WO2023279785A1 - Method and apparatus for detecting whether staff member is compatible with post, computer device, and storage medium - Google Patents

Abstract

The present disclosure provides a method and apparatus for detecting whether a staff member is compatible with a post, a computer device, and a storage medium. The method comprises: obtaining a video to be processed of a target time period obtained by image capturing a staff member of a target post; performing face recognition and action recognition on the video to obtain a face recognition result and an action recognition result of the staff member; associating the face recognition result with the action recognition result to obtain an association; and determining, on the basis of the association and post information corresponding to the target post, whether the staff member is compatible with the target post.

Description

Person-post matching detection method, device, computer equipment and storage medium

Cross References to Related Art

This disclosure claims the priority of Chinese patent application No. 202110763737.9 filed on July 6, 2021, the entire content of which is incorporated herein by reference.

technical field

The present disclosure relates to the technical field of image processing, and in particular to a human-post matching detection method, device, computer equipment, and storage medium.

Background technique

In the process of industrial production, it is mainly through human management to check whether the personnel match the post. This method not only consumes manpower and material resources, but also depends too much on the subjective consciousness of the managers to obtain the data on whether the personnel match the post.

Contents of the invention

Embodiments of the present disclosure at least provide a person-post matching detection method, device, computer equipment, and storage medium.

In the first aspect, an embodiment of the present disclosure provides a method for detecting person-post matching, including: acquiring a video to be processed at a target time period obtained by shooting a staff member at a target position; performing face recognition and action on the video to be processed Recognizing, obtaining the face recognition result and action recognition result of the staff member; associating the face recognition result with the action recognition result to obtain an association relationship; based on the association relationship and the position information corresponding to the target position, A determination is made as to whether the worker is a match for the target position.

In this way, by detecting whether the staff in the target positions match the target positions, the identities of the staff in each target position can be known, so as to know whether the staff and positions match, so as to achieve the purpose of personnel-post matching detection .

In an optional implementation manner, the position information includes personnel identity information corresponding to the target position in the target time period; based on the association relationship and the position information corresponding to the target position, the determination of the Whether the staff and the target post match, including: when the post information includes the personal identity information, in response to the face recognition result matching the personal identity information, obtain the staff and the target Job matching.

In this way, since the method of using the face recognition result for matching is relatively simple, based on the personnel identity information, it is relatively easy and fast to determine whether the staff matches the target position.

In an optional implementation manner, the post information includes a reference action sequence corresponding to the target post; based on the association relationship and the post information corresponding to the target post, it is determined that the staff member is related to the target post. Whether the target position matches includes: when the position information includes the reference action sequence, in response to the match between the action recognition result and the reference action sequence, it is obtained that the staff member matches the target position.

In this way, based on the reference action sequence, it can be more accurately determined whether the staff completes the production action according to the work requirements, and while it can be accurately determined whether the staff matches the target position, it can further ensure that the staff is in production. safety.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes at least one of the following: matching of working hours, matching of production norms, and matching of production efficiency.

In this way, using matching methods in different dimensions can ensure high-quality and high-efficiency production of products in different aspects.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching of working hours; the method further includes: based on the action recognition result, determining that the worker performs a production action The action duration; based on the action duration and the reference action duration corresponding to the reference action sequence, determine the matching result of the working duration.

In this way, through the matching of working hours, products can be produced at a faster speed.

In an optional implementation manner, the action duration of the worker performing the production action includes at least one of the following: the total duration of the worker performing the production action; the interval time between the worker performing adjacent production actions ; The sub-duration of the worker executing the production sub-action; the interval between the worker's execution of adjacent production sub-actions.

In this way, through the control of different time lengths, it can be ensured that the staff can carry out the relevant production actions efficiently and orderly.

In an optional implementation manner, the action recognition result includes: a recognition action sequence; the recognition action sequence includes: multiple sub-actions; the action recognition result matches the reference action sequence, including: producing normative Matching; the method further includes: matching the recognition action sequence with the reference action sequence to obtain the production normative matching result.

In this way, through the matching of action sequences, it can further ensure that the production is carried out in accordance with a certain order, and while ensuring the quality of the product, the safety of the staff can also be guaranteed.

In an optional implementation manner, the matching of the recognition action sequence and the reference action sequence to obtain the production normative matching result includes: dividing the recognition action sequence into multiple actions Subsequence: matching the reference action sequence with each action subsequence; obtaining the production normative matching result based on the matching results of the reference action sequence and each action subsequence.

In an optional implementation manner, the dividing the recognition action sequence into multiple action subsequences includes: from the recognition action sequence, determining A target sub-action; dividing the recognition action sequence into at least one action sub-sequence based on the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence .

In this way, by using the target sub-action and the target reference action to segment different action sequences, it is possible to obtain a production-standard matching result more accurately.

In an optional implementation manner, the production normative matching results include at least one of the following: correct content of production actions, wrong content of production actions, missing production actions, redundant production actions, and wrong order of actions in production actions .

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching work efficiency; the method further includes: determining that the worker performs a production action based on the action recognition result The total duration and/or interval duration; based on the total duration and/or interval duration of the staff performing production actions, and the reference action duration, determine the work efficiency of the staff; based on the work efficiency of the staff and the expected Work efficiency is assumed, and a matching result of the work efficiency is determined.

In this way, the efficiency of production can also be guaranteed.

In an optional implementation manner, the person-post matching detection method further includes: based on the face recognition result, determining at least one of the on-duty time and the off-duty time of the staff member; The on-duty time and/or the off-duty time are used to determine the attendance information of the staff member.

In the second aspect, the embodiment of the present disclosure also provides a person-post matching detection device, including: an acquisition module configured to acquire a video to be processed at a target time period obtained by shooting a staff member in a target position; an identification module configured to Perform face recognition and action recognition on the video to be processed to obtain the face recognition result and action recognition result of the staff; the association module is configured to associate the face recognition result and the action recognition result to obtain An association relationship; a determination module configured to determine whether the staff member matches the target position based on the association relationship and the position information corresponding to the target position.

In an optional implementation manner, the position information includes personnel identity information corresponding to the target position in the target time period; the determining module is based on the association relationship and the position information corresponding to the target position, When determining whether the staff member matches the target post, it is configured to: when the post information includes the personal identity information, in response to the face recognition result matching the personal identity information, obtain the Workers are matched to the stated target positions.

In an optional implementation manner, the post information includes a reference action sequence corresponding to the target post; the determining module determines the staff member based on the association relationship and the post information corresponding to the target post. When matching with the target position, it is configured to: when the position information includes the reference action sequence, in response to the action recognition result matching the reference action sequence, obtain the staff and the target Job matching.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes at least one of the following: matching of working hours, matching of production norms, and matching of production efficiency.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching of working hours; the determination module is further configured to: determine the staff member based on the action recognition result The action duration for executing the production action; based on the action duration and the reference action duration corresponding to the reference action sequence, determine the matching result of the working duration.

In an optional implementation manner, the action duration of the worker performing the production action includes at least one of the following: the total duration of the worker performing the production action; the interval time between the worker performing adjacent production actions ; The sub-duration of the worker executing the production sub-action; the interval between the worker's execution of adjacent production sub-actions.

In an optional implementation manner, the action recognition result includes: a recognition action sequence; the recognition action sequence includes: multiple sub-actions; the action recognition result matches the reference action sequence, including: producing normative Matching: the determining module is further configured to: match the recognition action sequence with the reference action sequence to obtain the production-standard matching result.

In an optional implementation manner, when the determination module matches the recognition action sequence with the reference action sequence to obtain the production normative matching result, it is configured to: match the recognition action sequence Segmenting into multiple action subsequences; matching the reference action sequence with each action subsequence; obtaining the production normative matching result based on the matching results of the reference action sequence and each action subsequence.

In an optional implementation manner, when the determining module divides the recognition action sequence into a plurality of action subsequences, it is configured to: from the recognition action sequence, determine The target sub-action matched by the target reference action; based on the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence, segment the recognition action sequence is at least one action subsequence.

In an optional implementation manner, the production normative matching results include at least one of the following: correct content of production actions, wrong content of production actions, missing production actions, redundant production actions, and wrong order of actions in production actions .

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching of work efficiency; the determination module is further configured to: determine the staff member based on the action recognition result The total duration and/or interval duration of performing production actions; based on the total duration and/or interval duration of the execution of production actions by the staff, as well as the reference action duration, determine the work efficiency of the staff; based on the work of the staff The efficiency and the preset work efficiency determine the matching result of the work efficiency.

In an optional implementation manner, the determination module is further configured to: determine at least one of the on-duty time and off-duty time of the staff member based on the face recognition result; time and/or the off-duty time, and determine the attendance information of the staff member.

In a third aspect, an optional implementation manner of the present disclosure further provides a computer device, a processor, and a memory, the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the instructions stored in the memory. machine-readable instructions, when the machine-readable instructions are executed by the processor, when the machine-readable instructions are executed by the processor, the above-mentioned first aspect is executed, or any possible implementation of the first aspect steps in the method.

In a fourth aspect, an optional implementation manner of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the above-mentioned first aspect, or any one of the first aspects in the first aspect, may be executed. Steps in one possible implementation.

In the fifth aspect, an optional implementation manner of the present disclosure further provides a computer program product, the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and when the computer program is read and executed by a computer, the above-mentioned first An aspect, or a step in any possible implementation of the first aspect.

For the effect description of the above-mentioned person-post matching detection device, computer equipment, and computer-readable storage medium, please refer to the description of the above-mentioned person-post matching detection method, which will not be repeated here.

In order to make the above-mentioned objects, features and advantages of the present disclosure more comprehensible, preferred embodiments will be described in detail below together with the accompanying drawings.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present disclosure more clearly, the following will briefly introduce the accompanying drawings used in the embodiments. The accompanying drawings here are incorporated into the specification and constitute a part of the specification. The drawings show the embodiments consistent with the present disclosure, and are used together with the description to explain the technical solution of the present disclosure. It should be understood that the following drawings only show some embodiments of the present disclosure, and therefore should not be regarded as limiting the scope. For those skilled in the art, they can also make From these drawings other related drawings are obtained.

FIG. 1 shows a flow chart of a method for detecting person-post matching provided by an embodiment of the present disclosure;

Fig. 2 shows a schematic diagram of performing action recognition on a video to be processed provided by an embodiment of the present disclosure;

Fig. 3 shows a schematic diagram of multiple reference actions provided by an embodiment of the present disclosure;

FIG. 4 shows a schematic diagram of a time axis corresponding to a production cycle provided by an embodiment of the present disclosure;

Fig. 5 shows a schematic diagram of a person-post matching detection device provided by an embodiment of the present disclosure;

Fig. 6 shows a schematic diagram of a computer device provided by an embodiment of the present disclosure.

detailed description

In order to make the purpose, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only It is a part of the embodiments of the present disclosure, but not all of them. The components of the disclosed embodiments generally described and illustrated herein may be arranged and designed in a variety of different configurations. Accordingly, the following detailed description of the embodiments of the present disclosure is not intended to limit the scope of the claimed disclosure, but merely represents selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present disclosure.

After research, it is found that there are a large number of problems in the industrial production process that the matching degree of personnel positions is not enough, such as staff waiting for posts, changing posts, leaving posts, etc. Under these circumstances, the staff cannot normally work in the corresponding positions according to the job requirements. This will easily lead to the production of product quality can not be guaranteed.

In addition, if the staff makes mistakes in their positions, it may also pose a threat to the staff's own safety.

Based on the above research, this disclosure provides a detection method for personnel-post matching. By detecting whether the staff in the target positions match with the target positions, the identities of the staff in each target position can be known, and thus the relationship between each staff and the target position can be known. Whether the positions are matched, in order to achieve the purpose of personnel-post matching detection.

In addition, according to whether the staff matches the target position, it can help the staff to work according to the normal work requirements, and improve the safety of the staff at work.

The defects in the prior art are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed by the present disclosure for the above-mentioned problems should be the result of the inventor's contribution to the present disclosure. contribution.

It should be noted that like numerals and letters denote similar items in the following figures, therefore, once an item is defined in one figure, it does not require further definition and explanation in subsequent figures.

In order to facilitate the understanding of this embodiment, a person-post matching detection method disclosed in the embodiments of the present disclosure is first introduced in detail. The person-post matching detection method provided in the embodiments of the present disclosure is generally executed by a computer with certain computing power equipment, the computer equipment includes, for example: terminal equipment or server or other processing equipment, the terminal equipment can be user equipment (User Equipment, UE), mobile equipment, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (Personal Digital Assistant) Assistant, PDA), handheld devices, computing devices, vehicle-mounted devices, wearable devices, etc. In some possible implementation manners, the method for detecting person-post matching may be implemented by a processor invoking computer-readable instructions stored in a memory.

The method for detecting person-post matching provided by the embodiments of the present disclosure will be described below.

Referring to FIG. 1 , it is a flow chart of a method for detecting person-post matching provided by an embodiment of the present disclosure. The method includes steps S101 to S104, wherein:

S101: Obtain the video to be processed in the target time period obtained by shooting the staff at the target position;

S102: Perform face recognition and action recognition on the video to be processed, and obtain a face recognition result and action recognition result of the staff member;

S103: Associating the face recognition result with the action recognition result to obtain an association relationship;

S104: Based on the association relationship and the position information corresponding to the target position, determine whether the worker matches the target position.

The person-post matching detection method provided by the embodiments of the present disclosure performs face recognition and action recognition on the to-be-processed video of the staff at the target position at the target time period, and determines the face recognition results and action recognition results of the staff, thereby using The association relationship obtained by associating the face recognition result with the action recognition result is matched with the position information corresponding to the target position, so as to determine whether the staff matches the target position. In this way, by detecting whether the staff in the target positions match the target positions, the identities of the staff in each target position can be known, so as to know whether the staff and positions match, so as to achieve the purpose of personnel-post matching detection .

The person-post matching detection method can be applied to, for example, an assembly line production workshop, or a production workshop with parallel operations. Among them, taking the assembly line production workshop as an example, the production workshop can include food processing production lines, automobile assembly production lines, pure water filling production lines, etc. Taking a food processing production line as an example, in order, the production line may include the following production steps (process): raw material screening, raw material cleaning, mixing processing, and filling. Different staffs can be equipped in different production steps. For example, in the production step of raw material screening, multiple staffs can be equipped to screen different raw materials, such as staff for screening vegetables, and staff for fruit screening. Screening staff, and staff screening spices. There may be one or more staff members for different production steps.

In addition, different production steps may also include more detailed production steps. Taking the mixing processing step as an example, it can also be refined into cutting processing of different raw materials, heating treatment, and mixing processing of different raw materials obtained after heating. The details may be determined according to the actual situation, and details are not repeated here.

In the following, in order to simplify the description, the production steps include raw material screening P1, raw material cleaning P2, and mixing processing P3, and a worker S1 is assigned for the production step of raw material screening P1 as an example.

The production step that worker S1 needs to perform is raw material screening P1, and the corresponding position of worker S1 during normal work is the position corresponding to raw material screening P1, that is, the target position of worker S1, which can be represented by A1, for example. Wherein, the target post A1 includes specific post information, for example, it may include the post name corresponding to the worker S1, such as a screener; in addition, the target post A1 may also include the specific work area of the worker S1, such as the material centralized warehouse in front of the conveyor belt.

Regarding the above S101, when acquiring the video to be processed, the target time period for shooting may be determined first. Specifically, for example, preset working hours may be used as the target time period, such as 9:00 am to 6:00 pm on weekdays. Alternatively, it is also possible to use the time period when the staff is prone to erroneous operations, the time period with low production efficiency, or the time period with high production efficiency, etc., as the target time period, such as 2:00 pm to 4:00 pm on weekdays.

After the target time period is determined, the staff at the target post can be photographed. In one possible situation, a video capture device can be directly installed in the area where the target post is located, and the area where the target post is located can be photographed within the target period; in another possible situation, if the staff work When the scope is large, for example, the staff need to go to and from different areas, the video capture device can also locate and track the staff at the target position in real time to obtain the video to be processed.

With respect to the above S102, after the video to be processed is obtained, the video to be processed can also be used to perform face recognition (Face Recognition, FR) and action recognition (Action Recognition, AR) on the staff. In a specific implementation, for example, face recognition and action recognition can be performed by means of deep learning.

Exemplarily, a pre-trained deep neural network may be used to identify workers in video frame images. Among them, since there may be multiple staff members displayed in the video frame image, when performing face recognition on the video frame image, for example, it can be determined in the video to be processed that the face of the staff member corresponding to the target position is included. Video frame image. Correspondingly, when the motion recognition is performed on the video to be processed, the motion recognition of the staff member can also be performed for the video frame image locked in the video to be processed and corresponding to the staff member corresponding to the target position. In this way, the obtained face recognition results and action recognition results are more targeted, and the accuracy is also higher when using the face recognition results and action recognition results to determine whether the staff matches the target post.

In addition, after face recognition is performed on the video to be processed, the obtained face recognition result of the staff member may include at least one of the recognized face sub-image or the matched staff related information, for example. Exemplarily, after face recognition is performed on the video to be processed corresponding to the staff S1, for example, multiple frames of sub-images containing the face of the staff S1 can be obtained, and the sub-image matching the staff S1 in the video to be processed can also be obtained. The result of face recognition, and the relevant information of the staff member S1 can be retrieved, such as job title, working hours and other information.

After performing action recognition on the video to be processed, the obtained staff action recognition result may include, for example, the recognized staff action information or matching results with multiple preset reference actions. Wherein, when performing action recognition, for example, the staff's limbs, torso, head, etc. may be identified in the video frame images included in the video to be processed, so as to determine the action recognition result. For example, refer to FIG. 2 , which is a schematic diagram of action recognition for a video to be processed provided by an embodiment of the present disclosure. Among them, (a) in FIG. 2 shows a frame of image in the video to be processed corresponding to worker S1. After performing motion recognition on the video frame image shown in (a) in FIG. 2, for example, the recognition result in (b) in FIG. The obtained sketch is used as the action recognition result; or, the basic action shown in (b) in Figure 2 can also be used to match multiple reference actions to obtain the action recognition result. Here, the obtained action recognition result can be specific to specific Action classification results, such as "walking action", "pickup action", etc. For example, referring to FIG. 3 , which is a schematic diagram of multiple reference actions provided by an embodiment of the present disclosure, the reference actions shown in FIG. 3 are shown in the form of simplified diagrams and corresponding to action identifiers. Using the result obtained in (b) of Figure 2 to match the multiple reference actions shown in Figure 3, the action recognition result "walking action" can be obtained.

Here, only some examples are shown, and the details may be determined according to actual conditions, and no limitation is made here.

With regard to the above S103, after the face recognition result and the action recognition result are determined according to the above S102, they can be matched to obtain associated information corresponding to different staff members.

Specifically, for worker S1, at least one of the sub-image of worker S1 included in the face recognition result, the "walking motion" determined in the motion recognition result, and the relevant information of worker S1 can be regarded as corresponding to worker S1. matching information, and further use the matching information to associate with the worker S1 to obtain the related information of the worker S1.

For the above S104, for different staff members, since each staff member has a specific target position, and the position information corresponding to the target position is fixed, it is possible to determine the job by using the association relationship determined in the above S103 and the position information corresponding to the target position. Whether the personnel match the target position.

In the following, how to determine whether the staff member matches the target position will be described for different position information. Among them, job information includes but not limited to the following (A) and (B):

(A): The position information includes the identity information of the person corresponding to the target position in the target time period.

In this case, when determining whether the staff matches the target position based on the association relationship and the position information corresponding to the target position, for example, the following method can be adopted: when the position information includes the When the personal identity information is mentioned, in response to the matching of the face recognition result with the personal identity information, it is obtained that the staff member matches the target position.

When the post information includes personnel identity information, for example, the face recognition result may include, for example, a sub-image containing the face of the staff, and correspondingly, the post information corresponding to the staff working at the target post may specifically include, for example: At least one of the staff's facial image information or facial feature information. In this case, the sub-image containing the face of the staff can be used to match at least one of the face image information or face feature information in the post information to determine whether it is the job specified on the target post Personnel come to work on the job, rather than other staff to replace the post or work on behalf of the post. In addition, the position information may also include, for example, the position information corresponding to the target position. Using the face recognition results, the position corresponding to the staff member can be determined, and the position corresponding to the staff member can be matched with the position information corresponding to the target position. The result of obtaining whether the staff matches the target post can be selected according to the actual situation, and no limitation is made here.

In this way, by using the method of matching the face recognition result with the personnel identity information, it is possible to simply and clearly determine whether the staff matches the target position.

(B): The position information includes a reference action sequence corresponding to the target position.

In this case, when determining whether the staff matches the target position based on the association relationship and the position information corresponding to the target position, for example, the following method can be adopted: when the position information includes the When the reference action sequence is used, in response to the action recognition result matching the reference action sequence, it is obtained that the worker matches the target position.

In a specific implementation, the action recognition result is matched with the reference action sequence, including at least one of the following: matching of working hours, matching of production norms, and matching of production efficiency.

Next, the matching of different action recognition results with the reference action sequence will be described, including the following (B1)-(B3):

(B1): The matching of the action recognition result with the reference action sequence includes: matching of working hours.

In this case, for example, the following method can be used to determine the matching result of the working time: based on the action recognition result, determine the action time for the worker to perform the production action; based on the action time and the reference action The reference action duration corresponding to the sequence determines the matching result of the working duration.

Wherein, the action duration for the worker to perform the production action may include at least one of the following (b1)-(b4), for example:

(b1): the total time spent by the worker performing production actions;

(b2): The length of time between the workers performing adjacent production actions;

(b3): the sub-duration of the worker performing the production sub-action;

(b4): The time interval between the workers performing adjacent production sub-actions.

Next, for the convenience of describing different action durations, refer to FIG. 4 , which is a schematic diagram of a production cycle corresponding to a time axis provided by an embodiment of the present disclosure. Wherein, different production periods are marked on the time axis, for example, the illustrated production period T1 and the production period T2 (the time axis also includes unillustrated parts, such as the production period T3, etc.). In addition, the start time and end time corresponding to the production actions corresponding to the three reference actions p1-p3 are also marked on the time axis. Here, the time axis is the corresponding time axis when the workers actually perform production actions.

Exemplarily, the worker S1 can complete the production actions of three production cycles (corresponding to T1-T3 respectively) in one day, and in the first production cycle T1, the corresponding time when the production action corresponding to the reference action p1 is started is t1_1, the corresponding time when the production action corresponding to the reference action p3 is finished is t1_6; in the second production cycle T2, the corresponding time when the production action corresponding to the reference action p1 is started is t2_1, and the execution of the reference action p3 is corresponding to The corresponding time of production action is t2_6; and so on.

For the above (b1), when the action duration of the worker performing the production action includes the total duration of the worker performing the production action, for example, the matching result of the working duration can be determined according to the total duration of the reference action in the reference action duration corresponding to the reference action sequence.

Exemplarily, taking worker S1 as an example, the production step performed is raw material screening P1. In this step, for example, three reference actions p1-p3 are included, which respectively correspond to the walking action corresponding to unpacking material boxes, the standing action corresponding to counting raw materials, and the picking action corresponding to screening inferior products. Through pre-testing or judging based on experience, for example, the total duration of the reference actions corresponding to the three reference actions p1-p3 performed within a production cycle can be reasonably determined. For example, for the three reference actions p1-p3, it can be determined that the total duration of the reference actions is 1 hour; that is, it can be determined that the worker S1 can complete a production cycle within 1 hour under the condition of safety and ensuring efficiency The production sub-actions corresponding to the three reference actions p1~p3.

Specifically, by using the action recognition result, the total time for the worker to perform the production action can be determined accordingly. Taking worker S1 as an example, the total duration of production actions performed by worker S1 may include, for example, the total duration required to complete all the production actions corresponding to the three reference actions p1-p3 within one cycle. For example, it can be the time between the start time t1_1 when the production sub-action corresponding to the reference action p1 is executed in the first production cycle T1 in Figure 4, and the end time t1_6 when the production sub-action corresponding to the reference action p3 is completed. The total duration; it can also be between the start time t2_1 when the production sub-action corresponding to the reference action p1 is executed in the second production cycle T2 in Figure 4, and the end time t2_6 when the production sub-action corresponding to the reference action p3 is completed total time between.

Alternatively, the total time spent by worker S1 on performing production actions may also include, for example, the total time spent on average determined after different execution cycles have ended. production subaction. Corresponding to the time axis shown in Figure 4, in the first production cycle T1, the total duration of the worker S1 performing production actions is the total duration between t1_1 and t1_6, for example, 1 hour; in the second production cycle T2 , the total duration of worker S1 performing production actions is the total duration between t2_1 and t2_6, for example, 1.5 hours; in the third production cycle T3 (not shown in Figure 4), the total duration of worker S1 performing production actions The duration is the total duration between t3_1 and t3_6, for example, 0.5 hour. Then, an average value can be calculated for the determined 1 hour, 1.5 hour, and 0.5 hour, that is, 1 hour, and then 1 hour is used as the total time for the worker S1 to perform production actions. In this way, the interference to the determination result caused by unexpected events or other situations can also be reduced.

Specifically, for example, when it is determined that the total duration of the production actions performed by the worker S1 is 50 minutes, since the total duration of the production actions performed by the worker S1 is less than the reference action duration corresponding to the reference action sequence, it can be considered that the worker S1 is normal. The corresponding production action is completed, that is, the action recognition result of the worker S1 matches the reference action sequence.

In another possible situation, if the total duration of the production action performed by the worker S1 is 65 minutes, it may be determined that the duration of the production action performed by the worker S1 exceeds the reference duration of the action. In this case, for example, a timeout threshold can be set. For example, based on the determined reference action duration, it can be allowed to exceed 10 minutes, that is, if the timeout threshold is determined to be 70 minutes, then the worker S1 will perform the production action. When the action duration of is 65 minutes, it can also be considered that the action recognition result matches the reference action sequence. On the other hand, if the total time for worker S1 to perform the production action is 75 minutes, it can be determined that worker S1 failed to complete the production action within the specified time, so it is determined that the action recognition result does not match the reference action sequence.

For the above (b2), when the duration of the worker’s execution of the production action includes the interval of the worker’s execution of adjacent production actions, it can be based on the interval between adjacent production cycles included in the reference action duration (that is, adjacent Refer to the interval between production actions) to determine the matching result of the working hours.

When determining the interval time between the execution of adjacent production actions by the worker S1, for example, any production cycle Ti (i is an integer greater than 1) in the multiple production cycles corresponding to the worker S1 except the first production cycle T1 ) as a reference, determine the corresponding time ti_1 when the production action corresponding to the reference action p1 starts to be executed in the production cycle Ti, and the time corresponding to the end of the reference action p3 in the previous production cycle T(i-1) adjacent to the production cycle Ti The time t(i-1)_6 corresponding to the production action, by calculating the time interval between t(i-1)_6 and ti_1, can determine the length of time interval between worker S1 performing adjacent production actions. Similarly, the time interval between ti_6 and t(i+1)_1 may also be determined based on the production cycle Ti and the subsequent production cycle T(i+1) adjacent to the production cycle Ti. Of course, it is also possible to obtain the total duration of the cumulative time interval based on multiple adjacent production cycles, and calculate the average value based on the number of time intervals to obtain an interval duration.

When using the action duration of the worker to perform the production action to match the reference action duration corresponding to the reference action sequence, for example, the interval between the execution of adjacent production actions by the worker S1 included in the action duration is, for example, 10 minutes, and the reference action duration The interval between adjacent reference production actions contained in is, for example, 15 minutes. In this case, it can be determined that worker S1 has started the execution of the next cycle of production actions within the specified 15 minutes, and determined accordingly The action recognition result of worker S1 matches the reference action sequence.

In another case, similar to the above (b1), it is also possible to set a timeout threshold. For example, 15 minutes can be used as a basis to determine that the interval between adjacent production actions cannot exceed 18 minutes at most. If the interval between the execution of adjacent production actions by the worker S1 is 17 minutes, it can also be determined that the action recognition result of the worker S1 matches the reference action sequence. However, if the interval between the execution of adjacent production actions by the worker S1 exceeds 18 minutes, it is determined that the action recognition result of the worker S1 does not match the reference action sequence.

For the above (b3), when the action duration of the worker performing the production action includes the sub-duration of the worker performing the production sub-action, for example, the working duration can be determined according to the reference sub-duration corresponding to multiple reference actions included in the reference action duration matching results.

Exemplarily, taking the worker S1 as an example, the sub-actions performed by him include the production sub-actions respectively corresponding to the reference actions p1-p3 under the normal workflow. For any production sub-action, the time required to complete the production sub-action can be reasonably determined based on pre-test or empirical judgment. For example, referring to the production sub-action corresponding to the reference action p1, it can be determined that the time required to normally complete the production sub-action is 20 minutes, that is, the reference action duration of the reference action p1 is 20 minutes.

Specifically, using the result of the worker's action recognition, the time spent by the worker to perform each of the multiple production sub-actions can be determined. Taking worker S1 completing the production sub-action corresponding to reference action p1 in the first period T1 as an example, the sub-duration for worker S1 to perform the production sub-action corresponding to reference action p1 can be determined according to the period from t1_1 to t1_2, for example for 18 minutes. In this case, by comparing the action duration of 18 minutes with the reference action duration of the reference action p1 of 20 minutes, and the actual execution time of 18 minutes is less than the specified execution time of 20 minutes, it is determined that the worker S1 can complete the reference action p1 on time. production sub-actions, so the worker's action recognition results match the reference action sequence.

In another possible situation, different allowable overtime thresholds may also be determined for different production sub-actions. For details, refer to the description of (b1) above, which will not be repeated here. Wherein, the overtime allowable time threshold can be adjusted based on at least one of multiple dimensions such as production line requirements and difficulty of actions. In addition, for other production sub-actions different from the production sub-action corresponding to the reference action p1, the method of determining whether the action recognition result matches the reference action sequence is the same as the sub-duration of the production sub-action corresponding to the reference action p1 mentioned above. The method of matching the above-mentioned reference action sequence is similar, and will not be repeated here.

For the above (b4), when the duration of the worker's execution of the production action includes the interval of the worker's execution of adjacent production sub-actions, for example, it can be based on the preset time between two adjacent reference actions in the reference action sequence within a cycle. The interval length of determines the matching result of the working hours. Exemplarily, for any two adjacent production sub-actions, the interval between two adjacent production sub-actions can be reasonably determined based on pre-test or empirical judgment, for example, the production sub-action corresponding to the reference action p1 and the reference The time interval between the production sub-actions corresponding to action p2 can be set to 8 minutes.

Taking worker S1 as an example, in the first period T1, the time t1_2 when the production sub-action corresponding to the reference action p1 is completed, and the time t1_3 when the production sub-action corresponding to the reference action p2 is started can be determined, and according to t1_2 and t1_3 determine the interval between two production sub-actions, for example, 5 minutes. In this case, it can be determined that the worker S1 can execute the production sub-action corresponding to the next reference action on time, so the action recognition result of the worker S1 matches the reference action sequence.

In another possible situation, different timeout thresholds may also be determined for intervals between different production sub-actions. For details, please refer to the description of (b2) above, which will not be repeated here. In addition, for the interval between other production sub-actions, the way of determining whether the action recognition result matches the reference action sequence is the time interval between the production sub-actions corresponding to the above-mentioned reference actions p1 and p2 and The matching manner of the reference action sequence is similar, and will not be repeated here.

(B2): The action recognition result is matched with the reference action sequence, including: production normative matching.

In some examples, the action recognition result includes: a recognition action sequence; the recognition action sequence includes: multiple sub-actions. In this case, for example, the production normative matching result may be determined in the following manner: the recognition action sequence is matched with the reference action sequence to obtain the production normative matching result.

Exemplarily, taking the worker S1 as an example, when performing action recognition on the corresponding video to be processed, for example, its corresponding recognition action sequence can be obtained. Specifically, the recognition action sequence includes, for example, walking action, standing action, and picking action, which are the specific sequence of the worker S1 during actual operation, for example, first perform the walking action, then perform the standing action, and finally perform the picking action; The corresponding sub-actions may include specific actions of walking action, standing action, and picking action.

Among them, for the convenience of description, the sub-actions of specific actions including walking action, standing action, and picking action are denoted as z1, z2, and z3 respectively; Perform a walking action, then perform a standing action, and finally perform a picking action, which is expressed as "z1→z2→z3".

Specifically, when matching the recognition action sequence with the reference action sequence to obtain the production normative matching result, for example, the following method may be adopted: the recognition action sequence is divided into multiple action sub-sequences sequence; matching the reference action sequence with each action subsequence; and obtaining the production normative matching result based on the matching results of the reference action sequence and each action subsequence.

Wherein, when the recognition action sequence is divided into multiple action sub-sequences, for example, the following method may be adopted: from the recognition action sequence, determine a target sub-action that matches the target reference action in the reference action sequence; Based on the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence, the recognition action sequence is divided into at least one action sub-sequence.

Specifically, the target reference action may be determined first in the reference action sequence. When selecting the target reference action, it may be determined according to the degree of differentiation between different reference actions. For example, picking up is easier to distinguish than walking and standing, so picking up can be used as a target reference action.

In addition, the target reference action can also be a sub-action at a preset position in the reference action sequence; for example, a sub-action at the head and/or end of the reference action sequence; The required sub-actions are determined from the recognition action sequence, so that the action sub-sequences corresponding to the production actions performed by the worker multiple times can be obtained.

After the target reference action is determined, the target sub-action matching the target reference action in the reference action sequence can be relatively easily determined according to the recognition action sequence.

After the target sub-action is determined, the position of the target sub-action in the recognition action sequence can be determined, for example, which position is in the recognition action sequence. Take the recognition action sequence including z1→z2→z3→z4→z5→z6→z7→z8→z9→z10→z11→z12 as an example, where the action type of z1, z4, z7, and z10 is A; refer to the action sequence Including: A→B→C, using A as the target reference action, the recognition action sequence can be divided into: z1→z2→z3, z4→z5→z6, z7→z8→z9, and z10→z11→z12 four action subsequence.

In this way, the recognition action sequence can be divided into at least one action sub-sequence by using the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence.

In the case of determining the reference action sequence and the action subsequence, the production normative matching result can be obtained based on the matching result of the reference action sequence and each action subsequence.

In the above example, after obtaining four action subsequences, the four action subsequences of z1→z2→z3, z4→z5→z6, z7→z8→z9, and z10→z11→z12 can be combined with A→B →C matches the action type, and obtains the matching result between each action subsequence and the reference action sequence.

The production normative matching result includes at least one of the following: correct production action content, wrong production action content, missing production action, redundant production action, and wrong order of each action in the production action.

Wherein, if the reference action sequence matches each action subsequence, it can be determined that the content of the production action is correct. In the case that the reference action sequence does not match an action subsequence, if there is an action in the action subsequence that does not exist in the reference action sequence, then determine the redundant production action; if the action subsequence is missing in the reference action sequence If the matching degree between the sub-action included in the action sub-sequence and the reference action in the reference action sequence is less than the preset matching degree threshold, it is determined that the production action content is wrong; if If the order of the sub-actions included in the action sub-sequence is inconsistent with the order of the sub-actions in the reference action sequence, it is determined that the order of the actions in the production action is wrong.

In this way, different types of warning information can be sent to the staff according to different matching results of production norms, so that the staff can correct action errors in time according to different types of warning information, or stop the operation urgently to ensure their own work safety .

(B3): matching the action recognition result with the reference action sequence, including: matching of work efficiency.

In this case, for example, the matching result of work efficiency can be determined in the following manner: based on the action recognition result, determine the total duration and/or interval duration for the worker to perform the production action; The total duration and/or interval duration of the action, and the reference action duration determine the work efficiency of the staff; based on the work efficiency of the staff and the preset work efficiency, determine the matching result of the work efficiency.

Wherein, the manner of determining the total duration and the interval duration of the production actions performed by the staff can refer to the relevant descriptions of (b1)-(b4) above, which will not be repeated here.

After determining the total time for the staff to perform production actions and the interval time, the work efficiency can be determined for the staff. Among them, when determining the work efficiency, a reasonable reference action duration can be determined first, including the reference total duration and the reference interval duration, and then the work efficiency can be calculated according to the total duration of the staff performing production actions and the interval duration.

Exemplarily, in the case that the total duration of production actions performed by worker S1 is 50 minutes, and the reference total duration is 60 minutes, the work can be determined by the ratio of the reference total duration to the total duration of production actions performed by worker S1 Personnel S1 completes the production action with an efficiency of 120%, that is, the corresponding work efficiency of the staff member S1 is 120%.

In addition, a preset working efficiency may also be determined, for example, 90%. If the working efficiency of the staff is greater than or equal to the preset working efficiency, it is determined that the working efficiency matches. For example, when the above-mentioned work efficiency corresponding to the worker S1 is 120%, it may be determined that the work efficiency matches. If the work efficiency of the staff is less than the preset work efficiency, it is determined that the work efficiency does not match.

In this way, through the matching of work efficiency, it can further ensure that the staff can complete the work tasks within the specified time, and can ensure the progress of the work while ensuring the quality of the products.

In another embodiment of the present disclosure, another detection method is provided, including: determining at least one of the on-duty time and off-duty time of the staff based on the face recognition result; The on-duty time and/or the off-duty time are used to determine the attendance information of the staff member.

In this case, for example, based on the results of face recognition, it is possible to determine the on-duty time of the staff within the target period, and determine whether the staff is on duty normally according to the on-duty time.

For example, it can be determined whether the on-duty time of the staff is before the prescribed on-duty time within the target time period, and whether the off-duty time of the staff is after the prescribed off-duty time, so as to determine whether the attendance information of the staff is normal.

As another example, it is possible to determine the time when the staff member changes in the position within the target time period: for example, at 9:00 am, it is determined that the staff member arrives at the post; at 10:00 am, it is determined that the staff member leaves the post; Work position; at 5:00 pm, it is determined that the staff leave the post. Correspondingly, it can be determined that the on-duty time of the staff is 7.5 hours. If the target position requires the staff member to work for 6 hours, it can be determined that the staff member is on duty normally. In this way, it is more flexible to determine whether the attendance information is normal.

Those skilled in the art can understand that in the above method of specific implementation, the writing order of each step does not mean a strict execution order and constitutes any limitation on the implementation process. The specific execution order of each step should be based on its function and possible The inner logic is OK.

Based on the same inventive concept, the embodiment of the present disclosure also provides a person-post matching detection device corresponding to the person-post matching detection method. Since the problem-solving principle of the device in the embodiment of the present disclosure is the same as that of the above-mentioned person-post matching detection method in the embodiment of the present disclosure Similar, therefore, the implementation of the device can refer to the implementation of the method, and repeated descriptions will not be repeated.

Referring to FIG. 5 , it is a schematic diagram of a person-post matching detection device provided by an embodiment of the present disclosure. The device includes: an acquisition module 51, an identification module 52, an association module 53, and a determination module 54; wherein,

The acquisition module 51 is configured to acquire the video to be processed in the target time period obtained by shooting the staff at the target position; the recognition module 52 is configured to perform face recognition and action recognition on the video to be processed to obtain the job The face recognition result and the action recognition result of the person; the association module 53 is configured to associate the face recognition result with the action recognition result to obtain an association relationship; the determination module 54 is configured to be based on the association relationship, the The position information corresponding to the target position is determined to determine whether the staff member matches the target position.

In an optional implementation manner, the post information includes personnel identity information corresponding to the target post in the target time period; , when determining whether the staff member matches the target post, it is configured to: when the post information includes the personal identity information, in response to the face recognition result matching the personal identity information, obtain the The above-mentioned staff members are matched with the above-mentioned target positions.

In an optional implementation manner, the position information includes a reference action sequence corresponding to the target position; Whether the person matches the target position is configured to: when the position information includes the reference action sequence, in response to the action recognition result matching the reference action sequence, obtain the Target job matching.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes at least one of the following: matching of working hours, matching of production norms, and matching of production efficiency.

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching of working hours; the determining module 54 is further configured to: determine the working duration based on the action recognition result. The action duration of the personnel performing the production action; based on the action duration and the reference action duration corresponding to the reference action sequence, the matching result of the working duration is determined.

In an optional implementation manner, the action duration of the worker performing the production action includes at least one of the following: the total duration of the worker performing the production action; the interval time between the worker performing adjacent production actions ; The sub-duration of the worker executing the production sub-action; the interval between the worker's execution of adjacent production sub-actions.

In an optional implementation manner, the action recognition result includes: a recognition action sequence; the recognition action sequence includes: multiple sub-actions; the action recognition result matches the reference action sequence, including: producing normative Matching: the determination module 54 is further configured to: match the recognition action sequence with the reference action sequence to obtain the production normative matching result.

In an optional implementation manner, when the determining module 54 matches the recognition action sequence with the reference action sequence to obtain the production normative matching result, it is configured to: The sequence is divided into multiple action sub-sequences; the reference action sequence is matched with each action sub-sequence; based on the matching results of the reference action sequence and each action sub-sequence, the production normative matching result is obtained.

In an optional implementation manner, when the determining module 54 divides the recognition action sequence into a plurality of action subsequences, it is configured to: from the recognition action sequence, determine The target sub-action matched by the target reference action in ; based on the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence, cut the recognition action sequence into at least one action subsequence.

In an optional implementation manner, the production normative matching results include at least one of the following: correct content of production actions, wrong content of production actions, missing production actions, redundant production actions, and wrong order of actions in production actions .

In an optional implementation manner, the matching of the action recognition result with the reference action sequence includes: matching of work efficiency; the determination module 54 is further configured to: determine the work efficiency based on the action recognition result The total duration and/or interval duration of the staff performing production actions; based on the total duration and/or interval duration of the staff performing production actions, as well as the reference action duration, the work efficiency of the staff is determined; based on the staff's The working efficiency and the preset working efficiency determine the matching result of the working efficiency.

In an optional implementation manner, the determining module 54 is further configured to: determine at least one of the on-duty time and off-duty time of the staff member based on the face recognition result; The on-duty time and/or the off-duty time are used to determine the attendance information of the staff member.

For the description of the processing flow of each module in the device and the interaction flow between the modules, reference may be made to the relevant description in the above method embodiment, and details will not be described here.

The embodiment of the present disclosure also provides a computer device, as shown in FIG. 6 , which is a schematic structural diagram of the computer device provided by the embodiment of the present disclosure, including:

Processor 10 and memory 20; the memory 20 stores machine-readable instructions executable by the processor 10, the processor 10 is used to execute the machine-readable instructions stored in the memory 20, and the machine-readable instructions are executed by the processor 10 During execution, the processor 10 performs the following steps:

Obtain the video to be processed in the target time period obtained by shooting the staff at the target post; perform face recognition and action recognition on the video to be processed, and obtain the face recognition result and action recognition result of the staff; associate the An association relationship is obtained between the face recognition result and the action recognition result; based on the association relationship and the position information corresponding to the target position, it is determined whether the staff matches the target position.

Above-mentioned memory 20 comprises memory 210 and external memory 220; Memory 210 here is also called internal memory, is used for temporarily storing the operation data in processor 10, and the data exchanged with external memory 220 such as hard disk, processor 10 communicates with memory 210 through memory 210. The external memory 220 performs data exchange.

For the specific execution process of the above instructions, reference may be made to the steps of the method for detecting person-post matching described in the embodiments of the present disclosure, which will not be repeated here.

An embodiment of the present disclosure also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the steps of the human-position matching detection method described in the above-mentioned method embodiments are executed. . Wherein, the storage medium may be a volatile or non-volatile computer-readable storage medium.

An embodiment of the present disclosure also provides a computer program product, the computer program product carries a program code, and the instructions included in the program code can be used to execute the steps of the person-post matching detection method described in the method embodiment above, for details, see The foregoing method embodiments are not described in detail here.

Wherein, the above-mentioned computer program product may be specifically implemented by means of hardware, software or a combination thereof. In an optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK) etc. Wait.

Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the above-described system and device can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here. In the several embodiments provided in the present disclosure, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.

If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present disclosure. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disc and other media that can store program codes. .

Finally, it should be noted that: the above-mentioned embodiments are only specific implementations of the present disclosure, and are used to illustrate the technical solutions of the present disclosure, rather than limit them, and the protection scope of the present disclosure is not limited thereto, although referring to the aforementioned The embodiments have described the present disclosure in detail, and those skilled in the art should understand that any person familiar with the technical field can still modify the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present disclosure Changes can be easily imagined, or equivalent replacements can be made to some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in this disclosure. within the scope of protection. Therefore, the protection scope of the present disclosure should be defined by the protection scope of the claims.

Industrial Applicability

In the embodiment of the present application, the video to be processed in the target time period is obtained by shooting the staff at the target position; the video to be processed is identified by using the image recognition technology in the field of image processing technology, and the identification result of the staff is obtained , such as face recognition results and action recognition results; according to the characteristics of the recognition results and the position information corresponding to the target position, determine whether the staff matches the target position, so as to achieve the purpose of person-position matching detection. The solution of the present application can avoid the detection of person-post matching from relying on manual management, and improve the accuracy and detection efficiency of person-post matching detection.

Claims (16)

A person-post matching detection method, characterized in that, comprising: Obtain the video to be processed at the target time period obtained by shooting the staff at the target position; Perform face recognition and action recognition on the video to be processed, and obtain the face recognition result and action recognition result of the staff member; Associating the face recognition result with the action recognition result to obtain an association relationship; Based on the association relationship and the position information corresponding to the target position, it is determined whether the staff member matches the target position. The person-post matching detection method according to claim 1, wherein the post information includes personnel identity information corresponding to the target post during the target period; The determining whether the staff matches the target post based on the association relationship and the post information corresponding to the target post includes: When the position information includes the personal identity information, in response to the face recognition result matching the personal identity information, it is obtained that the staff member matches the target position. The person-post matching detection method according to claim 1 or 2, wherein the post information includes a reference action sequence corresponding to the target post; The determining whether the staff matches the target post based on the association relationship and the post information corresponding to the target post includes: When the position information includes the reference action sequence, in response to the action recognition result matching the reference action sequence, it is obtained that the worker matches the target position. The method according to claim 3, wherein the matching of the action recognition result and the reference action sequence includes at least one of the following: matching of working hours, matching of production norms, and matching of production efficiency. The person-post matching detection method according to claim 3 or 4, wherein the matching of the action recognition result with the reference action sequence includes: matching of working hours; The method also includes: Based on the action recognition result, determine the action duration of the worker performing the production action; Based on the action duration and the reference action duration corresponding to the reference action sequence, a matching result of the working duration is determined. The human-post matching detection method according to claim 5, wherein the duration of the worker performing the production action includes at least one of the following: The total time spent by said worker performing production actions; The length of time intervals between the execution of adjacent production actions by the staff; The sub-duration for the worker to perform the production sub-action; The time interval between the workers performing adjacent production sub-actions. The human-post matching detection method according to any one of claims 3 to 6, wherein the action recognition result includes: a recognition action sequence; the recognition action sequence includes: a plurality of sub-actions; The action recognition result is matched with the reference action sequence, including: production normative matching; The method also includes: Matching the recognition action sequence and the reference action sequence to obtain the production normative matching result. The person-post matching detection method according to claim 7, wherein the matching of the recognition action sequence and the reference action sequence to obtain the production normative matching result includes: Dividing the recognition action sequence into multiple action subsequences; matching the reference action sequence with each action subsequence; Based on the matching results of the reference action sequence and each action subsequence, the production normative matching result is obtained. The person-post matching detection method according to claim 8, wherein said dividing said recognition action sequence into a plurality of action subsequences comprises: From said sequence of identified actions, determining a target sub-action that matches a target reference action in said sequence of reference actions; Based on the position of the target sub-action in the recognition action sequence and the position of the target reference action in the reference action sequence, the recognition action sequence is divided into at least one action sub-sequence. The person-post matching detection method according to any one of claims 7 to 9, characterized in that the production-standard matching results include at least one of the following: The content of the production action is correct, the content of the production action is wrong, the production action is missing, the production action is redundant, and the order of each action in the production action is wrong. The person-post matching detection method according to any one of claims 3 to 10, wherein the matching of the action recognition result with the reference action sequence includes: matching of work efficiency; The method also includes: Based on the action recognition result, determine the total duration and/or interval duration for the worker to perform the production action; determining the work efficiency of the staff based on the total duration and/or interval duration of the production actions performed by the staff and the reference action duration; A matching result of the work efficiency is determined based on the work efficiency of the staff member and the preset work efficiency. The person-post matching detection method according to claim 2, wherein the person-post matching detection method further comprises: Based on the face recognition result, determine at least one of the on-duty time and off-duty time of the staff member; Based on the obtained on-duty time and/or off-duty time, the attendance information of the staff is determined. A person-post matching detection device, characterized in that it includes: The obtaining module is configured to obtain the video to be processed in the target time period obtained by shooting the staff at the target position; The recognition module is configured to perform face recognition and action recognition on the video to be processed, and obtain the face recognition result and action recognition result of the staff member; An association module configured to associate the face recognition result with the action recognition result to obtain an association relationship; A determining module, configured to determine whether the staff matches the target post based on the association relationship and the post information corresponding to the target post. A computer device, characterized by comprising: a processor and a memory, the memory stores machine-readable instructions executable by the processor, and the processor is configured to execute the machine-readable instructions stored in the memory, When the machine-readable instructions are executed by the processor, the processor executes the steps of the human-post matching detection method according to any one of claims 1 to 12. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is run by a computer device, the computer device executes the computer program described in any one of claims 1 to 12. The steps of the person-post matching detection method described above. A computer program product, the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and when the computer program is read and executed by a computer, the method described in any one of claims 1 to 12 is realized. The steps of the person-post matching detection method.

Images