Automatic identification method and system based on YOLOv5

Abstract

The invention provides an automatic identification method and system based on YOLOv5. The method comprises the steps of obtaining an initial monitoring image corresponding to a real railway scene, performing image recognition on the initial monitoring image, and determining a moving element located in the initial monitoring image; in response to identification of any moving element in the initial monitoring image, determining a wearing attribute corresponding to the moving element, the wearing attribute including a normal wearing attribute and an abnormal wearing attribute; carrying out evacuation planning on the mobile element with the abnormal wearing attribute based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; and obtaining a current monitoring image corresponding to the real railway scene based on a preset interval time, carrying out image recognition on the current monitoring image, and determining an evacuation state corresponding to the moving element. The invention at least improves the evacuation efficiency.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to an automatic recognition method and system based on YOLOv5. Background Art

[0002] With the rapid development of railway transportation in my country, railway safety issues have become increasingly prominent, but there are still cases where non-railway workers enter railway areas.

[0003] The inventors found in their research that existing recognition technology is unable to identify whether the person appearing in the surveillance image is a railway staff member, and is unable to quickly plan an evacuation route and instruct non-railway staff members to evacuate accordingly. Summary of the invention

[0004] Based on the above problems, the present invention is proposed to provide an automatic recognition method and system based on YOLOv5 that overcomes the above problems or at least partially solves the above problems.

[0005] According to one aspect of the present invention, there is provided an automatic recognition method based on YOLOv5, comprising the following steps: Acquire an initial monitoring image corresponding to a real railway scene, and perform image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image; In response to identifying any moving element in the initial monitoring image, determining a wearing attribute corresponding to the moving element, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute; Performing evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; A current monitoring image corresponding to the actual railway scene is acquired based on a preset interval time, and image recognition is performed on the current monitoring image to determine an evacuation state corresponding to the moving element.

[0006] Optionally, in the method according to the present invention, obtaining an initial monitoring image corresponding to a real railway scene, performing image recognition on the initial monitoring image, and determining a moving element located in the initial monitoring image includes: Acquire an element training sample set, wherein the element training sample set includes a plurality of element training images corresponding to different moving elements; Training a pre-built YOLOv5 model based on the element training sample set to obtain a mobile element training model; An initial monitoring image corresponding to a real railway scene is obtained, and the initial monitoring image is input into the moving element training model to determine the moving elements in the initial monitoring image.

[0007] Optionally, in the method according to the present invention, performing evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element includes: determining an initial position of the mobile element based on the initial monitoring image; Retrieve the time monitoring table to determine the safe evacuation time and evacuation route; In response to the generation of the evacuation path, a voice broadcasting device is triggered to broadcast the evacuation path.

[0008] Optionally, in the method according to the present invention, retrieving the time monitoring table to determine the safe evacuation time and evacuation path includes: Acquire an initial time corresponding to the initial monitoring image; Retrieving the time monitoring table to determine each arrival time; Based on the difference calculation between each arrival time and the initial time, the obtained time difference corresponding to the minimum difference is determined as the evacuation time; Retrieving a preset safety evacuation coefficient, and determining the product of the preset safety evacuation coefficient and the evacuation time as the safety evacuation time; An evacuation path corresponding to the moving element is determined based on a preset direction recognition strategy.

[0009] Optionally, in the method according to the present invention, determining an evacuation path corresponding to the mobile element based on a preset direction recognition strategy includes: Performing image recognition on the initial monitoring image to determine various environmental elements in the initial monitoring image, wherein the environmental elements include an exit environment, a road environment, a forest environment, and a railway environment; Establishing an initial display layer corresponding to the initial monitoring image, and performing fuzzy processing on environmental elements except for the exit environment and the road environment in the initial display layer; An evacuation path is generated through each of the road environments, with the initial position as an indicated starting point and the exit environment as an indicated end point.

[0010] Optionally, in the method according to the present invention, in response to the generation of the evacuation path, triggering a voice broadcasting device to broadcast the evacuation path includes: In response to the generation of the evacuation path, retrieving a preset evacuation time; When the safe evacuation time is less than the preset evacuation time, the voice broadcasting device is triggered to broadcast a stop movement voice; When the safe evacuation time is greater than or equal to the evacuation stay time, the voice broadcasting device is triggered to broadcast the evacuation route.

[0011] Optionally, in the method according to the present invention, when the time difference is greater than or equal to the preset evacuation time, triggering the voice broadcasting device to broadcast the evacuation route includes: Retrieving a path broadcast template corresponding to the evacuation path, wherein the path broadcast template includes standard text characters and position description slots, time description slots, and direction description slots interspersed between the standard text characters; Fill the initial position, the safe evacuation time, and the evacuation path into the corresponding position description slot, time description slot, and direction description slot respectively to obtain the current path report; The voice broadcasting device is triggered to broadcast the current path.

[0012] Optionally, in the method according to the present invention, acquiring a current monitoring image corresponding to the actual railway scene based on a preset interval time, and performing image recognition on the current monitoring image to determine an evacuation state corresponding to the moving element includes: Retrieving the preset interval time, and performing a sum calculation based on the preset interval time and the initial time to obtain a secondary monitoring time; In response to the current moment being the secondary monitoring time, acquiring a current monitoring image corresponding to the actual railway scene; Performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state; When the moving element is not located in the preset area, the evacuation state corresponding to the moving element is determined as an abnormal state, the voice broadcast device is triggered to repeat the current path broadcast, and the current monitoring image is sent to the management terminal for display.

[0013] Optionally, in the method according to the present invention, performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state, comprises: Determining a moving distance corresponding to the preset interval time based on the preset interval time; Retrieving a preset distance coefficient, and determining the product of the preset distance coefficient and the moving distance as a reference radius; Generating a moving connection line with a moving distance along the extension direction of the evacuation path with the initial position as a starting point, and determining an end point of the moving connection line as a connection end point; A preset area with a reference radius is generated with the connection end point as a circle.

[0014] According to another aspect of the present invention, there is provided an automatic recognition system based on YOLOv5, comprising: An image recognition module is configured to obtain an initial monitoring image corresponding to a real railway scene, and perform image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image; an attribute determination module, configured to determine, in response to identifying any moving element in the initial monitoring image, a wearing attribute corresponding to the moving element, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute; A path planning module is configured to perform evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; The state determination module is configured to obtain a current monitoring image corresponding to the actual railway scene based on a preset interval time, and perform image recognition on the current monitoring image to determine an evacuation state corresponding to the moving element.

[0015] According to the solution of the present invention, the server will obtain the initial monitoring image corresponding to the actual railway scene through the monitoring equipment set on both sides of the railway, and perform image recognition on the initial monitoring image, so as to determine the mobile elements located in the initial monitoring image and the wearing attributes of the mobile elements. Then, the server will make corresponding evacuation plans for the mobile elements with abnormal wearing attributes according to the preset evacuation strategy, so as to obtain the evacuation path corresponding to the mobile elements, so as to improve the evacuation efficiency to a certain extent. Next, the server will obtain the current monitoring image corresponding to the actual railway scene according to the preset interval time, and perform image recognition on the current monitoring image, so as to determine the evacuation status of the mobile element and ensure the safety of the corresponding personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of an automatic recognition method based on YOLOv5 according to an embodiment of the present invention is shown; Figure 2 A schematic diagram showing an evacuation path according to an embodiment of the present invention is shown; Figure 3 A schematic diagram showing a preset area according to an embodiment of the present invention is shown; Figure 4 A structural block diagram of an automatic recognition system based on YOLOv5 according to another embodiment of the present invention is shown. DETAILED DESCRIPTION

[0017] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0018] With the rapid development of railway transportation in my country, railway safety issues have become increasingly prominent, but there are still cases where non-railway workers enter railway areas.

[0019] The inventors found in their research that existing recognition technology is unable to identify whether the person appearing in the surveillance image is a railway staff member, and is unable to quickly plan an evacuation route and instruct non-railway staff members to evacuate accordingly.

[0020] In order to solve the problems existing in the above-mentioned prior art, the inventor proposes the solution of the present invention. An embodiment of the present invention provides an automatic recognition method based on YOLOv5, which can be executed in a computing device.

[0021] Figure 1 A flowchart of an automatic recognition method based on YOLOv5 according to an embodiment of the present invention is shown, and the method is suitable for being executed in a computing device.

[0022] like Figure 1 As shown, the automatic recognition method based on YOLOv5 proposed in this embodiment starts from step S102. In step S102, the following contents are included: An initial monitoring image corresponding to a real railway scene is acquired, and image recognition is performed on the initial monitoring image to determine moving elements located in the initial monitoring image.

[0023] For example, in this embodiment, the server obtains an initial monitoring image corresponding to the actual railway scene through monitoring devices installed on both sides of the railway, and performs image recognition on the initial monitoring image to determine the people, that is, the moving elements, located in the initial monitoring image.

[0024] Furthermore, the above-mentioned “obtaining an initial monitoring image corresponding to the actual railway scene, and performing image recognition on the initial monitoring image to determine the moving elements in the initial monitoring image” further includes the following steps: Acquire an element training sample set, wherein the element training sample set includes a plurality of element training images corresponding to different moving elements; Training a pre-built YOLOv5 model based on the element training sample set to obtain a mobile element training model; An initial monitoring image corresponding to a real railway scene is acquired, and the initial monitoring image is input into the moving element training model to determine the moving elements located in the initial monitoring image.

[0025] For example, in this embodiment, the mobile element training model can be obtained by performing corresponding model training on a pre-built image recognition model. Before performing the model training, an element training sample set can be obtained in advance. The element training sample set can include multiple element training images corresponding to different mobile elements.

[0026] The server will then train the pre-built image recognition model using the element training sample set to obtain a mobile element training model. Next, the server will obtain an initial monitoring image corresponding to the actual railway scene and input the obtained initial monitoring image into the mobile element training model. At this point, the mobile element training model will perform image recognition on the initial monitoring image to determine each mobile element in the initial monitoring image.

[0027] In step S104, the following contents are included: In response to the recognition of any moving element in the initial monitoring image, a wearing attribute corresponding to the moving element is determined, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute.

[0028] For example, in this embodiment, when any moving element is identified in the initial monitoring image, the server can make a corresponding determination as to whether the moving element is wearing work clothes or a reflective vest.

[0029] When the server recognizes that the mobile element is wearing work clothes or a reflective vest, it will determine the wearing attributes corresponding to the mobile element as normal wearing attributes; when the server recognizes that the mobile element is not wearing work clothes or a reflective vest, it means that the mobile element may not be a railway staff member. In order to ensure the safety of the corresponding personnel, the server will determine the wearing attributes corresponding to the mobile element as abnormal wearing attributes, so as to facilitate the subsequent persuasion of the mobile element.

[0030] In step S106, the following contents are included: Based on a preset evacuation strategy, an evacuation plan is performed on the mobile element with abnormal wearing attributes to obtain an evacuation path corresponding to the mobile element.

[0031] For example, in this embodiment, the server will perform corresponding evacuation planning for the mobile element with abnormal wearing attributes according to the preset evacuation strategy, so as to obtain an evacuation path corresponding to the mobile element, so as to improve the evacuation efficiency to a certain extent.

[0032] Furthermore, the above-mentioned “performing an evacuation plan for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element” further includes the following steps: determining an initial position of the mobile element based on the initial monitoring image; Retrieve the time monitoring table to determine the safe evacuation time and evacuation route; In response to the generation of the evacuation path, a voice broadcasting device is triggered to broadcast the evacuation path.

[0033] For example, in this embodiment, the server will determine the specific location of the mobile element based on the initial monitoring image, that is, the initial location. Then, the server will call the time monitoring table, which contains the specific time when each train arrives at each location, so as to determine how long the mobile element needs to complete the evacuation, that is, the safe evacuation time, and thus the server will determine the safe evacuation time and evacuation path of the mobile element.

[0034] When the evacuation path is generated, the server will trigger the voice broadcast device to broadcast the evacuation path accordingly, so that the mobile elements can evacuate according to the broadcast content.

[0035] Furthermore, the above-mentioned "retrieve the time monitoring table to determine the safe evacuation time and evacuation route" also includes the following steps: Acquire an initial time corresponding to the initial monitoring image; Retrieving the time monitoring table to determine each arrival time; Based on the difference calculation between each arrival time and the initial time, the obtained time difference corresponding to the minimum difference is determined as the evacuation time; Retrieving a preset safety evacuation coefficient, and determining the product of the preset safety evacuation coefficient and the evacuation time as the safety evacuation time; An evacuation path corresponding to the moving element is determined based on a preset direction recognition strategy.

[0036] For example, in this embodiment, the server obtains the monitoring time of the initial monitoring image, that is, the initial time, and then retrieves the time monitoring table to determine the arrival time of each train.

[0037] Next, the server will calculate the difference between each arrival time and the initial time. Since the train corresponding to the minimum difference will arrive at the initial point first, the server will determine the time difference corresponding to the minimum difference as the evacuation time, that is, the mobile element needs to complete the evacuation within the evacuation time.

[0038] The ideal evacuation effect is that the mobile element can be evacuated before the train arrives, so the server will call the preset safe evacuation coefficient and multiply the preset safe evacuation coefficient with the evacuation time, and then determine the product as the safe evacuation time to ensure that the mobile element can be evacuated before the train arrives. Then, the server will determine the evacuation path of the mobile element according to the preset direction recognition strategy.

[0039] Furthermore, the above-mentioned “determining the evacuation path corresponding to the mobile element based on a preset direction recognition strategy” further includes the following steps: Performing image recognition on the initial monitoring image to determine various environmental elements in the initial monitoring image, wherein the environmental elements include an exit environment, a road environment, a forest environment, and a railway environment; Establishing an initial display layer corresponding to the initial monitoring image, and performing fuzzy processing on environmental elements except for the exit environment and the road environment in the initial display layer; An evacuation path is generated through each of the road environments, with the initial position as an indicated starting point and the exit environment as an indicated end point.

[0040] For example, in this embodiment, the server performs image recognition on the initial monitoring image to determine the various environmental elements located in the initial monitoring image, such as a walkable road environment, a railway environment located in the railway area, a forest environment with large trees planted in groups, and an exit environment located in the area connecting the railway area and the external road.

[0041] Next, the server will create an initial display layer corresponding to the initial monitoring image, and blur the environmental elements except the exit environment and the road environment in the initial display layer to facilitate the subsequent planning of the evacuation route.

[0042] Then, the server will generate an evacuation path through various road environments with the initial location as the starting point and the exit environment as the end point, such as Figure 2 shown.

[0043] Furthermore, the above-mentioned “in response to the generation of the evacuation path, triggering the voice broadcasting device to broadcast the evacuation path” further includes the following steps: In response to the generation of the evacuation path, retrieving a preset evacuation time; When the safe evacuation time is less than the preset evacuation time, the voice broadcasting device is triggered to broadcast a stop movement voice; When the safe evacuation time is greater than or equal to the evacuation stay time, the voice broadcasting device is triggered to broadcast the evacuation route.

[0044] For example, in this embodiment, after the evacuation route is generated, the server will retrieve the preset evacuation time.

[0045] When the safe evacuation time is less than the preset evacuation time, it means that the safe evacuation time is too short and the moving element may not be able to complete the evacuation before the train arrives. Therefore, the server will trigger the voice broadcast device to broadcast the stop movement voice, so that the moving element remains motionless at the initial position to avoid danger.

[0046] When the safe evacuation time is greater than or equal to the evacuation stay time, it means that the safe evacuation time is sufficient and the mobile element can complete the evacuation before the train arrives, so the server triggers the voice broadcast device to broadcast the evacuation path.

[0047] Furthermore, the above-mentioned “when the time difference is greater than or equal to the preset evacuation time, triggering the voice broadcasting device to broadcast the evacuation route” further includes the following steps: Retrieving a path broadcast template corresponding to the evacuation path, wherein the path broadcast template includes standard text characters and position description slots, time description slots, and direction description slots interspersed between the standard text characters; Fill the initial position, the safe evacuation time, and the evacuation path into the corresponding position description slot, time description slot, and direction description slot respectively to obtain the current path report; The voice broadcasting device is triggered to broadcast the current path.

[0048] For example, in this embodiment, the server will retrieve a path broadcast template corresponding to the evacuation path, in which there are standard text characters and position description slots, time description slots, and direction description slots interspersed between the standard text characters.

[0049] The server will fill the initial position, safe evacuation time, and evacuation path into the corresponding position description slot, time description slot, and direction description slot respectively, thereby obtaining the current path broadcast and triggering the voice broadcast device to broadcast the current path broadcast. For example, the current path broadcast is: To ensure your safety, people on the left side of Track 1 are requested to walk forward 300 meters within five minutes to evacuate.

[0050] In step S108, the following contents are included: A current monitoring image corresponding to the actual railway scene is acquired based on a preset interval time, and image recognition is performed on the current monitoring image to determine an evacuation state corresponding to the moving element.

[0051] For example, in this embodiment, the server acquires a current monitoring image corresponding to the actual railway scene according to a preset interval, and performs image recognition on the current monitoring image to determine the evacuation status of the mobile element.

[0052] Furthermore, the above-mentioned “acquiring a current monitoring image corresponding to the actual railway scene based on a preset interval, and performing image recognition on the current monitoring image to determine the evacuation state corresponding to the moving element” further includes the following steps: Retrieving the preset interval time, and performing a sum calculation based on the preset interval time and the initial time to obtain a secondary monitoring time; In response to the current moment being the secondary monitoring time, acquiring a current monitoring image corresponding to the actual railway scene; Performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state; When the moving element is not located in the preset area, the evacuation state corresponding to the moving element is determined as an abnormal state, the voice broadcast device is triggered to repeat the current path broadcast, and the current monitoring image is sent to the management terminal for display.

[0053] For example, in this embodiment, in order to ensure that personnel can evacuate in time, the server will retrieve the preset interval time, and then sum the preset interval time and the initial time to obtain the secondary monitoring time.

[0054] When the current moment arrives at the second monitoring time, the server will obtain the current monitoring image corresponding to the actual railway scene and perform image recognition on the current monitoring image. If the server recognizes that the moving element is located in the preset area, it means that the moving element is evacuating accordingly according to the evacuation path, and the server will determine the evacuation status of the moving element as normal.

[0055] If the server recognizes that the mobile element is not within the preset area, it means that the mobile element has not evacuated according to the evacuation path. Therefore, the server will determine the evacuation status of the mobile element as an abnormal state and trigger the voice broadcast device to repeat the current path broadcast. At the same time, the server will send the current monitoring image to the management end for display, so that the management end can send staff in time to persuade the mobile element to leave and ensure the safety of the corresponding personnel.

[0056] Furthermore, the above-mentioned “performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state” further includes the following steps: Determining a moving distance corresponding to the preset interval time based on the preset interval time; Retrieving a preset distance coefficient, and determining the product of the preset distance coefficient and the moving distance as a reference radius; Generating a moving connection line with a moving distance along the extension direction of the evacuation path with the initial position as a starting point, and determining an end point of the moving connection line as a connection end point; A preset area with a reference radius is generated with the connection end point as a circle.

[0057] For example, in this embodiment, the server determines the distance that the mobile element can move in the preset interval time, that is, the moving distance, according to the preset interval time, and then retrieves the preset distance coefficient, and determines the product of the preset distance coefficient and the moving distance as the reference radius.

[0058] At this time, the server will generate a moving connection line with a length equal to the moving distance along the extension direction of the evacuation path with the initial position as the starting point, and determine the end point of the moving connection line as the connection end point, thereby generating a circle with a reference radius with the connection end point as the circle point, that is, the preset area, such as Figure 3 shown.

[0059] According to the solution of the present invention, the server will obtain the initial monitoring image corresponding to the actual railway scene through the monitoring equipment set on both sides of the railway, and perform image recognition on the initial monitoring image, so as to determine the mobile elements located in the initial monitoring image and the wearing attributes of the mobile elements. Then, the server will make corresponding evacuation plans for the mobile elements with abnormal wearing attributes according to the preset evacuation strategy, so as to obtain the evacuation path corresponding to the mobile elements, so as to improve the evacuation efficiency to a certain extent. Next, the server will obtain the current monitoring image corresponding to the actual railway scene according to the preset interval time, and perform image recognition on the current monitoring image, so as to determine the evacuation status of the mobile element and ensure the safety of the corresponding personnel.

[0060] Another embodiment of the present invention provides an automatic recognition system based on YOLOv5, Figure 4 As a corresponding device block diagram, the system includes: An image recognition module is configured to obtain an initial monitoring image corresponding to a real railway scene, and perform image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image; an attribute determination module, configured to determine, in response to identifying any moving element in the initial monitoring image, a wearing attribute corresponding to the moving element, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute; A path planning module is configured to perform evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; The state determination module is configured to obtain a current monitoring image corresponding to the actual railway scene based on a preset interval time, and perform image recognition on the current monitoring image to determine an evacuation state corresponding to the moving element.

[0061] In the description provided herein, algorithms and displays are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the examples of the present invention. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to implement the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the preferred embodiment of the present invention.

[0062] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0063] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof.

[0064] Those skilled in the art will appreciate that the modules or units or components of the devices in the examples disclosed herein may be arranged in the devices described in the embodiment, or alternatively may be located in one or more devices different from the devices in the examples. The modules in the foregoing examples may be combined into one module or may be divided into multiple submodules.

[0065] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore may be divided into a plurality of submodules or subunits or subcomponents.

[0066] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present invention and to form different embodiments.

[0067] In addition, some of the embodiments are described herein as methods or combinations of method elements that can be implemented by a processor of a computer system or by other devices that perform the functions. Therefore, a processor with necessary instructions for implementing the method or method elements forms a device for implementing the method or method elements. In addition, the elements described herein of the device embodiments are examples of devices for implementing the functions performed by the elements for the purpose of implementing the invention.

[0068] As used herein, unless otherwise specified, the use of ordinal numbers "first," "second," "third," etc. to describe common objects merely indicates that different instances of similar objects are involved, and is not intended to imply that the objects so described must have a given order in time, space, order, or in any other manner.

[0069] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is primarily selected for readability and instructional purposes, rather than for explaining or limiting the subject matter of the present invention.

Claims

1. An automatic recognition method based on YOLOv5, characterized in that: The following steps are involved: Acquire an initial monitoring image corresponding to a real railway scene, and perform image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image; In response to identifying any moving element in the initial monitoring image, determining a wearing attribute corresponding to the moving element, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute; Performing evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; A current monitoring image corresponding to the actual railway scene is acquired based on a preset interval time, and image recognition is performed on the current monitoring image to determine an evacuation state corresponding to the moving element.

2. The automatic recognition method based on YOLOv5 according to claim 1, characterized in that Acquiring an initial monitoring image corresponding to a real railway scene, and performing image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image, including: Acquire an element training sample set, wherein the element training sample set includes a plurality of element training images corresponding to different moving elements; Training a pre-built YOLOv5 model based on the element training sample set to obtain a mobile element training model; An initial monitoring image corresponding to a real railway scene is obtained, and the initial monitoring image is input into the moving element training model to determine the moving elements in the initial monitoring image.

3. The automatic recognition method based on YOLOv5 according to claim 1, characterized in that: Based on a preset evacuation strategy, an evacuation plan is performed for a mobile element with abnormal wearing attributes to obtain an evacuation path corresponding to the mobile element, including: determining an initial position of the mobile element based on the initial monitoring image; Retrieve the time monitoring table to determine the safe evacuation time and evacuation route; In response to the generation of the evacuation path, a voice broadcasting device is triggered to broadcast the evacuation path.

4. The automatic recognition method based on YOLOv5 according to claim 3, characterized in that: Retrieve the time monitoring table to determine the safe evacuation time and evacuation route, including: Acquire an initial time corresponding to the initial monitoring image; Retrieving the time monitoring table to determine each arrival time; Based on the difference calculation between each arrival time and the initial time, the obtained time difference corresponding to the minimum difference is determined as the evacuation time; Retrieving a preset safety evacuation coefficient, and determining the product of the preset safety evacuation coefficient and the evacuation time as the safety evacuation time; An evacuation path corresponding to the moving element is determined based on a preset direction recognition strategy.

5. The automatic recognition method based on YOLOv5 according to claim 3, characterized in that: Determining an evacuation path corresponding to the moving element based on a preset direction recognition strategy includes: Performing image recognition on the initial monitoring image to determine various environmental elements in the initial monitoring image, wherein the environmental elements include an exit environment, a road environment, a forest environment, and a railway environment; Establishing an initial display layer corresponding to the initial monitoring image, and performing fuzzy processing on environmental elements except for the exit environment and the road environment in the initial display layer; An evacuation path is generated through each of the road environments, with the initial position as an indicated starting point and the exit environment as an indicated end point.

6. The automatic recognition method based on YOLOv5 according to claim 3, characterized in that: In response to the generation of the evacuation path, triggering a voice broadcasting device to broadcast the evacuation path includes: In response to the generation of the evacuation path, retrieving a preset evacuation time; When the safe evacuation time is less than the preset evacuation time, the voice broadcasting device is triggered to broadcast a stop movement voice; When the safe evacuation time is greater than or equal to the evacuation stay time, the voice broadcasting device is triggered to broadcast the evacuation route.

7. The automatic recognition method based on YOLOv5 according to claim 6, characterized in that: When the time difference is greater than or equal to the preset evacuation time, the voice broadcasting device is triggered to broadcast the evacuation route, including: Retrieving a path broadcast template corresponding to the evacuation path, wherein the path broadcast template includes standard text characters and position description slots, time description slots, and direction description slots interspersed between the standard text characters; Fill the initial position, the safe evacuation time, and the evacuation path into the corresponding position description slot, time description slot, and direction description slot respectively to obtain the current path report; The voice broadcasting device is triggered to broadcast the current path.

8. The automatic recognition method based on YOLOv5 according to claim 1, characterized in that: Acquiring a current monitoring image corresponding to the actual railway scene based on a preset interval, and performing image recognition on the current monitoring image to determine an evacuation state corresponding to the moving element, including: Retrieving the preset interval time, and performing a sum calculation based on the preset interval time and the initial time to obtain a secondary monitoring time; In response to the current moment being the secondary monitoring time, acquiring a current monitoring image corresponding to the actual railway scene; Performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state; When the moving element is not located in the preset area, the evacuation state corresponding to the moving element is determined as an abnormal state, the voice broadcast device is triggered to repeat the current path broadcast, and the current monitoring image is sent to the management terminal for display.

9. The automatic recognition method based on YOLOv5 according to claim 1, characterized in that: Performing image recognition on the current monitoring image, and when the moving element is located in a preset area, determining the evacuation state corresponding to the moving element as a normal state, includes: Determining a moving distance corresponding to the preset interval time based on the preset interval time; Retrieving a preset distance coefficient, and determining the product of the preset distance coefficient and the moving distance as a reference radius; Generating a moving connection line with a moving distance along the extension direction of the evacuation path with the initial position as a starting point, and determining an end point of the moving connection line as a connection end point; A preset area with a reference radius is generated with the connection end point as a circle.

10. An automatic recognition system based on YOLOv5, characterized in that: include: An image recognition module is configured to obtain an initial monitoring image corresponding to a real railway scene, and perform image recognition on the initial monitoring image to determine a moving element located in the initial monitoring image; an attribute determination module, configured to determine, in response to identifying any moving element in the initial monitoring image, a wearing attribute corresponding to the moving element, wherein the wearing attribute includes a normal wearing attribute and an abnormal wearing attribute; A path planning module is configured to perform evacuation planning for a mobile element with abnormal wearing attributes based on a preset evacuation strategy to obtain an evacuation path corresponding to the mobile element; The state determination module is configured to obtain a current monitoring image corresponding to the actual railway scene based on a preset interval time, and perform image recognition on the current monitoring image to determine an evacuation state corresponding to the moving element.

Images