Tunnel construction process management method, device and system based on AI identification

Abstract

According to the AI identification-based tunnel construction process management method, device and system provided by the invention, the face image of the constructor is acquired through the image acquisition equipment to obtain the personnel identification result, the first process identification result is determined according to the preset personnel and process corresponding relation, and the identity information of the construction carrier is determined through the radio frequency identification equipment; and determining a first process identification result according to a preset carrier and process corresponding relation, determining a second process identification result according to a preset carrier and process corresponding relation, and determining a construction process stage of the target position according to the first process identification result, a first weight corresponding to the first process identification result, the second process identification result and a second weight corresponding to the second process identification result. Therefore, the construction process stage of the tunnel can be determined more accurately, and accurate data support is provided for follow-up construction progress management and control and construction safety management and control.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a tunnel construction process management method based on AI recognition. Background Art

[0002] In tunnel construction scenarios, there are high requirements for construction safety and there is a problem of small construction space. Therefore, in order to ensure construction safety and improve construction efficiency, it is necessary to accurately and timely understand the current construction process or construction progress. The existing method of manually recording and reporting construction processes or construction progress is prone to misreporting or untimely information updates. Summary of the invention

[0003] In order to overcome the above-mentioned deficiencies in the prior art, the purpose of this application is to provide a tunnel construction process management method based on AI recognition, the method comprising: Acquire facial images of construction workers through image acquisition equipment; Inputting the facial image into a personnel recognition model for processing to obtain a personnel recognition result output by the personnel recognition model; Determine a first process identification result according to the personnel identification result and a preset correspondence between personnel and processes; Obtain the identity information of construction vehicles entering the tunnel through radio frequency identification equipment; Determine the second process identification result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process; The construction process stage is determined according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result, and the second weight corresponding to the second process identification result.

[0004] In some possible implementations, the method further includes: Determine a corresponding construction status identification model according to the construction process stage; Acquire a tunnel construction monitoring image of a target position in the tunnel, and input the tunnel construction monitoring image into the construction status recognition model; Processing the tunnel construction monitoring image by the personnel action feature extraction unit of the construction state recognition model to obtain a personnel action feature vector; Processing the tunnel construction monitoring image by a vehicle state feature extraction unit of the construction state recognition model to obtain a vehicle state feature vector; Processing the tunnel construction monitoring image by a personnel vehicle position feature extraction unit of the construction state recognition model to obtain a personnel vehicle position feature vector; Performing feature fusion on the personnel action feature vector, the vehicle state feature vector and the personnel vehicle position feature vector to obtain a fused feature vector; The fused feature vector is input into the classification processing unit of the construction status recognition model to obtain the construction status recognition result of the target position output by the classification processing unit.

[0005] In some possible implementations, the step of processing the tunnel construction monitoring image by the personnel vehicle position feature extraction unit of the construction state recognition model to obtain the personnel vehicle position feature vector includes: The tunnel construction monitoring image is processed by the first submodule of the personnel vehicle position feature extraction unit to obtain the first position data of at least one personnel operation area of ​​the construction vehicle; the personnel operation area includes at least one of a driving area, a lifting platform area and a control console area; the first position data includes a vehicle type data item and position data items corresponding to a plurality of personnel operation areas respectively, the vehicle type data item is used to indicate the type of the construction vehicle, and each of the position data items is used to indicate the position information of one of the personnel operation areas in the image; Processing the tunnel construction monitoring image through the second submodule of the personnel vehicle position feature extraction unit to obtain second position data of the construction personnel; The first position data and the second position data are concatenated into the personnel vehicle position feature vector.

[0006] In some possible implementations, the method further includes: The vehicle state feature vector and the personnel vehicle position feature vector are input into a pre-trained construction risk prediction module to obtain a construction risk detection result output by the construction risk prediction module.

[0007] In some possible implementations, the method further includes: Determining a moving direction of a construction vehicle according to the tunnel construction monitoring image of the target location; The traffic lights in the tunnel are controlled according to the construction status recognition result of the target position and the moving direction of the construction vehicle.

[0008] In some possible implementations, the method further includes: Determine corresponding construction progress prediction results according to the construction status identification results corresponding to multiple time points; The entry order sequence of the construction vehicles to enter the tunnel is adjusted according to the construction progress prediction result.

[0009] In some possible implementations, the method further includes: Acquire a first training sample, where the first training sample includes a training image and a first label, where the first label includes location frame information and human key point information of a construction worker in the training image; Training the personnel action feature extraction unit through the first training sample; Acquire a second training sample, wherein the second training sample includes a sample image and a second label, wherein the second label includes the type of the construction vehicle and key point information of the equipment in the sample image; Training the vehicle state feature extraction unit using the second training sample; Acquire a third training sample, wherein the third training sample includes a sample image and a third label, wherein the third label includes a position of a construction worker, a position of a construction vehicle, and a personnel operation area of ​​the construction vehicle marked in the sample image; The personnel carrier position feature extraction unit is trained using the third training sample.

[0010] In some possible implementations, the step of determining the construction process stage according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result, and the second weight corresponding to the second process identification result includes: Calculate the first The fusion recognition execution degree corresponding to the preset construction process :

[0011] in, In the identification result of the first process, The first recognition confidence corresponding to a preset construction process; The result of the first process identification is The first weight corresponding to the preset construction process; In the second process identification result, The second recognition confidence corresponding to a preset construction process; The second process identification result is The second weight corresponding to the preset construction process; The preset construction process with the highest degree of fusion recognition execution is determined as the current construction process stage.

[0012] Another object of the present application is to provide a tunnel construction process management device based on AI recognition, the device comprising: An image acquisition module, used to acquire facial images of construction workers through an image acquisition device; A first recognition module, used for inputting the face image into a person recognition model for processing, and obtaining a person recognition result output by the person recognition model; A second identification module, used to determine the first process identification result according to the personnel identification result and the preset correspondence between personnel and processes; The radio frequency acquisition module is used to obtain the identity information of the construction vehicles entering the tunnel through the radio frequency identification device; A third identification module, used to determine the second process identification result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process; The process judgment module is used to determine the construction process stage according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result and the second weight corresponding to the second process identification result.

[0013] Another object of the present application is to provide a tunnel construction process management system based on AI recognition, the system comprising an image acquisition device, a radio frequency identification device and a control device; The image acquisition device is used to obtain facial images of construction workers; The radio frequency identification device is used to obtain the identity information of the construction vehicle entering the tunnel; The control device is used to input the facial image into a personnel recognition model for processing to obtain a personnel recognition result output by the personnel recognition model; determine a first process recognition result based on the personnel recognition result and a preset correspondence between personnel and processes; determine a second process recognition result based on the identity information of the construction vehicle and a preset correspondence between the vehicle and processes; determine a construction process stage based on the first process recognition result, a first weight corresponding to the first process recognition result, the second process recognition result, and a second weight corresponding to the second process recognition result.

[0014] Compared with the prior art, this application has the following beneficial effects: The tunnel construction process management method, device and system based on AI recognition provided in this application obtains personnel recognition results by collecting facial images of construction personnel through image acquisition equipment, and determines the first process recognition result according to the preset correspondence between personnel and process, and at the same time determines the identity information of the construction vehicle through radio frequency identification equipment, and determines the second process recognition result according to the preset correspondence between vehicle and process, and then determines the construction process stage of the target location according to the first process recognition result, the first weight corresponding to the first process recognition result, the second process recognition result and the second weight corresponding to the second process recognition result. In this way, the construction process stage of the tunnel can be determined more accurately, providing accurate data support for subsequent construction progress control and construction safety control. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A schematic diagram of a tunnel construction process management system based on AI recognition provided in an embodiment of the present application; Figure 2 One of the step flow diagrams of the tunnel construction process management method based on AI recognition provided in an embodiment of the present application; Figure 3 The second step flow diagram of the tunnel construction process management method based on AI recognition provided in the embodiment of the present application; Figure 4 A schematic diagram of a control device provided in an embodiment of the present application; Figure 5 A schematic diagram of the functional modules of the device for the tunnel construction process management method based on AI recognition provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0018] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0019] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0020] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, or the positions or positional relationships in which the invented product is usually placed when in use. They are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific position, be constructed and operated in a specific position, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second", "third", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0021] In addition, the terms "horizontal", "vertical", "overhanging" and the like do not mean that the components are required to be absolutely horizontal or overhanging, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0022] In the description of this application, it should also be noted that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0023] See also Figure 1 , Figure 1 A schematic diagram of a tunnel construction process management system based on AI recognition is provided in this embodiment. The system may include a control device 100, an image acquisition device 200 and a radio frequency identification device 300. The image acquisition device 200 and the radio frequency identification device 300 may be connected to the control device 100 for communication via a wired or wireless network.

[0024] The image acquisition device 200 can be set at the entrance of the tunnel construction site and various places in the tunnel to collect images of people entering the construction site, images of vehicles, or construction monitoring images in the tunnel.

[0025] The RFID device 300 may be disposed at the entrance of a tunnel construction site and at various locations within the tunnel, and is used to identify a RFID (Radio Frequency Identification) tag on a construction vehicle to determine the identity or type information of the construction vehicle.

[0026] Please refer to Figure 2 , Figure 2For application Figure 1 The control device 100 shown is a flow chart of a tunnel construction process management method based on AI recognition. The method including each step will be described in detail below.

[0027] Step S110, acquiring a facial image of a construction worker through an image acquisition device.

[0028] In this embodiment, facial images of construction workers may be collected by an image acquisition device disposed at the entrance of a tunnel construction site or disposed in the tunnel.

[0029] Optionally, in order to ensure the recognizability of the collected facial images, in this embodiment, the facial images of the construction workers can be collected by an image acquisition device at a location with better ambient light, for example, the facial images of the construction workers can be collected by an image acquisition device at the entrance of the construction site.

[0030] Step S120: input the face image into a person recognition model for processing to obtain a person recognition result output by the person recognition model.

[0031] In this embodiment, the key facial features of the facial image can be extracted using a pre-trained personnel recognition model, and a search and match can be performed in a stored construction personnel facial feature library to determine the identity information of the construction personnel and obtain the personnel recognition result.

[0032] Step S130, determining a first process identification result according to the personnel identification result and a preset correspondence between personnel and processes.

[0033] In this embodiment, the correspondence between personnel and processes may be recorded in advance, and after the personnel identification result is obtained, the corresponding first process identification result is determined according to the correspondence between personnel and processes.

[0034] Optionally, in this embodiment, the first process identification result may be determined according to the personnel identification results corresponding to a plurality of construction personnel, thereby reducing the risk of identification errors.

[0035] Step S140, obtaining the identity information of the construction vehicle entering the tunnel through a radio frequency identification device.

[0036] In this embodiment, the construction vehicle may be provided with an RFID tag that records its identity information or equipment type information. The identity information of the construction vehicle entering the tunnel may be determined by identifying the RFID tag of the construction vehicle through a radio frequency identification device.

[0037] Step S150, determining the second process identification result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process.

[0038] Step S160, determining the construction process stage according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result and the second weight corresponding to the second process identification result.

[0039] In some cases, the construction process stage can be determined relatively more clearly based on the type of construction vehicle; in other cases, the types of construction vehicles used in multiple construction process stages may be the same, but the construction personnel are different, and the construction process stage needs to be determined based on the identity information of the construction personnel.

[0040] Therefore, in this embodiment, for different construction stages, the identification of construction personnel (i.e., the first process identification result) and the identification of construction vehicles (i.e., the second process identification result) may have different weights. Combining the first process identification result, the first weight corresponding to the first process identification result, the second process identification result, and the second weight corresponding to the second process identification result, the construction process stage of the target location can be determined more accurately.

[0041] In some possible implementations, in step S160, different construction processes may correspond to different combinations of first weights and second weights, and the first weight may be calculated by the following formula: The fusion recognition execution degree corresponding to the preset construction process :

[0042] in, In the identification result of the first process, The first recognition confidence corresponding to a preset construction process; The result of the first process identification is The first weight corresponding to the preset construction process; In the second process identification result, The second recognition confidence corresponding to a preset construction process; The second process identification result is The second weight corresponding to the preset construction process; Then the preset construction process with the highest fusion recognition execution degree is determined as the current construction process stage. The highest construction process is determined as the current construction process stage.

[0043] In some possible implementations, the construction process stage corresponding to the tunnel can be determined through steps S110 to S160. In order to further determine the corresponding construction progress status of the current construction process, please refer to Figure 3, the method provided in this embodiment may also include the following steps.

[0044] Step S210, determining a corresponding construction status recognition model according to the construction process stage.

[0045] In this embodiment, in different stages of the construction process, there may be significant differences in the actions of construction workers, types and states of construction vehicles, so different stages of the construction process may correspond to different construction state recognition models.

[0046] Step S220: acquiring a tunnel construction monitoring image of a target position in the tunnel, and inputting the tunnel construction monitoring image into the construction status recognition model.

[0047] Step S230: Processing the tunnel construction monitoring image by the personnel action feature extraction unit of the construction state recognition model to obtain a personnel action feature vector.

[0048] In this embodiment, different actions of construction workers can characterize different construction states to a certain extent. Therefore, the construction worker action feature extraction unit can be configured or trained to extract features for the actions of construction workers. For example, in the preliminary training stage, the first training sample can be used to train the personnel action feature extraction unit, and the first training sample includes a training image and a first label, and the first label includes the position frame information and human body key point information of the construction workers in the training image. For example, the first label of the first training sample may include the position frame information and human body key point information of the construction workers in the training image. In this way, after the preliminary training is completed, the construction worker action feature extraction unit can extract the position features related to the position of the construction workers and the human body key points in the image. During use, the position features related to the position of the construction workers and the human body key points corresponding to multiple continuous tunnel construction monitoring images can be spliced ​​in chronological order to obtain the personnel action feature vector.

[0049] Step S240: Processing the tunnel construction monitoring image through the vehicle state feature extraction unit of the construction state recognition model to obtain a vehicle state feature vector.

[0050] In this embodiment, different vehicle states of the construction vehicle can represent different construction states to a certain extent. For example, when the rock drilling arm of a construction vehicle with a rock drilling arm is extended and when the rock drilling arm is retracted, different construction states can be represented.

[0051] Therefore, the vehicle state feature extraction unit can be configured or trained to extract features for the type and state of the construction vehicle. For example, in the preliminary training stage, the vehicle state feature extraction unit can be trained using a second training sample, the second training sample includes a sample image and a second label, and the second label includes the type of construction vehicle and equipment key point information in the sample image. For example, the label of the second training sample may include the type of construction vehicle and equipment key point information marked in the training image (such as the vehicle body, lifting platform, rock drilling arm, arch anchor installation arm, etc.). In this way, after the preliminary training is completed, the vehicle state feature extraction unit can extract position features related to the type of construction vehicle and equipment key points in the image as the vehicle state feature vector.

[0052] Step S250: Processing the tunnel construction monitoring image by the personnel vehicle position feature extraction unit of the construction status recognition model to obtain a personnel vehicle position feature vector.

[0053] In this embodiment, when the construction workers are in different positions relative to the construction vehicle, different construction states can be represented to a certain extent. For example, for a construction vehicle with a lifting platform, different construction states can be represented when there are construction workers on the lifting platform and when there are no construction workers on the lifting platform.

[0054] Therefore, the personnel carrier position feature extraction unit can be configured or trained to extract features for the relative positions of construction personnel and construction vehicles. For example, in the preliminary training stage, the third training sample can be used to train the personnel carrier position feature extraction unit, and the third training sample includes a sample image and a third label, and the third label includes the position of the construction personnel, the position of the construction vehicle, and the personnel operation area of ​​the construction vehicle marked in the sample image. For example, the label of the third training sample may include the position of the construction personnel, the position of the construction vehicle, and the personnel operation area of ​​the construction vehicle marked in the training image. In this way, after the preliminary training is completed, the vehicle state feature extraction unit can extract features related to the position of the construction personnel, the position of the construction vehicle, and the position of the personnel operation area in the image as the personnel carrier position feature vector.

[0055] Step S260, performing feature fusion on the personnel action feature vector, the vehicle state feature vector and the personnel vehicle position feature vector to obtain a fused feature vector.

[0056] Step S270: input the fused feature vector into the classification processing unit of the construction status recognition model to obtain the construction status recognition result of the target position output by the classification processing unit.

[0057] In this embodiment, under different construction conditions, construction workers may have different actions, construction vehicles may have different states, and construction workers may be in different positions relative to the construction vehicles. The fused feature vector includes feature vectors that characterize the actions of construction workers, the states of construction vehicles, and the relative positional relationship between construction workers and construction vehicles. In this way, the corresponding construction state in the current tunnel construction monitoring image can be determined more accurately based on the fused feature vector.

[0058] It can be understood that after completing the preliminary training of the personnel action feature extraction unit, the vehicle state feature extraction unit and the personnel vehicle position feature extraction unit in the preliminary training stage, advanced training can be carried out in combination with multiple feature extraction units and classification processing units to obtain the construction status recognition model after the training is completed.

[0059] Further, in some possible implementations, step S250 may include the following sub-steps.

[0060] Step S251: Process the tunnel construction monitoring image through the first submodule of the personnel vehicle position feature extraction unit to obtain first position data of at least one personnel operation area of ​​the construction vehicle.

[0061] Among them, the personnel operation area includes at least one of a driving area, a lifting platform area and a control console area; the first position data includes a vehicle type data item and position data items corresponding to multiple personnel operation areas respectively, the vehicle type data item is used to indicate the type of the construction vehicle, and each of the position data items is used to indicate the position information of one of the personnel operation areas in the image.

[0062] Specifically, most construction vehicles include a driving area, which may be the area where construction workers are located when driving the construction vehicle, such as the main driver's seat. Some construction vehicles with a lifting platform also include a lifting platform for lifting construction workers to a certain height, and the lifting platform area is the area where construction workers are located when they are on the lifting platform. In order to have a better operating field of view, some construction vehicles also have a control console area located outside the driving area, and the control console area may include a joystick or a control button for controlling components such as the lifting platform, the rock drilling arm, and the arch anchor installation arm.

[0063] In this embodiment, the first submodule of the personnel vehicle position feature extraction unit can directly identify and determine the personnel operation area, or after identifying and determining the type of the construction vehicle, determine the relative position of the personnel operation area on the construction vehicle according to the type of the construction vehicle.

[0064] Step S252: Process the tunnel construction monitoring image through the second submodule of the personnel vehicle position feature extraction unit to obtain the second position data of the construction personnel.

[0065] In this embodiment, the second submodule of the personnel vehicle position feature extraction unit performs human body recognition on the construction personnel in the tunnel construction monitoring image, thereby determining the second position data of the construction personnel.

[0066] Step S253: concatenate the first position data and the second position data into the personnel vehicle position feature vector.

[0067] In this embodiment, the personnel vehicle position feature vector obtained by splicing the first position data and the second position data includes the position of the construction vehicle, the position of the personnel operating area of ​​the construction vehicle and the position of the construction personnel. When the personnel vehicle position feature vector is subsequently used, the relative position relationship between the construction personnel and the construction vehicle can be characterized based on the degree of overlap between the position of the construction personnel and the position of the construction vehicle and the position of the personnel operating area of ​​the construction vehicle.

[0068] In some possible implementations, after determining the personnel vehicle position feature vector, the vehicle state feature vector and the personnel vehicle position feature vector may be input into a pre-trained construction risk prediction module to obtain a construction risk detection result output by the construction risk prediction module.

[0069] For example, in a construction vehicle with a lifting platform, a rock drilling arm or an arch anchor installation arm, some positions are not allowed for construction personnel to enter during the construction operation. In step S251, the dangerous area corresponding to the construction vehicle can also be determined. After determining the position feature vector of the personnel vehicle, it can be determined whether there is a construction risk based on whether the position of the construction personnel overlaps with the position of the dangerous area, and the construction risk detection result can be obtained.

[0070] In some possible implementations, after determining the construction status recognition result, the moving direction of the construction vehicle can also be determined based on the tunnel construction monitoring image of the target location, and then the traffic lights in the tunnel can be controlled based on the construction status recognition result of the target location and the moving direction of the construction vehicle.

[0071] For example, when it is determined that the current construction process has been completed according to the construction status recognition result, and when it is determined that the construction vehicle has begun to move out of the tunnel according to the moving direction of the construction vehicle, a driving channel can be automatically reserved for the construction vehicle by controlling the traffic lights in the tunnel. In this way, the efficiency of large construction vehicles entering and exiting the tunnel when the construction phases are alternating can be effectively improved, thereby improving the overall efficiency of tunnel construction.

[0072] In some possible implementations, after determining the construction status identification result, the corresponding construction progress prediction result can also be determined based on the construction status identification results corresponding to multiple time points, and then the entry order sequence of the construction vehicles to enter the tunnel can be adjusted based on the construction progress prediction result.

[0073] In this way, in scenarios where tunnel traffic space is relatively tight, the order in which construction vehicles corresponding to different construction processes enter the tunnel can be arranged more reasonably to prevent construction vehicles from blocking the tunnel and affecting construction efficiency.

[0074] Please refer to Figure 4 , Figure 4 yes Figure 1 The block diagram of the control device 100 is shown. The control device 100 includes a tunnel construction process management device 110 based on AI recognition, a machine-readable storage medium 120, and a processor 130.

[0075] The components of the machine-readable storage medium 120 and the processor 130 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The tunnel construction process management device 110 based on AI recognition includes at least one software function module that can be stored in the machine-readable storage medium 120 in the form of software or firmware or solidified in the operating system (OS) of the control device 100. The processor 130 is used to execute the executable modules stored in the machine-readable storage medium 120, such as the software function modules and computer programs included in the tunnel construction process management device 110 based on AI recognition.

[0076] The machine-readable storage medium 120 may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc. The machine-readable storage medium 120 is used to store a program, and the processor 130 executes the program / executable the tunnel construction process management method based on AI recognition provided in this embodiment after receiving the execution instruction.

[0077] The processor 130 may be an integrated circuit chip with signal processing capabilities. The above-mentioned processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0078] Please refer to Figure 5 This embodiment further provides a tunnel construction process management device 110 based on AI recognition, and the tunnel construction process management device 110 based on AI recognition includes at least one functional module that can be stored in a machine-readable storage medium 120 in the form of software. From a functional perspective, the tunnel construction process management device 110 based on AI recognition can include an image acquisition module 111, a first recognition module 112, a second recognition module 113, a radio frequency acquisition module 114, a third recognition module 115 and a process judgment module 116.

[0079] The image acquisition module 111 is used to acquire the facial image of the construction worker through an image acquisition device.

[0080] In this embodiment, the image acquisition module 111 can be used to perform Figure 2 As shown in step S110 , for a detailed description of the image acquisition module 111 , please refer to the description of step S110 .

[0081] The first recognition module 112 is used to input the face image into a person recognition model for processing, and obtain a person recognition result output by the person recognition model.

[0082] In this embodiment, the first identification module 112 may be used to perform Figure 2 As shown in step S120, for a detailed description of the first identification module 112, please refer to the description of step S120.

[0083] The second identification module 113 is used to determine the first process identification result according to the personnel identification result and the preset corresponding relationship between personnel and processes.

[0084] In this embodiment, the second identification module 113 can be used to perform Figure 2 As shown in step S130, for a detailed description of the second identification module 113, please refer to the description of step S130.

[0085] The radio frequency acquisition module 114 is used to obtain the identity information of the construction vehicle entering the tunnel through a radio frequency identification device.

[0086] In this embodiment, the RF acquisition module 114 can be used to perform Figure 2 As shown in step S140, for a detailed description of the RF acquisition module 114, please refer to the description of step S140.

[0087] The third recognition module 115 is used to determine the second process recognition result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process.

[0088] In this embodiment, the third identification module 115 can be used to perform Figure 2 As shown in step S150, for a detailed description of the third identification module 115, please refer to the description of step S150.

[0089] The process judgment module 116 is used to determine the construction process stage of the target location according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result and the second weight corresponding to the second process identification result.

[0090] In this embodiment, the process determination module 116 can be used to execute Figure 2 As shown in step S160 , for a detailed description of the process determination module 116 , please refer to the description of step S160 .

[0091] In summary, the tunnel construction process management system based on AI recognition provided by the present application obtains personnel recognition results by collecting facial images of construction personnel through image acquisition equipment, and determines the first process recognition result according to the preset correspondence between personnel and processes, and at the same time determines the identity information of the construction vehicle through radio frequency identification equipment, and determines the second process recognition result according to the preset correspondence between the vehicle and the process, and then determines the construction process stage of the target location according to the first process recognition result, the first weight corresponding to the first process recognition result, the second process recognition result and the second weight corresponding to the second process recognition result. In this way, the construction process stage of the tunnel can be determined more accurately, providing accurate data support for subsequent construction progress control and construction safety control.

[0092] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0093] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0094] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0095] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0096] The above are only various implementations of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A tunnel construction process management method based on AI recognition, characterized in that: The method comprises: Acquire facial images of construction workers through image acquisition equipment; Inputting the facial image into a personnel recognition model for processing to obtain a personnel recognition result output by the personnel recognition model; Determine a first process identification result according to the personnel identification result and a preset correspondence between personnel and processes; Obtain the identity information of construction vehicles entering the tunnel through radio frequency identification equipment; Determine the second process identification result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process; The construction process stage is determined according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result, and the second weight corresponding to the second process identification result.

2. The method according to claim 1, characterized in that: The method further comprises: Determine a corresponding construction status identification model according to the construction process stage; Acquire a tunnel construction monitoring image of a target position in the tunnel, and input the tunnel construction monitoring image into the construction status recognition model; Processing the tunnel construction monitoring image by the personnel action feature extraction unit of the construction state recognition model to obtain a personnel action feature vector; Processing the tunnel construction monitoring image by a vehicle state feature extraction unit of the construction state recognition model to obtain a vehicle state feature vector; Processing the tunnel construction monitoring image by a personnel vehicle position feature extraction unit of the construction state recognition model to obtain a personnel vehicle position feature vector; Performing feature fusion on the personnel action feature vector, the vehicle state feature vector and the personnel vehicle position feature vector to obtain a fused feature vector; The fused feature vector is input into the classification processing unit of the construction status recognition model to obtain the construction status recognition result of the target position output by the classification processing unit.

3. The method according to claim 2, characterized in that The step of processing the tunnel construction monitoring image by the personnel vehicle position feature extraction unit of the construction state recognition model to obtain the personnel vehicle position feature vector comprises: The tunnel construction monitoring image is processed by the first submodule of the personnel vehicle position feature extraction unit to obtain the first position data of at least one personnel operation area of ​​the construction vehicle; the personnel operation area includes at least one of a driving area, a lifting platform area and a control console area; the first position data includes a vehicle type data item and position data items corresponding to a plurality of personnel operation areas respectively, the vehicle type data item is used to indicate the type of the construction vehicle, and each of the position data items is used to indicate the position information of one of the personnel operation areas in the image; Processing the tunnel construction monitoring image through the second submodule of the personnel vehicle position feature extraction unit to obtain second position data of the construction personnel; The first position data and the second position data are concatenated into the personnel vehicle position feature vector.

4. The method according to claim 2, characterized in that: The method further comprises: The vehicle state feature vector and the personnel vehicle position feature vector are input into a pre-trained construction risk prediction module to obtain a construction risk detection result output by the construction risk prediction module.

5. The method according to claim 2, characterized in that: The method further comprises: Determining a moving direction of a construction vehicle according to the tunnel construction monitoring image of the target location; The traffic lights in the tunnel are controlled according to the construction status recognition result of the target position and the moving direction of the construction vehicle.

6. The method according to claim 2, characterized in that The method further comprises: Determine corresponding construction progress prediction results according to the construction status identification results corresponding to multiple time points; The entry order sequence of the construction vehicles to enter the tunnel is adjusted according to the construction progress prediction result.

7. The method according to claim 3, characterized in that The method further comprises: Acquire a first training sample, where the first training sample includes a training image and a first label, where the first label includes location frame information and human key point information of a construction worker in the training image; Training the personnel action feature extraction unit through the first training sample; Acquire a second training sample, wherein the second training sample includes a sample image and a second label, wherein the second label includes the type of the construction vehicle and key point information of the equipment in the sample image; Training the vehicle state feature extraction unit using the second training sample; Acquire a third training sample, wherein the third training sample includes a sample image and a third label, wherein the third label includes a position of a construction worker, a position of a construction vehicle, and a personnel operation area of ​​the construction vehicle marked in the sample image; The personnel carrier position feature extraction unit is trained using the third training sample.

8. The method according to claim 1, characterized in that The step of determining the construction process stage according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result, and the second weight corresponding to the second process identification result includes: Calculate the first The fusion recognition execution degree corresponding to the preset construction process : in, In the identification result of the first process, The first recognition confidence corresponding to a preset construction process; The first step identification result is The first weight corresponding to the preset construction process; In the second process identification result, The second recognition confidence corresponding to a preset construction process; The second process identification result is The second weight corresponding to the preset construction process; The preset construction process with the highest degree of fusion recognition execution is determined as the current construction process stage.

9. A tunnel construction process management device based on AI recognition, characterized in that: The device comprises: An image acquisition module, used to acquire facial images of construction workers through an image acquisition device; A first recognition module, used for inputting the face image into a person recognition model for processing, and obtaining a person recognition result output by the person recognition model; A second identification module, used to determine the first process identification result according to the personnel identification result and the preset correspondence between personnel and processes; The radio frequency acquisition module is used to obtain the identity information of the construction vehicles entering the tunnel through the radio frequency identification device; A third identification module, used to determine the second process identification result according to the identity information of the construction vehicle and the preset correspondence between the vehicle and the process; The process judgment module is used to determine the construction process stage according to the first process identification result, the first weight corresponding to the first process identification result, the second process identification result and the second weight corresponding to the second process identification result.

10. A tunnel construction process management system based on AI recognition, characterized in that: The system includes an image acquisition device, a radio frequency identification device and a control device; The image acquisition device is used to obtain facial images of construction workers; The radio frequency identification device is used to obtain the identity information of the construction vehicle entering the tunnel; The control device is used to input the facial image into a personnel recognition model for processing to obtain a personnel recognition result output by the personnel recognition model; determine a first process recognition result based on the personnel recognition result and a preset correspondence between personnel and processes; determine a second process recognition result based on the identity information of the construction vehicle and a preset correspondence between the vehicle and processes; determine a construction process stage based on the first process recognition result, a first weight corresponding to the first process recognition result, the second process recognition result, and a second weight corresponding to the second process recognition result.

Images