiring image information and analyzing it in real time can quickly discover problems and make adjustments, thereby improving construction efficiency. By using the comprehensive monitoring of image information and initial inspection information, quality problems in construction can be discovered in real time, and early warning information can be generated in time to reduce the occurrence of quality problems. When the construction area does not meet the quality requirements, early warning information can be generated in time to help the construction team take timely measures to avoid potential risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are only used to illustrate specific embodiments and are not considered to limit the present application. In the entire drawings, the same reference symbols represent the same components. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a flow chart of a comprehensive monitoring method for road construction provided in an embodiment of the present application.

[0054] Figure 2 It is a flow chart of obtaining image monitoring results provided in an embodiment of the present application.

[0055] Figure 3 It is a flow chart of determining the second construction quality requirement provided in an embodiment of the present application.

[0056] Figure 4 It is a flow chart of obtaining first initial inspection information provided in an embodiment of the present application.

[0057] Figure 5 It is a flow chart of obtaining second initial verification information provided in an embodiment of the present application.

[0058] Figure 6It is a structural diagram of a comprehensive monitoring system provided in an embodiment of the present application.

[0059] Figure 7a It is a flow chart of a stone baking process of a comprehensive monitoring method for road construction provided in an embodiment of the present application.

[0060] Figure 7b It is a flow chart of the paving and compaction process of a comprehensive monitoring method for road construction provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present application, the technical solutions of the present application will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present application. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present application.

[0062] Furthermore, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily confusing the concepts disclosed in the present application.

[0063] Deep learning is a machine learning method based on artificial neural networks. A network with a multi-layer structure is called a deep neural network. A deep learning model consists of multiple layers of neurons, each of which can extract and transform different levels of features of the data. Deep learning models usually require a large amount of data for training in order to capture the statistical characteristics of the data. At the same time, deep learning models can achieve highly complex input-output mapping through nonlinear activation functions.

[0064] In deep learning, establishing a functional relationship from input to output is an automated process that is achieved through the following steps: First, the parameters of the model (including weights and biases) are randomly initialized. Then, in each training iteration, the input data is fed into the model and forward propagated through a series of computational layers (such as convolutional layers, pooling layers, and fully connected layers) to finally generate a predicted output. The difference between the predicted output and the true label is quantified by the loss function, which defines the accuracy of the prediction. Using the gradient information of the loss function, the backpropagation algorithm calculates the gradient of each parameter, which indicates how to adjust the parameter to reduce the loss. Then, based on the gradient and the preset learning rate, the parameters of the model are updated to make more accurate predictions in the next iteration. This process is repeated in multiple training iterations, and may be accompanied by tuning strategies such as adjusting the learning rate, introducing regularization techniques, and optimizing the network structure to improve the performance of the model and avoid overfitting. Finally, after sufficient training and tuning, the model learns how to map the features of the input data to the output results, thereby establishing a complex nonlinear functional relationship from input to output. The learned function can capture patterns and associations in the data, allowing the model to make accurate predictions on new, unseen data and exhibit good generalization capabilities.

[0065] Industrial big data refers to the large amount of diverse and potentially valuable data collected and generated in the industrial field. The analysis and application of industrial big data can help improve production efficiency, optimize operational decisions, reduce costs and risks, and promote innovation.

[0066] Computer vision is an important branch of artificial intelligence that enables computers to understand visual information by processing and analyzing digital images or videos. Its goal is to simulate the human visual system so that computers can recognize, locate, measure, interpret and understand image content and make decisions or perform tasks based on this information.

[0067] The existing comprehensive monitoring process of road construction is mainly through on-site manual observation and recording, which is not only time-consuming and labor-intensive, but also leads to inaccurate monitoring results due to the influence of human subjective factors. In order to solve this problem, some construction units have begun to consider the use of industrial big data and adopt drone technology to obtain more information to realize road construction monitoring. Specifically, road construction monitoring is realized through a system composed of multiple drones. Each drone can work together to improve the efficiency and effectiveness of task execution. Each drone can realize information exchange and collaborative control through technologies such as wireless communication, sensor sharing, and collaborative navigation to make it more flexible, scalable, and fault-tolerant. The drone can be equipped with a laser radar for three-dimensional point cloud data of the area to be measured; a global positioning system (GPS) for positioning and navigation to provide flight position and trajectory information; and a camera for taking photos or videos.

[0068] In the related art, computer vision technology is used to monitor the current image information obtained by drones and the construction plan information in the project establishment stage in real time, and to check the road construction quality through three-dimensional modeling and other methods. For example, taking the comprehensive monitoring method and system for highway road construction disclosed in application number CN202310053317.0 as an example, a three-dimensional real-scene model of the engineering structure in the current construction stage is obtained, and the road settlement deformation monitoring value is obtained according to the three-dimensional real-scene model of the engineering structure and the monitoring data of the monitoring point, and a road settlement prediction model is constructed based on deep learning, and the road settlement deformation monitoring value is predicted and analyzed; the predicted value of the road settlement deformation is obtained to generate early warning information, and the staged construction constraints of the current road construction stage are adjusted and optimized according to the early warning information. It mainly processes the monitoring data in real time to reduce the lag of artificial processing of monitoring data. Although it is also mentioned that the use of drones for site scanning and point cloud data for three-dimensional reconstruction are used for users to judge the quality of construction, the construction materials (asphalt mixture) are also not taken into account in the monitoring process. This is because in its technical solution, construction raw materials (information) are only used to generate construction features. Construction features are used to set construction parameter standards in combination with construction quality requirements. Construction constraints are set through construction parameter standards and dynamic management of the construction process is carried out to obtain a three-dimensional real-life model of the engineering structure at the current road construction stage. In other words, it focuses on single-point (single process) monitoring of the construction road on site. Construction raw materials (information) can be obtained in any form, and it does not care about the comprehensive monitoring of the entire process from stone baking, asphalt mixing to paving and compaction.

[0069] Similarly, illustratively, taking the Internet of Things-based road construction monitoring system disclosed in application number CN202410438312.4 as an example, industrial big data is used to analyze the image data of the construction site in real time to accurately identify progress deviations and the actual status of key sections. The actual status of the construction is identified through image data, and a decision tree analysis is performed based on the analysis results of the problem area to generate a road section efficiency optimization plan. Although it is also mentioned that the feasibility and potential risks of construction under various environmental conditions are judged by comparing and analyzing the environmental sensitivity of real-time environmental monitoring data with construction materials and processes, and generating environmental sensitivity analysis results, the construction materials (asphalt mixtures) are also not considered as a type of industrial big data in the monitoring process. This is because it also regards paving, compaction and maintenance as key activities in the construction process, while ignoring the importance of the pre-process.

[0070] The present application proposes a comprehensive monitoring method and system for road construction, which utilizes industrial big data of a large amount of construction data and construction plan information collected in real time from multiple processes, through double-layer (first construction quality and second construction quality) hierarchical monitoring, as well as the introduction of an initial inspection information set when judging the first construction quality and a second construction quality judgment process of analyzing the information change trend over time after the first construction quality monitoring is qualified, to ensure that the adjustment of the construction plan information can take into account the monitoring results of each process, avoid the impact of incomplete monitoring when only focusing on a single process, and further affect the adjustment of the construction plan information.

[0071] The following will be combined with the accompanying drawings and specific implementation methods to specifically describe in detail the technical solutions of the embodiments of the present application and how the technical solutions of the embodiments of the present application solve the above-mentioned technical problems. It should be noted that the embodiments described below or the technical features can be arbitrarily combined to form new embodiments. The description order of the following embodiments is not intended to limit the preferred order of the embodiments. The same or similar concepts or processes may not be repeated in some embodiments. Obviously, the described embodiments are part of the embodiments of the present application, not all of them.

[0072] Method embodiment.

[0073] See also Figure 1 , Figure 1 1 is a flow chart of a comprehensive monitoring method for road construction provided by an embodiment of the present application. Specifically, the comprehensive monitoring method for road construction is applied to a comprehensive monitoring system for road construction, aiming to improve the overall quality of road construction through multi-level monitoring and real-time early warning. The comprehensive monitoring method for road construction includes steps S101 to S105.

[0074] S101, obtaining construction plan information, and periodically obtaining image information of each construction area of ​​the current paving and compacting process as an image information set according to the construction plan information; the construction plan information includes paving and compacting plan information for indicating the construction arrangement of the paving and compacting process; it can also be understood that the construction plan information is used to indicate the construction arrangement of each construction process, and the construction process includes a stone baking process, a mixing process and a paving and compacting process;

[0075] S102, obtaining a preliminary inspection information set, wherein the preliminary inspection information set is used to indicate the monitoring results corresponding to the asphalt mixture in each of the construction areas;

[0076] S103, obtaining an image monitoring result of each of the construction areas according to the image information set and the initial inspection information set, wherein the image monitoring result is used to indicate whether the corresponding construction area meets the first construction quality requirement;

[0077] S104, taking one or more construction areas that meet the first construction quality requirement as intermediate construction areas, and determining whether each of the intermediate construction areas meets the second construction quality requirement within a preset time period from the current moment; the second construction quality requirement is higher than the first construction quality requirement, and the preset time period includes multiple cycles for acquiring image information;

[0078] S105: One or more areas that do not meet the second construction quality requirement within a preset period from the current moment are taken as target construction areas, and monitoring prompt information indicating the scope of the target construction area is generated and sent to the on-site user equipment.

[0079] In this embodiment, a drone is used to obtain image information of each construction area of ​​the current paving and compaction process, and the image information is sent to the central controller of the construction site. The drone is used to collect image information of multiple construction areas; the central controller is connected to one or more drones for communication and is used to execute the above-mentioned comprehensive monitoring method. It can be understood that the central controller and the stone baking controller and mixing controller mentioned below may include a memory and at least one processor respectively, the memory stores a computer program, and the at least one processor is configured to execute the computer program and is combined with other processors to implement the method described in any one of the method embodiments. The current moment in this application, for example, refers to the time point when the image monitoring results are recorded. In specific applications, the data collected from different construction processes (such as stone baking, mixing, paving and compaction) can be managed in a unified manner through a big data platform.

[0080] It can be understood that the technical solution provided in this embodiment first obtains the construction plan information, and based on this information, periodically obtains the image information of each construction area of ​​the current paving and compaction process to form an image information set. The construction plan information includes the construction arrangements of the stone baking process, the mixing process and the paving and compaction process. The periodicity is, for example, based on a time interval of 15 minutes, 25 minutes, etc.; it is also based on the progress of the construction area as a cycle. The construction area that meets the first construction quality requirement is marked as an intermediate construction area, and continues to monitor whether it meets the higher second construction quality requirement within a preset period of time (including multiple image information acquisition cycles). The monitoring prompt information and the monitoring abnormal information mentioned below are, for example, sound warnings, image warnings or text warnings, which are displayed to users using on-site user devices in the form of pop-up windows, text messages or emails. On-site user devices are, for example, laptops, industrial computers, tablet computers, etc.

[0081] Compared with the related art that only considers the monitoring of a single type of process, the production information of related raw materials will be considered only after problems occur in the key process, that is, the monitoring and monitoring information of other construction processes are not taken seriously. The troubleshooting and early warning method provided in this embodiment directly uses the industrial big data of the initial inspection information set of the stone baking process and the mixing process to obtain the image monitoring results, completes the judgment of the first construction quality requirement, and it can be considered that the first construction quality requirement is a judgment of the conventional quality standard; the construction area that meets the conventional quality standard is judged by the second construction quality with a higher quality evaluation standard within the time period. The above technical solution has the following advantages: by continuously monitoring the construction process, ensure that the construction quality meets the predetermined standards. Periodically obtain image information and analyze it in real time, you can quickly find problems and make adjustments, and improve construction efficiency. By using the comprehensive monitoring of image information and initial inspection information, quality problems in construction can be found in real time, and early warning information can be generated in time to reduce the occurrence of quality problems. When the construction area does not meet the quality requirements, early warning information can be generated in time to help the construction team take measures in time to avoid potential risks. This technical solution covers the stone baking, mixing and paving compaction processes. By using computer vision technology, it can comprehensively monitor the entire construction process and improve the overall construction quality through data analysis of multiple processes.

[0082] See also Figure 2 , Figure 2 It is a flow chart of obtaining image monitoring results provided in an embodiment of the present application.

[0083] In some embodiments, the image information set includes first and second preliminary inspection information corresponding to the asphalt mixture of each of the construction areas, the second preliminary inspection information is used to indicate the mixing monitoring result of the mixture in the mixing process, and the first preliminary inspection information is used to indicate the stone baking monitoring result of the stone in the mixture in the stone baking process. In this case, the monitoring prompt information also includes the second preliminary inspection information corresponding to the target construction area (mixing collection information of the asphalt mixture corresponding to the target construction area), so that the monitoring method takes into account the impact of the pre-process including stone baking or asphalt mixing on the road construction.

[0084] The obtaining of image monitoring results of each of the construction areas according to the image information set and the initial inspection information set includes:

[0085] S201, determining a target acquisition mode for each of the construction areas according to the first initial inspection information in the initial inspection information set, the target acquisition mode including a first acquisition mode and a second acquisition mode; when the construction area corresponds to the first acquisition mode, the construction area is regarded as a first type of area; when the construction area corresponds to the second acquisition mode, the construction area is regarded as a second type of area;

[0086] S202, for each first type of area, according to its corresponding image information and construction plan information, push the image information of the first type of area to the on-site user equipment through three-dimensional modeling, and receive the monitoring score of the first type of area by the technician using the on-site user equipment as the image monitoring result;

[0087] S203: For each second type area, the corresponding image information and construction plan information are input into a preset image monitoring scoring model to obtain a predicted score as an image monitoring result.

[0088] Through this implementation, the target acquisition mode and the type of area of ​​each construction area are determined according to the initial inspection information set. For the first type of area, its image information and construction plan information are pushed to the on-site user device by means of three-dimensional modeling. The technician (i.e., the user) monitors and scores the area through the on-site user device, and the scoring method is, for example, keyboard input, voice input, etc., and the scoring result is used as the image monitoring result. For the second type of area, its image information and construction plan information are input into a pre-set image monitoring scoring model and a predicted score is generated, and this predicted score is used as the image monitoring result. Among them, the construction plan information may include paving and compaction plan information to indicate the construction arrangement of the paving and compaction process, and the paving and compaction plan information includes the predicted image information of the current construction area (each second type of area), and the predicted image information is predicted based on the image information before construction, the time arrangement, the tasks and work content to be completed in each construction stage (such as construction methods, construction sequence, process arrangement, etc.), quality requirements, etc.

[0089] It can be considered that the scoring results and the predicted scores belong to the same type of scoring, such as the 100-point scoring system, which makes the results of different acquisition modes (manual scoring and model prediction) consistent and comparable. According to the specific situation, the most suitable scoring acquisition mode can be selected, which can rely on the professional judgment of technicians and the advanced model prediction technology, and can flexibly respond to the evaluation needs of different construction areas and processes, ensuring the comprehensiveness and effectiveness of the monitoring method.

[0090] This implementation has the following benefits: by distinguishing between the first acquisition mode and the second acquisition mode, different monitoring methods are used for different types of construction areas, thereby improving the flexibility and adaptability of monitoring. For the first type of area, more accurate monitoring results can be obtained through on-site scoring by technicians combined with three-dimensional modeling; for the second type of area, the scoring results are automatically generated through the image monitoring scoring model, thereby improving the monitoring efficiency and consistency.

[0091] Before construction begins, the image monitoring scoring model can be trained using historical data and expert knowledge to enable it to identify image features of the construction area and match them with construction plan information. During construction, real-time image information is input into the image monitoring scoring model, which generates a predicted score based on the construction plan information and image features.

[0092] In some embodiments, the present application can train an image monitoring scoring model and the stone baking scoring model and stability calculation model described later. In other optional embodiments, the present application can use pre-trained image monitoring scoring models, stone baking scoring models, and stability calculation models. Taking the image monitoring scoring model as an example, it is used to predict the score of each second type area based on image information and construction plan information. The stone baking scoring model and stability calculation model mentioned below are similar to the training process of the above-mentioned image monitoring scoring model and will not be repeated.

[0093] Among them, the image monitoring scoring model can be obtained by training a preset first deep learning model using a first training set, and each first training data in the first training set includes a set of sample image information groups and corresponding scoring annotation data. The sample image data group includes first sample image data and second sample image data, the second sample image data can be the predicted image information recorded in a certain construction area in history, the first sample image data is the image information after actual construction corresponding to the second sample image data, and the corresponding score is the score given by an expert. An expert refers to a person who has an in-depth understanding of the construction process, construction materials, construction technology and construction standards, and can judge whether the construction quality meets the requirements through experience based on the construction plan, actual construction conditions and industry standards such as the construction and quality acceptance specifications for urban road projects.

[0094] Obtain a first training set, where the first training set includes a plurality of first training data; and perform the following processing on each first training data in the first training set:

[0095] Inputting the sample image data group in the first training data into a preset first depth model to obtain prediction data of the sample image data group;

[0096] Based on the prediction data and the annotation data of the sample image data group, updating the model parameters of the first depth model;

[0097] Detect whether a preset training end condition is met; if yes, use the trained first depth model as the image monitoring and scoring model; if not, continue to train the first depth model using the next first training data.

[0098] Therefore, by designing, establishing an appropriate amount of neuron computing nodes and a multi-layer operation hierarchy, and selecting appropriate input and output layers, a preset deep learning model can be obtained. Through the learning and tuning of the deep learning model, a functional relationship from input to output is established. Although the functional relationship between input and output cannot be found 100%, the actual correlation relationship can be approached as much as possible. The image monitoring and scoring model trained in this way can obtain the corresponding output data based on the input data. It has a wide range of applications, and the calculation results are highly accurate and reliable.

[0099] The embodiments of the present application do not limit the model training process, and for example, it may adopt a supervised learning training method, or a semi-supervised learning training method, or an unsupervised learning training method.

[0100] The embodiments of the present application do not limit the training termination conditions corresponding to each model. For example, it may be that the number of training times reaches a preset number (the preset number of times is, for example, 1 time, 3 times, 10 times, 100 times, 1000 times, 10000 times, etc.), or it may be that the training data in the corresponding training set has completed one or more trainings, or it may be that the total loss value obtained from this training is not greater than the preset loss value.

[0101] In some implementations, for example, data mining can be performed on historical data to obtain training data in a training set. That is, the training data can be information collected from real road construction records. In addition, the sample information can also be automatically generated using a generative network of a GAN model in the prior art, which is not limited in this application.

[0102] In some implementations, the method of determining whether the construction area meets the first construction quality requirement includes:

[0103] If the monitoring score or the predicted score is within the first preset score range, it is considered that the corresponding construction area meets the first construction quality requirement; otherwise, it is considered that the corresponding construction area does not meet the first construction quality requirement;

[0104] The method further includes: taking one or more construction areas that do not meet the first construction quality requirement as problematic construction areas.

[0105] It can be considered that the problematic construction area that does not meet the first construction quality requirement is an area that does not meet the conventional quality standards. There is no need to evaluate the second construction quality requirement, which also reduces the amount of data processing in subsequent steps. In a specific application, when a problematic construction area is identified, monitoring abnormality information is generated and sent to the on-site user device. Unlike the monitoring prompt information mentioned above, the monitoring abnormality information indicates that the corresponding area does not meet the basic construction quality standards, while the monitoring prompt information only indicates that the basic construction quality standards are met but there may be potential problems or risks that require further attention.

[0106] In this implementation mode, a first preset scoring range is pre-set to define the standard of construction quality; the on-site technicians actually score the construction area according to technical means such as three-dimensional modeling, or the scoring results are automatically generated by the image monitoring scoring model. For each construction area, the monitoring score or predicted score is compared to see whether it is within the range. If it is within the range, the construction area is considered to meet the first construction quality requirement, and if it is not within the range, the construction area is considered to not meet the first construction quality requirement. The construction area that does not meet the first construction quality requirement is marked as a problem construction area, which indicates that further processing or inspection is required to ensure the construction quality.

[0107] This implementation has the following benefits: comprehensive use of monitoring scores and prediction scores, combined with automated and manual evaluation, giving full play to the advantages of each, and improving the comprehensiveness and effectiveness of the monitoring process.

[0108] See also Figure 3 , Figure 3 It is a flow chart of determining the second construction quality requirement provided in an embodiment of the present application.

[0109] In some implementations, within a preset period from the current moment, for any intermediate construction area, a method for determining whether the second construction quality requirement is met includes:

[0110] S301, obtaining a plurality of verification models within a score range where the monitoring score or the predicted score of the intermediate construction area is located, each of the verification models including a plurality of verification curves and a reference score corresponding to each verification curve;

[0111] S302, obtaining actual curves of multiple monitoring scores or predicted scores within a preset period from the current moment, and taking the reference score corresponding to the verification curve with the highest similarity to the actual curve in each of the verification models as the second construction quality score;

[0112] S303: If the second construction quality score is within a second preset score range, it is considered that the corresponding intermediate construction area meets the second construction quality requirement; otherwise, it is considered that the corresponding intermediate construction area does not meet the second construction quality requirement.

[0113] In a specific application, when multiple monitoring scores or predicted scores within a preset time period starting from the current moment are not within the first preset score range, the corresponding intermediate construction area is used as a problem construction area, and monitoring abnormality information is generated and sent to the on-site user equipment, and the corresponding construction area no longer executes step S302.

[0114] Through this implementation, for any intermediate construction area within a preset period from the current moment, multiple verification models related to the monitoring score or predicted score of the area are obtained. Each verification model contains multiple verification curves and a reference score corresponding to each curve, and the verification model is used to evaluate the quality of the construction area. By obtaining multiple monitoring scores or predicted scores within a preset period from the current moment, a data drawing tool (such as Plotly) is used to form a time series curve as an actual curve. The actual curve is compared with the curves in each verification model to find the verification curve with the highest similarity to the actual curve. According to the reference score corresponding to the most similar verification curve, it is determined as the second construction quality score. The calculated second construction quality score is compared with the second preset score range. If the score is within this range, it is considered that the intermediate construction area meets the second construction quality requirement; otherwise, it is considered not to meet.

[0115] This implementation has the following benefits: The comparison between the actual curve and the calibration curve can improve the accuracy of the evaluation and ensure that the quality evaluation of the construction area is closer to the actual situation. Since the areas used for the second construction quality judgment all meet the first construction quality judgment, the second construction quality requirement is higher than the first construction quality requirement. Through higher standards of quality control, potential quality problems can be avoided and the long-term performance and safety of road construction can be ensured. It can be considered that the multiple calibration curves contained in the calibration model and the reference score corresponding to each curve are generated in advance by technicians based on historical data.

[0116] The similarity between the actual curve and the verification curve is obtained by, for example, measuring the similarity between the actual curve and the verification curve by using a dynamic time warping algorithm, or by using a cross-correlation value.

[0117] Measuring the similarity between the actual curve and the verification curve by means of the dynamic time warping algorithm refers to aligning the actual curve and the verification curve, and measuring the similarity by calculating the cumulative distance after alignment. Measuring the similarity between the actual curve and the verification curve by means of the cross-correlation value refers to obtaining the cross-correlation value between the actual curve and the verification curve, the range of which is, for example, -1 to 1, and the closer the value is to 1, the more similar it is.

[0118] See also Figure 4 , Figure 4It is a flow chart of obtaining first initial inspection information provided in an embodiment of the present application.

[0119] In some implementations, the construction plan information further includes stone baking plan information, and the first initial inspection information is obtained in the following manner:

[0120] S401, obtaining a stone image information set according to the stone baking plan information; the stone image information set includes a plurality of stone image information, each of which includes identification data of a corresponding sampling point and stone image data of a batch of stones at the sampling point, and each sampling point corresponds to a combustion furnace in the stone baking process; the stone baking plan information includes, for example, the type and specification of the stone, the particle size distribution of the stone, the start and end time of baking and the corresponding combustion furnace, and the control scheme of temperature change during baking, etc.; in this embodiment, the stone image data includes, for example, images of the stone at the same sampling point before and after baking;

[0121] S402, judging whether the batch of stones corresponding to each sampling point meets the first stone baking requirement according to each stone image information in the stone image information set;

[0122] S403, deleting the stone image information that does not meet the first stone baking requirement from the stone image information set, and using the batches of stones corresponding to the sampling points that meet the first stone baking requirement as aggregates for the asphalt mixing plant; judging whether the batches of stones corresponding to the remaining sampling points meet the second stone baking requirement based on the stone image information set after the deletion process;

[0123] S404, setting a null value for the stone image data in the stone image information of the sampling point that meets the second stone baking requirement, and associating the identification data of the sampling point that meets the second stone baking requirement with the stone image data with the null value and use them as the first initial inspection data in the first initial inspection information;

[0124] S405: Associating the stone image information of the sampling points that do not meet the second stone baking requirement with the corresponding identification data and using them as the first initial inspection data in the first initial inspection information.

[0125] Through this implementation, in the comprehensive monitoring process, by analyzing the stone image data of each sampling point, it is determined whether the batch of stones meets the first stone baking requirement, and the preliminary screening of the stone quality is completed. The stone image information that does not meet the first stone baking requirement is deleted from the information set, and the stones that meet the first stone baking requirement are used as aggregates for the asphalt mixing plant for further processing. The remaining stone image information set is further screened to determine whether these stones meet the second stone baking requirement, and the image data of the stones that meet the second stone baking requirement is set to a null value, and only the identification data is retained, and it is associated with the null value image data as the first initial inspection data in the first initial inspection information. The image information and identification data of the stones that do not meet the second stone baking requirement are also associated as the first initial inspection data in the first initial inspection information.

[0126] This implementation has the following benefits: Associating the stone image information with the specific sampling point identification data can ensure that the quality control can be accurate to the same batch of a single combustion furnace. If there is a problem with a batch of stones in a combustion furnace, it can be quickly located. The stone image information that does not meet the first stone baking requirements will be deleted from the stone image information set to ensure that only qualified stones are used for the subsequent asphalt mixing plant aggregate preparation. By setting the stone image data that meets the second stone baking requirements to null values, the size of the data set is reduced, simplifying subsequent data processing and analysis. Based on the analysis and quality requirements of image data, data-driven decision-making is supported, which helps scientific management of the construction process.

[0127] In specific applications, a relational database management system (RDBMS) can be used to store the above data sets. The database includes multiple data tables (such as stone image information table, sampling point identification data table, etc.), and the relationship between tables is established through foreign keys to ensure data consistency and integrity. At the same time, the use of big data technology can support real-time or near real-time data stream processing of data sets in relational databases.

[0128] In some embodiments, the method for determining whether the batch of stones corresponding to each of the sampling points meets the first stone baking requirement includes: inputting the stone image data in each stone image information into the stone baking scoring model to obtain the corresponding baking score, and using the obtained baking score as the stone baking score of the batch of stones corresponding to the stone image data; when the stone baking score is within the first baking scoring standard, it is considered that the batch of stones corresponding to the sampling point meets the first stone baking requirement; otherwise, it is considered that it does not meet the first stone baking requirement, and the corresponding stone image information is deleted from the stone image information set. The first baking scoring standard is, for example, a score of not less than 70 points, 85 points, etc. Among them, the stone baking score is used to indicate the combustion degree of the stone. When the stone baking score is not within the first baking scoring standard, it means that the corresponding combustion furnace has insufficient oil.

[0129] The batch of stones corresponding to the sampling points in this application refers to the stones that have been baked in a combustion furnace. The image data can reflect the uniformity of the combustion of the stone surface in terms of the consistency of color distribution, as well as the cracks, holes or other surface defects that may occur during the baking process, and the color changes of the stone after baking (the color changes are related to the changes in the chemical composition of the stone), which can be used to evaluate the degree of combustion of the stone.

[0130] The focus of collecting image data of baked stones is to capture features related to baking quality. In specific applications, the features extracted from the image data are input into the stone baking scoring model to evaluate the baking quality of the stone.

[0131] In some implementations, the method of determining whether the batch of stones corresponding to each sampling point meets the second stone baking requirement includes:

[0132] The stone image information set after the deletion process is used as the second stone image information set, and the stone image information within the preset time period of the same identification data in the second stone image information set is arranged and classified into one group according to the acquisition time; the preset time period; for example, 20 minutes, 30 minutes or 50 minutes; if there is only one stone image information within the preset time period, the preset time period is extended until at least N stone image information is obtained, where N is a positive integer greater than 4;

[0133] The same set of stone image information is input into the stability calculation model to obtain the stability score corresponding to the stone with the same identification data; the stability score is, for example, 60 points, 80 points, A, B or C; when the score is higher or closer to A, it means that the quality of the same sampling point (combustion furnace) corresponding to the identification data is more stable;

[0134] When the stability score is within the second baking score standard, it is considered that the multiple batches of stones corresponding to the sampling point meet the second stone baking requirement; otherwise, it is considered that they do not meet the second stone baking requirement. The second baking score standard is, for example, a score greater than 70 points and not less than C.

[0135] The stone image information set after preliminary screening (deleting unqualified stone image information) is used as the second stone image information set, and the stone image information of the same identification data (corresponding to the same combustion furnace) is arranged and classified within the preset time period according to the acquisition time. A certain amount of stone image information is collected for stability analysis. The same group of stone image information is input into the stability calculation model, which is used to evaluate the stability of the stone baking process. According to the second baking scoring standard, it is judged whether the stone meets the second baking requirements. The setting of the preset time period and the data classification method are used to ensure that enough data is collected for stability analysis. Through the stability score, the stone baking quality can be more comprehensively evaluated.

[0136] See also Figure 5 , Figure 5 It is a flow chart of obtaining second initial verification information provided in an embodiment of the present application.

[0137] In some embodiments, the construction plan information includes mixing plan information, which includes the ratio parameters of aggregate, powder and asphalt in each pot of mixture during mixing, and process parameters, and the process parameters include mixing temperature parameters and grading parameters; grading parameters refer to parameters used to describe the particle composition and distribution of aggregates (such as sand, gravel, etc.) in road construction, which determine the particle size distribution of aggregates and their proportion in the mixture, and affect the mechanical properties, durability and construction quality of the road.

[0138] The method for obtaining the second initial inspection information includes:

[0139] S501, according to the mixing plan information, using the mixing equipment of the asphalt mixing plant to perform the mixing operation, and using the mixing information collection equipment to obtain the mixing collection information during the mixing process of each pot of mixed material and form a mixing collection information set;

[0140] According to the mixing plan information, the asphalt mixing station performs mixing operations on the mixing equipment to ensure that the mixture is mixed according to the established ratio and process parameters. During the mixing process, the mixing information collection equipment monitors and records the mixing data of each pot of mixture in real time to form a mixing collection information set.

[0141] S502: Based on the collected mixing information, computer vision inspection technology is used to determine whether the asphalt mixture formed by each pot of mixture is qualified, and the collected mixing information of the qualified asphalt mixture is used as the second initial inspection information.

[0142] This application does not limit computer vision inspection technology, which is a method that uses image processing and pattern recognition technology to automatically detect the quality of products or materials. As an example, it may include the following:

[0143] The image acquisition step uses a high-resolution camera or other visual sensor to capture images of the asphalt mixture. The image of the asphalt mixture can be static or a continuous video stream during the mixing process.

[0144] Image preprocessing step, to process the acquired images. The preprocessing step may include denoising, contrast enhancement, brightness adjustment and color balance adjustment, etc., to improve the accuracy of subsequent analysis.

[0145] In the feature extraction step, key features such as color, texture, shape, and size are extracted from the preprocessed images for quality assessment of asphalt mixtures.

[0146] In the pattern recognition step, the extracted features are analyzed using machine learning algorithms (such as support vector machines, neural networks, or decision trees) to identify the quality and consistency of the mix.

[0147] In the decision-making step, based on the results of pattern recognition, it is automatically determined whether the mixture meets the preset quality standards. If not, the asphalt mixture is marked as unqualified.

[0148] In specific applications, for asphalt mixtures that are judged to be unqualified, abnormal mixture information is generated and sent to on-site user equipment to prompt that there are problems with the current mixture.

[0149] The powder is one or more of mineral powder, stone chips or sand, and the asphalt can be petroleum asphalt. The mixing plan information includes, for example, the ratio parameters of aggregate, powder and asphalt in each batch of mixture, as well as process parameters (such as mixing temperature and grading parameters) to define the preparation requirements and process standards of the asphalt mixture. Correspondingly, the mixing data of each batch of mixture refers to the ratio parameters of aggregate, powder and asphalt in the actual mixing process, as well as the actual process parameters (such as the actual mixing temperature and grading parameters).

[0150] Through this embodiment, the raw materials are mixed according to the predetermined ratio and process conditions, and the mixing equipment of the asphalt mixing station is used to perform actual mixing operations to generate asphalt mixture. During the mixing process, the mixing process data of each pot of mixture is collected using a mixing information collection device (for example, including a camera). By analyzing the collection of mixing information sets, the second preliminary inspection information of the asphalt mixture is obtained to further verify the degree to which the quality of the asphalt mixture meets the preset standards and requirements.

[0151] As an example, a batch of raw materials is added to the mixer according to the proportion, and the mixing is started. In the mixer, the aggregate, powder and asphalt are fully mixed at a preset temperature (e.g., 170 to 190 degrees) for a preset time (e.g., 5 minutes). After the mixing is completed, the mixed asphalt mixture is discharged from the discharge port as a pot of mixture.

[0152] This implementation has the following benefits: The proportion parameters and process parameters in the mixing plan information provide clear mixing standards, which help to ensure that the formula and production process of the mixture meet the requirements. The mixing process is monitored in real time through the mixing information collection equipment to ensure that the actual production of the mixture is consistent with the plan.

[0153] It is worth noting that in the comprehensive monitoring corresponding to the mixing process, the first preliminary inspection information is not referenced in the monitoring process of judging whether the asphalt mixture formed by each pot of mixture is qualified. This is because it is assumed that the stones in the mixture meet the first stone baking requirements. Even if they do not meet the second stone baking requirements, they meet the basic conditions as aggregates for asphalt mixing plants. In this case, there is no need to refer to the first preliminary inspection information, as long as the asphalt mixture is qualified. This reduces unnecessary information processing processes, reduces the demand for data processing and transmission, and helps reduce costs associated with data management.

[0154] Similarly, the first preliminary verification information in the present application is used to consider the early stage of the data, divide the acquisition mode, and realize the judgment of the first acquisition mode and the second acquisition mode.

[0155] In some implementations, determining the target acquisition mode for each of the construction areas according to the first initial verification information in the initial verification information set includes:

[0156] For each construction area, judging whether the asphalt mixture in the construction area meets the preset conditions of the first acquisition mode according to the first initial inspection information of multiple batches of stone materials in the corresponding asphalt mixture, and if so, the target acquisition mode of the construction area is the first acquisition mode, otherwise, it is the second acquisition mode;

[0157] The preset conditions of the first acquisition mode are:

[0158] Obtaining each first preliminary inspection data in the first preliminary inspection information of the stone in the corresponding asphalt mixture;

[0159] When the stone image data of the first initial inspection data is null, statistics are started and the number of statistics is increased by one;

[0160] Detecting whether the proportion of the statistical number of times in the number of all first preliminary inspection data obtained is greater than a preset proportion;

[0161] When the proportion of the statistical number of times in the number of all first initial inspection data obtained is greater than a preset proportion, the statistical number of times is cleared and it is considered that the preset condition of the first acquisition mode is met.

[0162] By counting the proportion of null values ​​in stone image data, effective data can be used more reasonably for quality assessment. Depending on the acquisition mode, the allocation of monitoring resources can be optimized and the monitoring efficiency can be improved.

[0163] As an example, a comprehensive monitoring method is provided, including: monitoring the stone baking process, obtaining a stone image information set, scoring each image information, and using the batch of stones corresponding to the sampling points according to the baking score and the first baking score standard as aggregate for the asphalt mixing plant; obtaining a second stone image information set; and obtaining the first preliminary inspection information according to the stability score and the second baking score standard.

[0164] Monitor the mixing process, perform mixing operations, obtain the mixing collection information set, use computer vision inspection technology to determine whether the mixed materials are qualified, and use the qualified information as the second initial inspection information.

[0165] Monitor the paving and compaction process, obtain construction plan information, obtain the initial inspection information set, determine the acquisition mode of the construction area, and divide the construction area into the first type of area and the second type of area. The first type of area monitoring method is to push image information to the on-site user equipment through three-dimensional modeling and receive the monitoring score of the technician. The second type of area monitoring method is to input image information and construction plan information into the image monitoring and scoring model, obtain the predicted score as the image monitoring result, determine whether the first construction quality requirement is met, and regard the area that does not meet the first construction quality requirement as the problem construction area. For the remaining area, determine whether the second construction quality requirement is met, and regard the area that does not meet the second construction quality requirement as the target construction area, and generate monitoring prompt information and send it to the on-site user equipment.

[0166] As an example, in the stone baking process, the stone baking plan information plans to bake 200 tons of stone, and 10 combustion furnaces work simultaneously. The first baking score standard is not less than 70 points. The second baking score standard is not less than 80 points. The N value (the minimum number of stone images required for stability scoring) is 5.

[0167] In the mixing process, the mixing plan information is that each pot of mixture consists of 50% aggregate, 30% powder and 20% asphalt, and the mixing temperature is 180°C. The standard for qualified asphalt mixture is a visual inspection score greater than 85 points.

[0168] In the paving and compaction process, the current construction plan information is 3 construction areas, and each area needs to pave and compact 80 tons of asphalt mixture. The preset proportion of the first acquisition mode is 85%. The preset score range of the first construction quality requirement is greater than 90 points. The preset score range of the second construction quality requirement is greater than 95 points. The preset time period is 10 hours (including 6 image information acquisition cycles, each cycle is 1.5 hours).

[0169] In the stone baking process, a stone image information set is obtained, and a total of 200 stone image information is collected. Each image information corresponds to the same batch of stone in a combustion furnace. After the model is scored, it is assumed that 180 stone image information scores are not less than 70, and 20 scores are less than 70. Delete the 20 stone image information with scores less than 70, and the remaining 180 are used as aggregates for the asphalt mixing plant. The remaining 180 stone image information are scored for stability, and it is assumed that 160 stones have a stability score of not less than 80. Set the 160 stone image data with a stability score of not less than 80 points to null, and retain the identification data as the first initial inspection information.

[0170] In the mixing process, the mixing operation is performed, and each pot of mixed material is mixed in proportion according to the mixing plan. Each pot of mixed material is visually inspected, assuming that 95 pots of mixed materials have a score greater than 85 points, and 5 pots have a score less than 85 points. The mixing collection information of the 95 pots of qualified mixed materials is used as the second initial inspection information.

[0171] During the paving and compaction process, the image information of the three construction areas is periodically obtained, and the first preliminary inspection information and the second preliminary inspection information are obtained from the cloud server. Assume that the proportion of null values ​​in the stone image data of all construction areas is greater than 85%, which meets the first acquisition mode condition. The technicians score the first type of area; and the predicted score of the second type of area is obtained through the model. The scores of the construction areas are: 91 points for construction area 1, 92 points for construction area 2, and 93 points for construction area 3. The scores of all construction areas are within the first preset score range and meet the first construction quality requirements. The score range includes 85 to 89, 90 to 95, and 96 to 100. The three scores are in the same score range, and multiple verification models corresponding to the score range of 90 to 95 are obtained. Then the image information of the three construction areas is periodically obtained every 1.5 hours, for a total of 10 hours. The three construction areas are scored continuously, and the scores obtained by time arrangement are: Construction Area 1: 91, 92, 91, 91, 92, etc. Construction Area 2: 92, 93, 91, 91, 92, etc. Construction area 3: 93, 91, 91, 92, 93, etc. It is obtained that the curve similarity between construction area 1 and model A is the highest, and the reference score of model A is 98; the curve similarity between construction area 2 and model C is the highest, and the reference score of model C is 96; the curve similarity between construction area 3 and model E is the highest, and the reference score of model E is 93. Construction area 3 does not meet the preset score range of the second construction quality requirement, and a monitoring prompt message is generated. It is worth noting that the above examples are only exemplary and should not be construed as limitations on the claims. In practice, those skilled in the art can flexibly select and adjust the scoring criteria, N value, qualified asphalt mixture standards, etc. according to actual needs, in order to obtain the best technical effect. At the same time, the details not described in detail in this embodiment are all well known in the art and will not be repeated here.

[0172] As an example, in the preset deep learning model, the image size of the input layer is 224×224, and the number of neurons in the input layer is 224×224×3=150,528, where 3 represents the color channel (RGB). Two convolutional layers are selected, each followed by a maximum pooling layer. The first convolutional layer contains 32 filters (neurons), each of which is 3×3 in size. The second convolutional layer contains 64 filters, also of 3×3 size. Two fully connected layers are set after the convolution and pooling layers. The output layer uses the softmax activation function for multi-class classification. Among them, the cross entropy loss function is used to measure the difference between the predicted data and the labeled data. The Adam optimizer is selected, and the initial learning rate is set to 0.001. The Dropout regularization technique is applied to reduce the risk of overfitting, and the Dropout rate is set to 0.5. The training cycle (epochs) is set to 50 cycles. During the training process, if it is found that the performance of the model on the validation set is no longer improved, the learning rate can be reduced. If the model is overfitting, data diversity can be increased through data augmentation techniques (such as rotating and scaling images). Through the back-propagation algorithm, the model learns the mapping from input images to output category probabilities. During the training process, the weights and biases of the model are continuously updated to minimize the loss function.

[0173] System Embodiments.

[0174] See also Figure 6 , Figure 6 It is a structural diagram of a comprehensive monitoring system provided in an embodiment of the present application.

[0175] As another specific embodiment of the present application, disclosed is a comprehensive monitoring system for implementing the comprehensive monitoring method as described in any one of the implementation methods in the above method embodiments. Its specific embodiments are consistent with the embodiments recorded in the above method embodiments and the technical effects achieved, and some contents will not be repeated here.

[0176] The comprehensive monitoring system includes a drone, a cloud server, and a stone baking controller, a mixing controller and a central controller that are communicatively connected to the cloud server;

[0177] The stone baking controller is set in the stone baking process, used to generate the first initial inspection information and transmit it to the cloud server; the mixing controller is set in the mixing process, used to obtain the first initial inspection information from the cloud server and generate the second initial inspection information, and transmit the second initial inspection information to the cloud server; the central controller is set in the paving and compacting process, used to obtain the construction plan information and the initial inspection information set composed of the first initial inspection information and the second initial inspection information from the cloud server, and generate the acquisition control information according to the construction plan information and send it to the drone, and control the drone to periodically obtain the image information of each construction area of ​​the current paving and compacting process as the image information set. The cloud server can dynamically adjust the computing resources, such as CPU, memory and storage, according to the amount of data to be processed and the processing requirements, to adapt to the volatility of big data; big data services including data warehouses, big data processing frameworks (such as Hadoop, Spark) and machine learning platforms can also be obtained through cloud service providers.

[0178] Among them, the stone baking process where the comprehensive monitoring system is located also includes a stone information collection device connected to the stone baking controller, which is used to collect images of the baked stones of each combustion furnace. The mixing process also includes a mixing information collection device connected to the mixing controller, which is used to obtain the mixing collection information of each pot of mixed material in the mixing equipment during the mixing process. The stone baking controller, the mixing controller and the central controller, for example, respectively include processors such as CPU (Central Processing Unit, central processing unit), computing circuits (integrated circuits) such as FPGA (Field Programmable Gate Array, field programmable gate array) and edge computing devices (such as industrial computers or embedded systems on site) of communication modules and storage modules. The stone information collection device, for example, includes multiple cameras, and the stone baking controller can control the cameras of all stone information collection devices to collect image information of the baked stones of the corresponding sampling points in real time, and the stone information collection device can send the collected image information of the stones to the stone baking controller.

[0179] See also Figure 7a and Figure 7b , Figure 7a It is a flow chart of a stone baking process of a comprehensive monitoring method for road construction provided in an embodiment of the present application. Figure 7b It is a flow chart of the paving and compaction process of a comprehensive monitoring method for road construction provided in an embodiment of the present application.

[0180] As an example, a comprehensive monitoring method for road construction implemented by a comprehensive monitoring system is provided, wherein the comprehensive monitoring system includes a drone, a cloud server, and a stone baking controller, a mixing controller, and a central controller that are communicatively connected to the cloud server;

[0181] The stone baking controller is arranged in the stone baking process, and is used to generate the first preliminary inspection information and transmit it to the cloud server; the mixing controller is arranged in the mixing process, and is used to obtain the first preliminary inspection information from the cloud server and generate the second preliminary inspection information, and transmit the second preliminary inspection information to the cloud server; the central controller is arranged in the paving and compacting process, and is used to obtain the construction plan information and the preliminary inspection information set consisting of the first preliminary inspection information and the second preliminary inspection information from the cloud server, and generate the acquisition control information according to the construction plan information and send it to the drone, so as to control the drone to periodically obtain the image information of each construction area of ​​the current paving and compaction process as the image information set.

[0182] 1. In the stone baking process, a stone baking controller is used to obtain stone baking plan information from a cloud server. According to the stone baking plan information, a stone image information set is obtained through a stone information acquisition device. The stone image information set includes a plurality of stone image information. Each of the stone image information includes identification data of a corresponding sampling point and stone image data of stones of the same batch at the sampling point. Each sampling point corresponds to a combustion furnace in the stone baking process.

[0183] According to each stone image information in the stone image information set, the stone image data in each stone image information is respectively input into the stone baking scoring model to obtain the corresponding baking score, and the obtained baking score is used as the stone baking score of the batch of stones corresponding to the stone image data; when the stone baking score is within the first baking scoring standard, it is considered that the batch of stones corresponding to the sampling point meets the first stone baking requirement; otherwise, it is considered that it does not meet the first stone baking requirement, and the corresponding stone image information is deleted from the stone image information set; wherein, the stone baking score is used to indicate the combustion degree of the stone, and when the stone baking score is not within the first baking scoring standard, it indicates that the corresponding combustion furnace has insufficient oil;

[0184] The stone image information that does not meet the first stone baking requirement is deleted from the stone image information set, and the batches of stones corresponding to the sampling points that meet the first stone baking requirement are used as aggregates for the asphalt mixing plant; the stone image information set after the deletion process is used as the second stone image information set, and the stone image information within the preset time period of the same identification data in the second stone image information set is arranged and classified into one group according to the acquisition time; if there is only one stone image information with the same identification data within the preset time period, the preset time period is extended until at least N stone image information is obtained, where N is a positive integer greater than 4; the same group of stone image information is input into the stability calculation model to obtain the stability score corresponding to the stone with the same identification data; when the stability score is within the second baking score standard, it is considered that the multiple batches of stones corresponding to the sampling points meet the second stone baking requirement; otherwise, it is considered that they do not meet the second stone baking requirement; the stone baking score model and the stability calculation model are executed in the stone baking controller;

[0185] The stone image data in the stone image information of the sampling point that meets the second stone baking requirement is set to a null value, and the identification data of the sampling point that meets the second stone baking requirement and the stone image data with the null value are associated and used as the first initial inspection data in the first initial inspection information;

[0186] The stone image information of the sampling point that does not meet the second stone baking requirement is associated with the corresponding identification data and used as the first preliminary inspection data in the first preliminary inspection information. The above-mentioned first preliminary inspection information is sent to the cloud server. In other words, the stone baking controller is responsible for obtaining the stone image information set through the stone information acquisition device according to the stone baking plan information, and integrates the image acquisition and preliminary processing functions in the comprehensive monitoring.

[0187] Second, in the mixing process, the mixing controller is used to obtain the mixing plan information from the cloud server. According to the mixing plan information, the mixing equipment of the asphalt mixing plant is used to perform the mixing operation, and the mixing information collection equipment is used to obtain the mixing collection information of each batch of mixed materials during the mixing process and form a mixing collection information set;

[0188] According to the mixing collection information, computer vision inspection technology is used to determine whether the asphalt mixture formed by each pot of mixture is qualified; the mixing collection information of the asphalt mixture that is judged to be qualified is used as the second initial inspection information, and the second initial inspection information is sent to the central server.

[0189] Third, the paving and compaction process uses the central controller to obtain the construction plan information from the cloud server. According to the construction plan information, the image information of each construction area of ​​the current paving and compaction process is periodically obtained as an image information set; the construction plan information is used to indicate the construction arrangement of each construction process, which includes the stone baking process, the mixing process and the paving and compaction process;

[0190] The central controller is used to obtain a preliminary inspection information set from the cloud server, wherein the preliminary inspection information set includes first preliminary inspection information and second preliminary inspection information corresponding to the asphalt mixture in each construction area, wherein the second preliminary inspection information is used to indicate a mixing monitoring result of the mixture in the mixing process, and the first preliminary inspection information is used to indicate a stone baking monitoring result of the stone in the mixture in the stone baking process;

[0191] For each construction area, according to the first preliminary inspection information of multiple batches of stone in the corresponding asphalt mixture, it is judged whether the asphalt mixture in the construction area meets the preset conditions of the first acquisition mode. If it meets the conditions, the target acquisition mode of the construction area is the first acquisition mode, otherwise, it is the second acquisition mode; when the construction area corresponds to the first acquisition mode, the construction area is regarded as the first type of area; when the construction area corresponds to the second acquisition mode, the construction area is regarded as the second type of area; the preset conditions of the first acquisition mode are: obtaining each first preliminary inspection data in the first preliminary inspection information of the stone in the corresponding asphalt mixture; when the stone image data of the first preliminary inspection data obtained is a null value, starting to count and increasing the number of statistics by one; detecting whether the proportion of the statistical number in the number of all the first preliminary inspection data obtained is greater than the preset proportion; when the proportion of the statistical number in the number of all the first preliminary inspection data obtained is greater than the preset proportion, the statistical number is cleared to zero, and it is considered that the preset conditions of the first acquisition mode are met.

[0192] For each first-type area, according to its corresponding image information and construction plan information, the image information of the first-type area is pushed to the on-site user equipment through 3D modeling, and the monitoring score of the first-type area by the technician using the on-site user equipment is received as the image monitoring result;

[0193] The image monitoring and scoring model is executed in the stone baking controller; for each second type area, the corresponding image information and construction plan information are input into the preset image monitoring and scoring model to obtain a predicted score as the image monitoring result, and the image monitoring result is used to indicate whether the corresponding construction area meets the first construction quality requirement; the method for judging whether the first construction quality requirement is met is: if the monitoring score or the predicted score is within the first preset score range, it is considered that the corresponding construction area meets the first construction quality requirement; otherwise, it is considered that the corresponding construction area does not meet the first construction quality requirement;

[0194] One or more construction areas that do not meet the first construction quality requirement are regarded as problem construction areas.

[0195] One or more construction areas that meet the first construction quality requirement are used as intermediate construction areas, and it is judged whether each intermediate construction area meets the second construction quality requirement within a preset time period from the current moment; the second construction quality requirement is higher than the first construction quality requirement, and the preset time period includes multiple cycles for obtaining image information; the method for judging whether the second construction quality requirement is met is: obtaining multiple verification models (within the score range where the monitoring score or predicted score of the intermediate construction area is located), each verification model includes multiple verification curves and a reference score corresponding to each verification curve; obtaining multiple actual curves of monitoring scores or predicted scores within the preset time period from the current moment, and taking the reference score corresponding to the verification curve with the highest similarity to the actual curve in each verification model as the second construction quality score; if the second construction quality score is within the second preset score range, it is considered that the corresponding intermediate construction area meets the second construction quality requirement; otherwise, it is considered that the corresponding intermediate construction area does not meet the second construction quality requirement;

[0196] One or more areas that do not meet the second construction quality requirement within a preset period from the current moment are taken as target construction areas, and monitoring prompt information indicating the scope of the target construction area is generated by the central controller and sent to the on-site user equipment.

[0197] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present application, rather than to limit them. Although the present application is described in detail with reference to the aforementioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. Any changes or replacements that can be easily thought of by a person of ordinary skill in the art within the technical scope disclosed in the present application should be covered within the scope of protection of the present application.

Claims

1. A comprehensive monitoring method for road construction, characterized in that: The comprehensive monitoring method comprises: Acquire construction plan information, and periodically acquire image information of each construction area of ​​the current paving and compaction process as an image information set according to the construction plan information; the construction plan information includes paving and compaction plan information for indicating the construction arrangement of the paving and compaction process; Acquire a preliminary inspection information set, where the preliminary inspection information set is used to indicate monitoring results corresponding to the asphalt mixture in each of the construction areas; Acquire image monitoring results of each of the construction areas according to the image information set and the initial inspection information set, wherein the image monitoring results are used to indicate whether the corresponding construction area meets the first construction quality requirement; One or more construction areas that meet the first construction quality requirement are used as intermediate construction areas, and it is determined whether each of the intermediate construction areas meets the second construction quality requirement within a preset time period from the current moment; the second construction quality requirement is higher than the first construction quality requirement, and the preset time period includes multiple cycles for acquiring image information; One or more areas that do not meet the second construction quality requirement within a preset period from the current moment are taken as target construction areas, and monitoring prompt information indicating the scope of the target construction area is generated and sent to the on-site user equipment.

2. The comprehensive monitoring method for road construction according to claim 1, characterized in that: The image information set includes first preliminary inspection information and second preliminary inspection information corresponding to the asphalt mixture of each of the construction areas, the second preliminary inspection information is used to indicate the mixing monitoring result of the mixture in the mixing process, and the first preliminary inspection information is used to indicate the stone baking monitoring result of the stone in the mixture in the stone baking process; The obtaining of image monitoring results of each construction area according to the image information set and the initial inspection information set includes: Determine, according to the first preliminary inspection information in the preliminary inspection information set, a target acquisition mode for each of the construction areas, the target acquisition mode including a first acquisition mode and a second acquisition mode; when the construction area corresponds to the first acquisition mode, the construction area is regarded as a first type of area; when the construction area corresponds to the second acquisition mode, the construction area is regarded as a second type of area; For each first-type area, according to its corresponding image information and construction plan information, push the image information of the first-type area to the on-site user equipment through three-dimensional modeling, and receive the monitoring score of the first-type area by the technician using the on-site user equipment as the image monitoring result; For each second type area, the corresponding image information and construction plan information are input into a preset image monitoring scoring model to obtain a predicted score as an image monitoring result.

3. The comprehensive monitoring method for road construction according to claim 2, characterized in that: Methods for determining whether the construction area meets the first construction quality requirement include: If the monitoring score or the predicted score is within the first preset score range, it is considered that the corresponding construction area meets the first construction quality requirement; otherwise, it is considered that the corresponding construction area does not meet the first construction quality requirement; The method further includes: taking one or more construction areas that do not meet the first construction quality requirement as problematic construction areas.

4. The comprehensive monitoring method for road construction according to claim 2, characterized in that: In a preset period from the current moment, for any intermediate construction area, the method of judging whether the second construction quality requirement is met includes: Acquire multiple verification models within a score range where the monitoring score or the predicted score of the intermediate construction area is located, each of the verification models including multiple verification curves and a reference score corresponding to each verification curve; Acquire actual curves of multiple monitoring scores or predicted scores within a preset time period starting from the current moment, and use the reference score corresponding to the verification curve with the highest similarity to the actual curve in each of the verification models as the second construction quality score; If the second construction quality score is within a second preset score range, it is considered that the corresponding intermediate construction area meets the second construction quality requirement; otherwise, it is considered that the corresponding intermediate construction area does not meet the second construction quality requirement.

5. The comprehensive monitoring method for road construction according to claim 2, characterized in that: The construction plan information also includes stone baking plan information, and the first initial inspection information is obtained by: Obtaining a stone image information set according to the stone baking plan information, the stone image information set including a plurality of stone image information, each of the stone image information including identification data of a corresponding sampling point and stone image data of stones of the same batch at the sampling point; each sampling point corresponds to a combustion furnace in the stone baking process; According to each piece of stone image information in the stone image information set, determining whether the batch of stones corresponding to each sampling point meets the first stone baking requirement; The stone image information that does not meet the first stone baking requirement is deleted from the stone image information set, and the batches of stones corresponding to the sampling points that meet the first stone baking requirement are used as aggregates for the asphalt mixing plant; according to the stone image information set after the deletion processing, it is determined whether the batches of stones corresponding to the remaining sampling points meet the second stone baking requirement; The stone image data in the stone image information of the sampling point that meets the second stone baking requirement is set to a null value, and the identification data of the sampling point that meets the second stone baking requirement and the stone image data with the null value are associated and used as the first initial inspection data in the first initial inspection information; The stone image information of the sampling points that do not meet the second stone baking requirement and the corresponding identification data are also associated and used as the first initial inspection data in the first initial inspection information.

6. The comprehensive monitoring method for road construction according to claim 5, characterized in that: Determining the target acquisition mode of each construction area according to the first preliminary inspection information in the preliminary inspection information set includes: For each construction area, judging whether the asphalt mixture in the construction area meets the preset conditions of the first acquisition mode according to the first initial inspection information of multiple batches of stone materials in the corresponding asphalt mixture, and if so, the target acquisition mode of the construction area is the first acquisition mode, otherwise, it is the second acquisition mode; The preset conditions of the first acquisition mode are: Obtaining each first preliminary inspection data in the first preliminary inspection information of the stone in the corresponding asphalt mixture; When the stone image data of the first initial inspection data is null, statistics are started and the number of statistics is increased by one; Detecting whether the proportion of the statistical number of times in the number of all first preliminary inspection data obtained is greater than a preset proportion; When the proportion of the statistical number of times in the number of all first initial inspection data obtained is greater than a preset proportion, the statistical number of times is cleared and it is considered that the preset condition of the first acquisition mode is met.

7. The comprehensive monitoring method for road construction according to claim 5, characterized in that: The construction plan information includes mixing plan information, which includes the mixing ratio parameters of aggregate, powder and asphalt in each pot of mixed material during mixing, and process parameters, which include mixing temperature parameters and grading parameters; The method for obtaining the second initial inspection information includes: According to the mixing plan information, the mixing operation is performed using the mixing equipment of the asphalt mixing plant, and the mixing information collection equipment is used to obtain the mixing collection information during the mixing process of each pot of mixed material and form a mixing collection information set; According to the mixing collection information, computer vision inspection technology is used to determine whether the asphalt mixture formed by each pot of mixture is qualified; the mixing collection information of the asphalt mixture that is judged to be qualified is used as the second initial inspection information.

8. The comprehensive monitoring method for road construction according to claim 7, characterized in that: Methods for judging whether the batches of stones corresponding to the sampling points meet the first stone baking requirements include: The stone image data in each stone image information are respectively input into the stone baking scoring model to obtain the corresponding baking score, and the obtained baking score is used as the stone baking score of the batch of stone corresponding to the stone image data; when the stone baking score is within the first baking scoring standard, it is considered that the batch of stone corresponding to the sampling point meets the first stone baking requirement; otherwise, it is considered that it does not meet the first stone baking requirement, and the corresponding stone image information is deleted from the stone image information set; wherein, the stone baking score is used to indicate the combustion degree of the stone, and when the stone baking score is not within the first baking scoring standard, it means that the corresponding combustion furnace has insufficient oil.

9. The comprehensive monitoring method for road construction according to claim 8, characterized in that: Methods for judging whether the batches of stones corresponding to the sampling points meet the second stone baking requirements include: The stone image information set after the deletion process is used as the second stone image information set, and the stone image information within the preset time period of the same identification data in the second stone image information set is arranged and classified into one group according to the acquisition time; if there is only one stone image information with the same identification data within the preset time period, the preset time period is extended until at least N stone image information is obtained, where N is a positive integer greater than 4; Input the same set of stone image information into the stability calculation model to obtain the stability scores corresponding to the stones with the same identification data; When the stability score is within the second baking score standard, it is considered that the multiple batches of stone corresponding to the sampling point meet the second stone baking requirement; otherwise, it is considered that they do not meet the second stone baking requirement.

10. A comprehensive monitoring system for implementing the comprehensive monitoring method according to claim 1, characterized in that: The comprehensive monitoring system includes a drone, a cloud server, and a stone baking controller, a mixing controller and a central controller that are communicatively connected to the cloud server; The stone baking controller is arranged in the stone baking process, and is used to generate the first preliminary inspection information and transmit it to the cloud server; the mixing controller is arranged in the mixing process, and is used to obtain the first preliminary inspection information from the cloud server and generate the second preliminary inspection information, and transmit the second preliminary inspection information to the cloud server; the central controller is arranged in the paving and compacting process, and is used to obtain the construction plan information and the preliminary inspection information set consisting of the first preliminary inspection information and the second preliminary inspection information from the cloud server, and generate the acquisition control information according to the construction plan information and send it to the drone, so as to control the drone to periodically obtain the image information of each construction area of ​​the current paving and compaction process as the image information set.

Images