Patrol inspection method for spinning workshop, patrol inspection device for spinning workshop, electronic device, storage medium, and program

Abstract

To provide a patrol inspection method for a spinning workshop, a patrol inspection device for a spinning workshop, an electronic device, a storage medium, and a program.SOLUTION: A patrol inspection method for a spinning workshop includes: acquiring first image information of a patrol inspection passage when a patrol inspection robot performs patrol inspection at a first position corresponding to a first spinning box in the patrol inspection passage; determining a second spinning box from a plurality of spinning boxes when the first image information includes a first inspection object; acquiring second image information of the patrol inspection passage when the patrol inspection robot performs patrol inspection at a second position corresponding to the second spinning box in the patrol inspection passage; and determining state information of the first inspection object based on the first image information and the second image information. A patrol inspection robot can inspect state information of workers in a patrol inspection passage during a patrol inspection process, so that the workers who may have safety problems can be timely found, thereby advantageously lowering the probability of safety problems occurring in worker operations and reducing safety hazards.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present disclosure relates to the field of computer technology, in particular to the field of automatic control and deep learning technology, and more particularly to a method for patrol inspection of a spinning workshop, a patrol inspection apparatus for a spinning workshop, an electronic device, a storage medium, and a program. [Background technology]

[0002] Currently, fiber raw materials need to be produced and processed in spinning workshops to form fiber yarn, but the temperature in spinning workshops is high, making it easy for workers to suffer from heat stroke or other health problems while working, posing a safety risk. Summary of the Invention [Problem to be solved by the invention]

[0003] The present disclosure provides a method for patrol inspection of a spinning workshop, a patrol inspection apparatus for a spinning workshop, an electronic device, a storage medium, and a program for reducing the probability of safety hazards occurring during worker work. [Means for solving the problem]

[0004] In a first aspect, the present disclosure provides a patrol inspection method for a spinning workshop, the spinning workshop comprises a plurality of spinning boxes arranged along a first direction, and a patrol inspection passage is provided on one side of the plurality of spinning boxes, and a patrol inspection robot is movable in the patrol inspection passage to sequentially inspect the operation states of the plurality of spinning boxes, and the patrol inspection method for the spinning workshop includes: When the patrol inspection robot patrol-inspects a first position corresponding to a first spinning box in the patrol inspection passage, first image information of the patrol inspection passage is obtained, the first spinning box being one of a plurality of spinning boxes, and the first image information is obtained by photographing the patrol inspection robot; When the first image information includes a first inspection object, a second spinning box is determined from among the plurality of spinning boxes, and an inspection order of the second spinning box is later than an inspection order of the first spinning box; When the patrol inspection robot patrol-inspects a second position corresponding to a second spinning box in the patrol inspection passage, second image information of the patrol inspection passage is acquired, the second image information being acquired by photographing the patrol inspection robot, and the second image information being an image including a position where the first inspection target is located in the first image information; determining condition information of the first inspection object based on the first image information and the second image information.

[0005] In a second aspect, the present disclosure provides a patrol inspection device for a spinning workshop, the spinning workshop comprises a plurality of spinning boxes arranged along a first direction, and a patrol inspection passage is provided on one side of the plurality of spinning boxes, and a patrol inspection robot is movable in the patrol inspection passage to sequentially inspect the operation states of the plurality of spinning boxes, and the patrol inspection device for the spinning workshop comprises: a first acquisition unit for acquiring first image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a first position corresponding to a first spinning box in the patrol inspection passage, the first spinning box being one of a plurality of spinning boxes, and the first image information being acquired by photographing the patrol inspection robot; A first determination unit for determining a second spinning box from among a plurality of spinning boxes when the first image information includes a first inspection object, and the inspection order of the second spinning box is later than the inspection order of the first spinning box; a second acquisition unit for acquiring second image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a second position corresponding to a second spinning box in the patrol inspection passage, the second image information being acquired by photographing the patrol inspection robot, and the second image information being an image including a position where the first inspection target is located in the first image information; and a second determination unit for determining state information of the first inspection object based on the first image information and the second image information.

[0006] In a third aspect, the present disclosure provides an electronic device, the device comprising: At least one processor; a memory in communication with the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the implementation of any one of the methods in the embodiments of the present disclosure.

[0007] In a fourth aspect, there is provided a non-transitory computer-readable storage medium having stored thereon computer instructions for causing a computer to perform any one of the methods in the embodiments of the present disclosure.

[0008] In a fifth aspect, there is provided a program for implementing any one of the methods in the embodiments of the present disclosure when executed by a processor.

[0009] According to the method for inspecting a spinning work area, the apparatus for inspecting a spinning work area, the electronic device, the storage medium, and the program provided by the present disclosure, when a patrol inspection robot inspects a first position corresponding to a first spinning box in a patrol inspection passage, first image information of the patrol inspection passage is obtained, where the first spinning box is one of a plurality of spinning boxes, and the first image information is obtained by photographing the patrol inspection robot; when the first image information includes a first inspection object, a second spinning box is determined from among the plurality of spinning boxes, and the inspection order of the second spinning box is later than the inspection order of the first spinning box; when a patrol inspection robot inspects a second position corresponding to the second spinning box in a patrol inspection passage, second image information of the patrol inspection passage is obtained, where the second image information is obtained by photographing the patrol inspection robot, and the second image information is an image including the position where the first inspection object is located in the first image information; and based on the first image information and the second image information, status information of the first inspection object is determined. During the patrol inspection process, the patrol inspection robot can check the status information of the workers in the patrol inspection corridor and promptly detect workers who may have safety issues, which helps to reduce the probability of safety issues occurring during the workers' work and reduce safety risks.

[0010] It should be understood that the contents described herein are not intended to describe key points or important features of the embodiments of the present disclosure, nor are they used to limit the scope of the present disclosure, other features of the present disclosure are facilitated by the following specification. [Brief description of the drawings]

[0011] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the several accompanying drawings indicate the same or similar components or elements. The accompanying drawings are not necessarily drawn to scale. It should be understood that the drawings depict only some examples provided by the present disclosure and should not be considered as limiting the scope of the present disclosure. [Figure 1]FIG. 2 is a schematic diagram illustrating an application scenario of a method for patrol inspection of a spinning workshop according to an embodiment of the present disclosure; [Diagram 2] FIG. 2 is a schematic flow chart of a method for patrol inspection of a spinning workshop according to an embodiment of the present disclosure. [Diagram 3] FIG. 13 is a schematic flow chart of a method for patrol inspection of a spinning workshop according to another embodiment of the present disclosure. [Figure 4] 1 is a schematic block diagram showing a patrol inspection device for a spinning workshop according to an embodiment of the present disclosure; [Diagram 5] FIG. 2 is a block diagram of an electronic device for implementing a method for patrol inspection of a spinning workshop according to an embodiment of the present disclosure; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] The present disclosure will now be described in more detail with reference to the accompanying drawings, in which like reference numbers represent the same or similar elements, and in which various aspects of the embodiments are illustrated, and which, unless otherwise noted, are not necessarily drawn to scale.

[0013] Furthermore, in order to better explain the present disclosure, many specific details are described in the following specific embodiments. Those skilled in the art should understand that the present disclosure can be similarly implemented without some details. In some embodiments, methods, means, components, circuits, etc. that are well known to those skilled in the art are not described in detail so that the gist of the present disclosure is clear.

[0014] In order to facilitate understanding of the spinning workshop patrol inspection method provided by the embodiments of the present disclosure, the following describes the related technologies of the embodiments of the present disclosure, and the following related technologies are provided as optional solutions and can be arbitrarily combined with the technical solutions of the embodiments of the present disclosure, all of which fall within the protection scope of the embodiments of the present disclosure.

[0015] In the related technology, the temperature in the spinning workshop is high, and workers are prone to heat stroke due to the high temperature while working in the workshop, posing a safety risk.

[0016] In order to solve at least one of the above problems, the present invention provides a method, apparatus, electronic device and storage medium for patrol inspection of a spinning workshop, which enables a patrol inspection robot to inspect the status information of workers in the patrol inspection passage during patrol inspection, thereby timely discovering workers who may have safety issues, reducing the probability of safety issues occurring during the workers' work, and reducing safety risks.

[0017] Embodiments of the present disclosure are described below with reference to the accompanying drawings.

[0018] Fig. 1 is a schematic diagram showing an application scenario of a method for patrol inspection of a spinning workshop according to an embodiment of the present disclosure. Fig. 2 is a schematic flow chart of a method for patrol inspection of a spinning workshop according to an embodiment of the present disclosure.

[0019] 1 and 2, an embodiment of the present disclosure provides a patrol inspection method 200 for a spinning workshop, in which the spinning workshop includes a plurality of spinning boxes 120 arranged along a first direction, and has a patrol inspection passage 130 on one side of the plurality of spinning boxes 120, and the patrol inspection robot 110 can move in the patrol inspection passage 130 to sequentially inspect the operation states of the plurality of spinning boxes 120. The method 200 includes the following steps S201 to S204.

[0020] In step S201, when the patrol inspection robot patrol-inspects a first position corresponding to a first spinning box in the patrol inspection passage, first image information of the patrol inspection passage is acquired, where the first spinning box is one of a plurality of spinning boxes, and the first image information is obtained by photographing the patrol inspection robot.

[0021] In step S202, when the first image information includes a first inspection object, a second spinning box is determined from among the multiple spinning boxes, and the inspection order of the second spinning box is later than the inspection order of the first spinning box.

[0022] In step S203, when the patrol inspection robot patrol-inspects a second position corresponding to the second spinning box in the patrol inspection passage, second image information of the patrol inspection passage is acquired, the second image information being obtained by photographing the patrol inspection robot, and the second image information being an image including the position where the first inspection object is located in the first image information.

[0023] In step S204, state information of the first inspection object is determined based on the first image information and the second image information.

[0024] The spinning box 120 may be a fiber yarn manufacturing device, and may have multiple spinning boxes 120 in a spinning workshop, and although only one row of spinning boxes 120 is shown in Figure 1, it can be understood that in another embodiment, multiple spinning boxes 120 may be arranged in an array in a spinning workshop. One side of the row of spinning boxes may have a patrol inspection corridor 130 where workers can walk around.

[0025] The patrol inspection robot may be a robot capable of patrol inspection in the related art, and an AGV (Automated Guided Vehicle) may be installed at the bottom of the patrol inspection robot to realize the patrol inspection robot to move freely in the spinning workshop. The patrol inspection robot may be provided with a robot arm capable of multi-degree-of-freedom movement, and further equipped with a camera to inspect the operation state such as whether the spinning box is broken. The inspection of the operation state may include filament fluctuation, filament breakage, or whether there is a foreign object on the path, the need for cleaning the spinneret, and whether there is an abnormality in the guide hook and nozzle.

[0026] The first direction is the direction of the arrow in Figure 1, and the patrol inspection robot sequentially inspects spinning boxes 120 No. 1 to No. 8 in the patrol inspection passage 130 along the first direction. Taking the first spinning box as an example, when the patrol inspection robot moves to a position corresponding to the first spinning box (for example, directly in front of the spinning box), it can inspect the operating status of the first spinning box.

[0027] The method 200 is used in a management system of a spinning workshop, such as a patrol inspection robot or a server, and the patrol inspection robot may be connected to the server in communication. The patrol inspection robot may be equipped with a camera for photographing the surrounding environment. Specifically, in step S201, when the patrol inspection robot reaches a first position of the first spinning box, the camera can photograph the first image information in the patrol inspection passage 130. Here, the first spinning box may be one of a plurality of spinning boxes, for example, the spinning box with the highest inspection order in the row of spinning boxes, such as the first spinning box in FIG. 1. The first position may be the patrol inspection position of the first spinning box, for example, the position directly in front of it, and when the patrol inspection robot 110 is located at the position, the operating state of the first spinning box can be inspected.

[0028] In step S202, the first image information includes the first inspection object 140, which can be understood as including a human body (i.e., an operator) in the first image information. For example, whether or not the first image information includes a human body is determined by a method such as image recognition, and if a human body is included, the operator is set as the first inspection object. Then, a second spinning box different from the first spinning box can be determined from the multiple spinning boxes 120. There are various methods for determining the second spinning box, and for example, the number of the second spinning box is calculated from the preset shooting time interval, the average inspection time of each spinning box, and the number of the first spinning box, and is, for example, the fifth spinning box. It can be understood that the inspection order of the second spinning box is after the inspection order of the first spinning box.

[0029] In step S203, when the patrol inspection robot reaches the second spinning box, the camera can take a second image information of the patrol inspection passage. Assuming that the first inspection object in the first image information is in the vicinity of the No. 8 spinning box, it can be understood that the second image information is also an image taken in the vicinity of the No. 8 spinning box, that is, the second image information is an image including the position where the first inspection object is located in the first image information.

[0030] It can be understood that each spinning box is provided with a corresponding spinning box number, and the number can be provided on the spinning box or in a corresponding position on the space on the top or side of the spinning box, etc. In the first image information, the first number of the spinning box closest to the first inspection object can be determined by image recognition, and in the captured second image information, the spinning box with the first number is found and an image of its vicinity is captured, which makes it easier to track the position of the first inspection object, and thus determines the status information of the first inspection object.

[0031] In step S204, the first image information and the second image information may be analyzed to determine condition information of the first inspection object, which may reflect a safe or unsafe condition of the first inspection object.

[0032] If the second image information also includes the first inspection subject, and there is no change in the behavior of the first inspection subject between the first image information and the second image information, it can be understood that it is possible that the first inspection subject is in a dangerous state where he or she may collapse from heat stroke, and an alarm may be sent to notify the administrator to check.

[0033] Alternatively, if the second image information does not include the first test subject, this indicates that the first test subject has moved away from its original position, and is therefore considered to be conscious and in good health, i.e., the first test subject is considered to be in a safe state.

[0034] In this embodiment, images are taken in the patrol inspection passage during the patrol inspection of the patrol inspection robot 110, and when the first image information includes the first inspection object, the second spinning box is determined, and when the patrol inspection robot performs a patrol inspection of the second spinning box, second image information is taken, and status information of the first inspection object can be determined based on the first image information and the second image information. This makes it possible to inspect the status information of workers in the patrol inspection passage, and to timely discover workers who have safety issues, thereby reducing the probability of safety issues occurring while the workers are working, and reducing safety risks.

[0035] In some embodiments, in step S202, when the first image information includes the first inspection object, determining a second spinning box from among the plurality of spinning boxes includes: When the first image information includes a first inspection object, determining a target inspection grade of the first inspection object based on posture information of the first inspection object in the first image information; and determining a second spinning box based on the target inspection grade.

[0036] It may be understood that there may be a correspondence between the posture information of the first inspection object and the inspection grade. For example, the posture information may be classified into a standing posture, a sitting posture, a lying posture, etc., and the inspection grade corresponding to the standing posture may be low, the inspection grade corresponding to the sitting posture may be medium, and the inspection grade corresponding to the lying posture may be high. The higher the inspection grade, the higher the probability that the first inspection object is in a dangerous state. In this case, the shooting time interval between the first image information and the second image information (or the difference in numbers between the first spinning box and the second spinning box) can be set to be small, and safety problems of the workers can be found early.

[0037] When the first image information includes a first inspection object, a corresponding inspection grade, i.e., a target inspection grade for the first inspection object, can be calculated based on the posture information of the first inspection object, and then, based on the target inspection grade, the second spinning box that was inspected when the second image information was photographed can be determined, and thus the photographing time of the second image information can be determined. This avoids a situation in which the delay in photographing the second image information delays the rescue time of a worker in a dangerous situation, and enables early detection of safety issues for the workers.

[0038] In some embodiments, determining a target inspection grade for the first inspection object based on the pose information of the first inspection object in the first image information includes: predicting the first image information using the human body pose prediction model to obtain pose information of the first test object and a reliability of the pose information; determining a test score for the first test object based on the posture information and a reliability of the posture information; determining a target inspection grade based on the inspection score and a correspondence between the inspection score and the inspection grade.

[0039] The human body posture prediction model may be a neural network in the related art that predicts the position coordinates of the main parts of the human body from image information and then predicts the posture of the human body, i.e., posture information.

[0040] The human body pose prediction model allows for determining pose information of the first examination object in the first image information and a reliability of the pose information.

[0041] In this embodiment, a basic score may be set in advance for each type of posture information. For example, posture information may include standing posture, sitting posture, and lying posture, which correspond to different basic scores such as 1, 5, and 10, respectively. The test score for the first test subject is obtained by exponentiating the basic score as a base and the reliability of the posture information as an index.

[0042] There may be a correspondence between the test score and the test grade, where a higher score indicates a higher probability that the current first test object is in a dangerous state, and a higher test grade indicates a higher need to pay attention to the first test object. By determining the measurement score, the target test grade can be determined relatively accurately.

[0043] Of course, in this embodiment, only simple classification is performed on the posture information, but in other embodiments, more detailed classification may be performed on the posture information.

[0044] In some examples, determining the second spinning box based on the target inspection grade can include determining a target inspection time interval corresponding to the target inspection grade based on the target inspection grade, and determining the second spinning box based on the target inspection time interval.

[0045] The inspection grade can be embodied as an inspection time interval, and a relationship between the inspection grade and the inspection time interval can be preset to determine a corresponding target inspection time interval according to the target inspection grade. It can be understood that the higher the inspection grade, the larger the target time interval, and the target time interval is the photographing time interval between the first image information and the second image information. This helps to accurately and easily identify the second spinning box.

[0046] In some embodiments, a method 300 for patrol inspection of a spinning workshop is provided, which includes the following steps S301 to S309.

[0047] In step S301, when the patrol inspection robot patrol-inspects a first position corresponding to a first spinning box in the patrol inspection passage, first image information of the patrol inspection passage is acquired, where the first spinning box is one of a plurality of spinning boxes, and the first image information is obtained by photographing the patrol inspection robot.

[0048] In step S302, the first image information is divided using the image division model to obtain a first division result.

[0049] In step S303, if an inspection object of the target category is included in the first division result, the inspection object of the target category in the first image information is determined as the first inspection object.

[0050] In step S304, when the first image information includes the first inspection object, a second spinning box is determined from among the multiple spinning boxes, and the inspection order of the second spinning box is later than the inspection order of the first spinning box.

[0051] In step S305, when the patrol inspection robot patrol-inspects a second position corresponding to the second spinning box in the patrol inspection passage, second image information of the patrol inspection passage is acquired, the second image information being obtained by photographing the patrol inspection robot, and the second image information being an image including the position where the first inspection object is located in the first image information.

[0052] In step S306, the second image information is segmented using the image segmentation model to obtain a second segmentation result.

[0053] In step S307, if an inspection object of the target category is included in the second division result, the inspection object of the target category in the second image information is determined as the second inspection object.

[0054] In step S308, the first test object and the second test object are compared to obtain a comparison result to indicate whether the second test object and the first test object are the same object.

[0055] In step S309, state information of the first inspection target is determined based on the comparison result.

[0056] Here, step S301, step S304 and step S305 correspond to the above-mentioned steps S201 to S203, respectively, and for specific details, reference can be made to the above-mentioned embodiments, and the description will not be repeated here.

[0057] Steps S307 to S309 are a further detailed version of the above-mentioned step S204, that is, step S204 can specifically include steps S307 to S309.

[0058] In step S302, the first image information can be segmented by an image segmentation model. The image segmentation model can segment the image based on region labels, for example, a Segment Anything model can be used to segment the image into small parts, and each region in the image and each region's corresponding category label can be obtained.

[0059] By dividing the first image information by the image division model, a plurality of region images in the first image information and categories corresponding to each region image can be obtained, that is, a first division result can be obtained.

[0060] In step S303, an inspection target of a target category can be obtained from the first segmentation result. The target category may be a human body, and if an area image with a category label of a human body exists in the first segmentation result, the operator (i.e., the inspection target) corresponding to this area image can be set as the first inspection target.

[0061] The division method in step S306 is similar to that in step S302, specifically, refer to step S302, and the second image information is divided into images to obtain a second division result.

[0062] The determination method of step S307 is similar to that of step S303, and specifically, step S303 can be referred to. If an area image whose category label is a human body exists in the second segmentation result, the worker corresponding to this area image can be the second inspection target.

[0063] In step S308, the first inspection object and the second inspection object can be compared to obtain a comparison result, which is used to determine whether they are the same object, i.e., the same operator.

[0064] In step S309, status information of the first test object may be determined based on the comparison result of step S308.

[0065] In this embodiment, the first image information and the second image information can be segmented using an image segmentation model, which allows for an accurate determination of whether or not a worker is included in the image, and is useful for subsequent comparison of the first inspection object with the second inspection object.

[0066] In some embodiments, comparing the first test object to the second test object to obtain a comparison result in step S308 includes: Obtaining first identification information of a first test object; Obtaining second identification information of a second test subject; Comparing the first identification information and the second identification information to obtain a comparison result.

[0067] The first identification information may be a human face or an employee number on a worker's clothing, and by comparing the first identification information, it can be determined whether the first inspection object and the second inspection object are the same object.

[0068] In some embodiments, the first identification information includes a first human face image of the first test object, and the second identification information includes a second human face image of the second test object. Obtaining the first identification information of the first test object in the above method includes determining the first human face image as the first identification information when the first segmentation result includes the first human face image whose category is human face.

[0069] Obtaining second identification information of the second inspection object in the above method includes determining the second human face image as the second identification information when the second segmentation result includes a second human face image whose category is a human face.

[0070] Image segmentation models such as the segment anything model can accurately identify regions where a human body exists, as well as accurately segment regions in the human body, for example, extracting human face regions.

[0071] Of course, the first division result may include the result of division multiple times using the image division model. For example, the first image information is first divided to extract an area image where the first inspection object exists, and then the area image is secondarily divided using the image division model again to obtain a first human face image whose category in the human body is a human face, and the first human face image is used as the first identification information.

[0072] Similarly, the method of extracting the second identification information is similar to the method of extracting the first identification information, and will not be described repeatedly here.

[0073] By using a human face image as the identification information, it is possible to accurately determine whether the first inspection object and the second inspection object are the same person.

[0074] In another embodiment, obtaining the first identification information of the first test subject includes: When the first division result does not include a first human face image whose category is a human face, determining a first photographing time of the first image information and target position information of the first inspection object, the target position information being number information of a third spinning box that is closest to the first inspection object, the target position information being obtained based on the first division result, and the third spinning box being one of the multiple spinning boxes; determining a first imaging device capable of imaging the position according to the target position information in the spinning workshop based on the target position information; acquiring a first auxiliary image captured at a first capturing time by a first imaging device; The method further includes determining a first human face image based on the first auxiliary image.

[0075] In this embodiment, when the first division result does not include a human face, i.e., when the first human face image of the first inspection object cannot be extracted from the first image information, the number information of the spinning box closest to the first inspection object can be obtained based on the first division result. For example, as shown in Fig. 1, the worker is located near the No. 8 spinning box, and the No. 8 spinning box is photographed in the first image information, and the area of ​​the spinning box and the number information of the area can be obtained through image division, i.e., the number information of the spinning box closest to the first inspection object can be determined.

[0076] Then, the number information is used as the position information of the first inspection target, and a first image capturing device capable of capturing an image of the position in the spinning workshop can be called. The first image capturing device may be an image capturing device installed in a position such as the ceiling of the workshop.

[0077] Based on the image captured by the first photographing device, an image frame (first auxiliary image) corresponding to the first photographing time when the first image information was photographed can be found, and a first human face image can be extracted based on that image frame.

[0078] The first photographing device can capture human face images of the first examination object from different angles, which helps to accurately extract the first human face image of the first examination object.

[0079] If the second division result of the second image information does not include a second human face image, the above-mentioned method can also be used to find a second human face image captured from a different angle using an imaging device in the spinning workshop.

[0080] 3 is a flow chart schematic diagram of a method for inspecting a spinning workshop according to another embodiment of the present disclosure. Referring to FIG. 3, in some embodiments, in step S308, comparing the first inspection object with the second inspection object to obtain a comparison result includes: If the first human face image of the first test subject or the second human face image of the second test subject cannot be obtained, the method further includes obtaining a first character on clothing of the first test subject and a second character on clothing of the second test subject; Obtaining the first identification information of the first inspection object includes determining a first character as the first identification information; Obtaining the second identification information of the second test object includes determining the second character as the second identification information.

[0081] It is understood that the name and employee number of each worker may be provided on the work clothes of the worker. When the first human face image and the second human face image cannot be acquired by the above-mentioned method, the identity of the worker can be determined by acquiring characters (e.g., name and employee number) on the worker's clothes as identification information.

[0082] In some embodiments, acquiring a first character on clothing of a first test subject includes: determining a first region in the first image information based on the first segmentation result, the first region being a region in the first image information where the first inspection object is located; If the first segmentation result indicates that the first region includes a first partial image whose category is a character, determining a first character based on the first partial image.

[0083] In this embodiment, the first segmentation result may include a first region in which the first inspection object is located, and the first region may be accurately extracted according to the target category.

[0084] Then, a secondary image division is performed on the first region, and if the first region includes a category whose category label is a character, a first local image corresponding to the character is extracted, and the corresponding first character may be determined by character recognition or the like.

[0085] The first character allows for a precise determination of the identity of the first test subject.

[0086] In some embodiments, acquiring a first character on clothing of a first test subject includes: determining a first photographing time of the first image information and target position information of the first inspection object when the first division result indicates that the first region does not include a first local image whose category is character, the target position information being number information of a third spinning box closest to the first inspection object, the target position information being obtained based on the first division result, and the third spinning box being one of the multiple spinning boxes; determining a first imaging device capable of imaging the position according to the target position information in the spinning workshop based on the target position information; acquiring a first auxiliary image captured at a first capturing time by a first imaging device; The method further includes determining a first character based on the first auxiliary image.

[0087] In this embodiment, if the first character cannot be determined by the first image information, the number information of the spinning box closest to the first inspection object can be obtained based on the first division result. For example, as shown in Fig. 1, an operator is located near the No. 8 spinning box, and can photograph the number of the No. 8 spinning box in the first image information, and the area of ​​this spinning box and the number information in the area can be obtained through image division, that is, the number information of the spinning box closest to the first inspection object can be determined.

[0088] Then, the number information is used as position information of the first inspection target, and a first imaging device in the spinning workshop capable of photographing the position can be called. The first imaging device may be an imaging device installed at a position such as the ceiling of the workshop.

[0089] Based on the image captured by the first photographing device, an image frame (first auxiliary image) corresponding to the first photographing time when the first image information was captured can be found, and the first character can be extracted based on the image frame. Specifically, the first character is recognized by using image segmentation and image recognition techniques.

[0090] The first photographing device can capture a first character corresponding to the first test subject from different angles, which is helpful in determining the identity information of the first test subject.

[0091] Similarly, a second character corresponding to the second test target may be determined in a similar manner to each of the above embodiments.

[0092] According to the above embodiment, step S309 determines the state information of the first inspection object based on the comparison result. determining a first region in the first image information where the first inspection object is located and a second region in the second image information where the second inspection object is located when the comparison result indicates that the second inspection object and the first inspection object are the same object; determining a similarity between the first region and the second region; and determining condition information of the first test object based on the similarity.

[0093] The first region and the second region may both be local regions extracted from the segmentation result and categorized as a human body.

[0094] It can be understood that when the first inspection object and the second inspection object are the same object, by calculating the similarity between the first area and the second area, it is possible to determine whether the worker's position or posture has changed, and further, to obtain status information of the first inspection object.

[0095] In some embodiments, determining the state information of the first test object based on the similarity comprises: When the similarity exceeds a similarity threshold, determining that the state information of the first inspection object is a dangerous state and issuing first alarm information; determining that the state information of the first test object is a safe state if the similarity does not exceed the similarity threshold.

[0096] If the similarity exceeds the similarity threshold, it indicates that there is a high probability that the position or posture of the first inspection object has not changed, and it is determined that the first inspection object may be in a dangerous state. In this case, a first alarm information is sent to the manager to notify him or her to check on the worker in question, so that the dangerous situation of the worker can be discovered in a timely manner.

[0097] Conversely, if the similarity is equal to or less than the threshold, this indicates that there is a high probability that the position or orientation of the first inspection object has changed, and it is determined that the first inspection object is in a safe state.

[0098] In some embodiments, determining that the condition information of the first inspection object is a hazardous condition if the similarity exceeds a similarity threshold includes: determining a first object information of a first inspection object when the similarity exceeds a similarity threshold; determining a wearable device worn by the first test subject based on the first subject information; acquiring current health data representative of a current vital sign indicator of a first test subject from a wearable device; and determining that the condition information of the first test subject is a risk condition when the current health data is outside a preset health range.

[0099] It can be understood that when workers work in the spinning workshop, they can wear wearable devices such as smart bracelets, which can detect the workers' health data such as blood pressure, body temperature, heart rate, etc.

[0100] In this embodiment, when the similarity exceeds the similarity threshold, the first target information may be determined as an identity identifier of the first test target, such as an employee number or a name. The method of acquiring the first target information may refer to the method of acquiring the first identification information, and when the first target information is a name, the first human face image in the first identification information is first acquired, and then the first target information can be obtained based on the correspondence data between the human face image and the name input by the employee.

[0101] When the first identification information is a first character and the first character is an employee number, the acquired first identification information can be the first target information.

[0102] Then, a wearable device corresponding to the first test subject is determined based on the first test subject information, and current health data is obtained from the wearable device. If any of the health data, such as blood pressure, body temperature, or heart rate, is outside a preset health range, the first test subject is deemed to be in danger. If all the health data is within the preset health range, the first test subject is deemed to be in a safe state.

[0103] In this embodiment, further determining the current vital sign status of the first test subject by the wearable device aids in the accuracy of the determination of the status information.

[0104] In some embodiments, determining the similarity between the first region and the second region comprises: obtaining a first set of position information for each keypoint of the first test object in the first image information using the human body pose prediction model; obtaining a second set of position information for each keypoint of the second test object in the second image information using the human pose prediction model; determining a similarity based on the first set of location information and the second set of location information.

[0105] In this embodiment, the similarity between the first region and the second region can be represented by the similarity between the orientation information of the first inspection object and the orientation information of the second inspection object in the first image information.

[0106] Specifically, a first position information set of each key point of the first test object can be obtained by the human body posture testing model, where each key point can be each key point of the human body, for example, can include key point parts such as nose, left eye, right eye, left ear, right ear, left shoulder, right shoulder, left elbow, right elbow, left wrist, right wrist, left thigh, right thigh, left knee, right knee, left ankle, right ankle, etc.

[0107] Similarly, the position of each keypoint of the second inspection object in the second image information can be predicted using the human body pose prediction model to obtain a second set of position information.

[0108] Then, a comparison can be made based on the position information set, for example, by using the position information set to reconstruct the pose of the human body and then comparing the similarity between the two poses to obtain a similarity.

[0109] According to the key points of the human body, it can be accurately determined whether the posture of the first inspection subject has changed.

[0110] In some embodiments, determining the similarity between the first region and the second region includes adjusting a size of the first region such that a size of a minimum bounding rectangle of the adjusted first region matches a size of a minimum bounding rectangle of the second region, and determining a similarity between a contour of the second region and a contour of the adjusted first region.

[0111] In this embodiment, a first region of the human body in the first image information and a second region of the human body in the second image information are cut out, and the sizes of the first and second regions are adjusted by scaling processing so that the sizes of the minimum bounding rectangular frames of both regions match, and then the similarity between the contours is compared to determine the similarity between the first and second regions.

[0112] Depending on subtle changes in the shooting angle, the size of the human body in the first image information and the human body in the second image information may be enlarged or reduced, and size adjustment can make the two relatively consistent in size, which helps to improve the accuracy of the status information results.

[0113] In addition, depending on the shooting angle, even if the position of the worker does not change, the human body in the two images may not be 100% similar. In this case, the degree of change in the camera angle is determined according to the range of the shooting time interval, and the similarity threshold is also determined. For example, the similarity threshold is set lower as the degree of change increases, for example to about 70%, and a more reasonable judgment result of the status information can be obtained.

[0114] In some embodiments, in step S309, determining status information of the first test object based on the comparison result further includes determining that the status information of the first test object is in a safe state if the comparison result indicates that the second test object and the first test object are not the same object.

[0115] If the first inspection object and the second inspection object are not the same object, it can be determined that the first inspection object has left its original position during the patrol inspection by the patrol inspection robot, which indicates that the first inspection object is in an active state, and it can be determined that the first inspection object is in a safe state and is not in a state where it will collapse from heat stroke.

[0116] In addition, when it is determined that the condition information or the first image information of the first inspection object does not include a human body, the above-mentioned patrol inspection method is re-executed based on a preset restart time, for example, 10 minutes, thereby making it possible to continuously inspect the patrol inspection passage during the patrol inspection and to timely discover workers in a dangerous condition.

[0117] In some embodiments, in step S204, determining status information of the first inspection object based on the first image information and the second image information further includes determining that the status information of the first inspection object is a safe state if the second segmentation result does not include an inspection object of the target category.

[0118] If the second image information does not include a human body, it may indicate that the first inspection object has left its original position during the patrol inspection by the patrol inspection robot, and it may be determined that the first inspection object is in a safe state.

[0119] 4 is a schematic block diagram of a patrol inspection device for a spinning workshop according to an embodiment of the present disclosure. Referring to FIG. 4, an embodiment of the present invention provides a patrol inspection device 400 for a spinning workshop, the spinning workshop includes a plurality of spinning boxes arranged along a first direction, and a patrol inspection passage is provided on one side of the plurality of spinning boxes, and a patrol inspection robot is movable in the patrol inspection passage to sequentially inspect the operation states of the plurality of spinning boxes, and the patrol inspection device 400 for the spinning workshop includes: A first acquisition unit 401 for acquiring first image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a first position corresponding to a first spinning box in the patrol inspection passage, the first spinning box being one of a plurality of spinning boxes, and the first image information being acquired by photographing the patrol inspection robot; A first determination unit 402 for determining a second spinning box from among a plurality of spinning boxes when the first image information includes a first inspection object, and the inspection order of the second spinning box is later than the inspection order of the first spinning box; a second acquisition unit 403 for acquiring second image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a second position corresponding to a second spinning box in the patrol inspection passage, the second image information being acquired by photographing the patrol inspection robot, and the second image information being an image including a position where the first inspection target is located in the first image information; A second determining unit 404 for determining condition information of the first inspection object based on the first image information and the second image information.

[0120] In some embodiments, when the first image information includes a first inspection object, the first determination unit 402 is used to determine a target inspection grade of the first inspection object based on posture information of the first inspection object in the first image information, and to determine a second spinning box based on the target inspection grade.

[0121] In some embodiments, the first determination unit 402 is used for predicting first image information using a human body posture prediction model to obtain posture information and a reliability of the posture information of the first examination object, determining an examination score of the first examination object based on the posture information and the reliability of the posture information, and determining a target examination grade based on the examination score and the correspondence between the examination score and the examination grade.

[0122] In some embodiments, the first determining unit 402 is used for determining a target inspection time interval corresponding to the target inspection grade based on the target inspection grade, and determining a second spinning box based on the target inspection time interval.

[0123] In some embodiments, the device 400 includes: a first segmentation unit for segmenting the first image information by using the image segmentation model to obtain a first segmentation result; a third determination unit for determining, when the first division result includes an inspection object of the target category, the inspection object of the target category in the first image information as a first inspection object;

[0124] A second division unit for dividing the second image information by using the image division model to obtain a second division result; The second determination unit 404 determines the inspection object of the target category in the second image information as the second inspection object when the second division result includes the inspection object of the target category; comparing the first test object with the second test object to obtain a comparison result indicating whether the second test object and the first test object are the same object; and determining status information of the first test object based on a result of the comparison.

[0125] In some embodiments, the second determination unit 404 is further used for obtaining first identification information of the first test subject, obtaining second identification information of the second test subject, and comparing the first identification information and the second identification information to obtain a comparison result.

[0126] In some embodiments, the first identification information includes a first human face image of a first inspection object, and the second identification information includes a second human face image of a second inspection object, and the first determination unit 402 is further used for determining the first human face image as the first identification information when the first segmentation result includes a first human face image whose category is a human face, and the first determination unit 402 is further used for determining the second human face image as the second identification information when the second segmentation result includes a second human face image whose category is a human face.

[0127] In some embodiments, the first determination unit 402 is further used for determining a first shooting time of the first image information and target position information of the first inspection object when the first division result does not include a first human face image whose category is a human face, where the target position information is number information of a third spinning box closest to the first inspection object, the target position information is obtained based on the first division result, and the third spinning box is one spinning box among the multiple spinning boxes; determining a first imaging device capable of photographing a position according to the target position information in the spinning workshop based on the target position information; obtaining a first auxiliary image photographed by the first imaging device at the first shooting time; and determining the first human face image based on the first auxiliary image.

[0128] In some embodiments, when a first human face image of the first test subject or a second human face image of the second test subject cannot be obtained, the second determination unit 404 is further used for obtaining a first character on the clothing of the first test subject and a second character on the clothing of the second test subject, determining the first character as first identification information, and determining the second character as second identification information.

[0129] In some embodiments, the second determination unit 404 is used for determining a first region in the first image information based on the first segmentation result, where the first region is an area in the first image information where the first inspection object is located, and determining a first character based on the first local image when the first segmentation result indicates that the first region includes a first local image whose category is character.

[0130] In some embodiments, when the first division result indicates that the first region does not include a first local image whose category is character, the second determination unit 404 is further used for determining a first shooting time of the first image information and target position information of the first inspection object, where the target position information is number information of a third spinning box closest to the first inspection object, the target position information is obtained based on the first division result, and the third spinning box is one spinning box among the multiple spinning boxes; determining a first imaging device capable of photographing a position according to the target position information in the spinning workshop based on the target position information; obtaining a first auxiliary image photographed by the first imaging device at the first shooting time; and determining a first character based on the first auxiliary image.

[0131] In some embodiments, when the comparison result indicates that the second test object and the first test object are the same object, the second determination unit 404 is further used for determining a first area in the first image information where the first test object is located and a second area in the second image information where the second test object is located, determining a similarity between the first area and the second area, and determining status information of the first test object based on the similarity.

[0132] In some embodiments, the second determination unit 404 is further used for determining that the status information of the first inspection object is in a dangerous state and issuing first alarm information when the similarity exceeds the similarity threshold, and for determining that the status information of the first inspection object is in a safe state when the similarity does not exceed the similarity threshold.

[0133] In some embodiments, the second determination unit 404 is further used for determining first object information of the first test subject when the similarity exceeds a similarity threshold, determining a wearable device worn by the first test subject based on the first object information, obtaining current health data of the wearable device to represent current vital sign indicators of the first test subject, and determining that the status information of the first test subject is in a dangerous state when the current health data exceeds a preset health range.

[0134] In some embodiments, the second determination unit 404 is further used for: using a body pose prediction model to obtain a first set of position information for each key point of the first inspection object in the first image information; using the body pose prediction model to obtain a second set of position information for each key point of the second inspection object in the second image information; and determining a similarity based on the first set of position information and the second set of position information.

[0135] In some embodiments, the second determination unit 404 is further used for adjusting the size of the first region such that the size of the minimum bounding rectangle of the adjusted first region matches the size of the minimum bounding rectangle of the second region, and determining a similarity between the contour of the second region and the contour of the adjusted first region.

[0136] In some embodiments, the second determination unit 404 is further used for determining that the status information of the first test object is in a safe state when the comparison result indicates that the second test object and the first test object are not the same object.

[0137] In some embodiments, the second determining unit 404 is further used for determining that the status information of the first inspection object is in a safe state when the second segmentation result does not include any inspection object of the target category.

[0138] The specific functions and exemplary descriptions of each module and sub-module of the apparatus in the embodiments of the present disclosure may refer to the relevant descriptions of the corresponding steps in the embodiments of the above method, and will not be repeated here.

[0139] According to an embodiment of the present disclosure, an electronic device is provided, At least one processor; a memory in communication with the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the implementation of any one of the methods in the embodiments of the present disclosure. In an embodiment of the present disclosure, a non-transitory computer-readable storage medium is provided having stored thereon computer instructions for causing a computer to perform any one of the methods in the embodiment of the present disclosure. In an embodiment of the present disclosure, a program is provided, which, when executed by a processor, realizes any one of the methods in the embodiments of the present disclosure. FIG. 5 is a structural block diagram of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 5, the electronic device includes a memory 510 and a processor 520, and the memory 510 stores a computer program executable by the processor 520. The number of the memory 510 and the processor 520 may be one or more. The memory 510 may store one or more computer programs, which, when executed by the electronic device, cause the electronic device to perform the method provided by the above method embodiment. The electronic device may further include: The communication interface 530 is used for communicating with an external device and performing data interaction and transmission.

[0140] When the memory 510, the processor 520, and the communication interface 530 are implemented independently, the memory 510, the processor 520, and the communication interface 530 are connected to each other via a bus and can communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus can be classified into an address bus, a data bus, a control bus, and the like. For ease of explanation, only one thick line is shown in FIG. 5, but this does not indicate only one bus or only one type of bus.

[0141] Optionally, in a specific implementation, when the memory 510, the processor 520, and the communication interface 530 are integrated on one chip, the memory 510, the processor 520, and the communication interface 530 can communicate with each other via an internal interface.

[0142] It should be understood that the processor may be a Central Processing Unit (CPU) or may be other general purpose processors, Digital Signal Processing (DSP), Application Specific Integrated Circuits (ASIC), Field Programmable Gate Arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware assemblies, etc. The general purpose processor may be a microprocessor or any conventional processor, etc. Additionally, the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0143] Additionally, optionally, the memory may include read-only memory and random access memory, and may further include non-volatile random access memory. The memory may be either volatile or non-volatile memory, or may include both volatile and non-volatile memory. Here, non-volatile memory may include Read-Only Memory (ROM), Programmable ROM (PROM), Erasable PROM (EPROM), Electrically EPROM (EEPROM), or flash memory. Volatile memory may include Random Access Memory (RAM), which acts as an external cache. By way of example and not limitation, many forms of RAM are available. For example, static random access memory (Static RAM, SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (Double Data Rate SDRAM, DDR SDRAM), enhanced SDRAM (Enhanced SDRAM, ESDRAM), synchlink DRAM (Synchlink DRAM, SLDRAM), and direct RAMBUS RAM (DR RAM).

[0144] The above-mentioned embodiments may be implemented in whole or in part in software, hardware, firmware, or any combination thereof. When implemented in software, the whole or in part may be implemented in the form of a computer program product. A computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, a process or function according to the embodiments of the present disclosure is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, Bluetooth, microwave, etc.). The computer-readable storage medium may be any available medium that can be accessed by a computer, or a data storage device including a server, a data center, etc. integrated with one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a Digital Versatile Disc (DVD)), or a semiconductor medium (e.g., a Solid State Disk (SSD)). Note that the computer-readable storage medium referred to in this disclosure may be a non-volatile storage medium, in other words, a non-transitory storage medium.

[0145] Those skilled in the art can understand that all or part of the steps for realizing the above embodiments may be implemented by hardware, or may be implemented by instructing relevant hardware by a program, and the program may be stored in a computer-readable storage medium, and the storage medium may be a read-only memory, a magnetic disk, an optical disk, etc.

[0146] In the description of the embodiments of the present disclosure, the reference words "one embodiment", "several embodiments", "examples", "specific examples", or "several examples" mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. And, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. Furthermore, a person skilled in the art may combine different embodiments or examples and features of different embodiments or examples described in the present disclosure to the extent that they are not mutually inconsistent.

[0147] In the description of the embodiments of the present disclosure, unless otherwise specified, " / " means "or", for example, A / B can mean either A or B. "And / or" in the present disclosure is merely for describing the related relationship of related objects, and indicates that three kinds of relationships may exist, for example, A and / or B can indicate the following three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0148] In the description of the embodiments of the present disclosure, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying a relative importance, nor should they be construed as implying a number of technical features shown. Thus, features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present disclosure, "plurality" means two or more, unless otherwise specified.

[0149] The above are merely illustrative examples of the present disclosure, and are not intended to limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure should be included within the scope of protection of the present disclosure.

Claims

1. A method for inspecting a spinning workshop, comprising the steps of: The spinning workshop includes a plurality of spinning boxes arranged along a first direction, and has a patrol inspection passage on one side of the plurality of spinning boxes, and a patrol inspection robot is movable along the patrol inspection passage to sequentially inspect the operation states of the plurality of spinning boxes; The method for inspecting the spinning workshop comprises: When the patrol inspection robot patrols and inspects a first position corresponding to a first spinning box in the patrol inspection passage, first image information of the patrol inspection passage is obtained, the first spinning box is one of the plurality of spinning boxes, and the first image information is obtained by photographing the patrol inspection robot; When the first image information includes a first inspection object, a second spinning box is determined from among the plurality of spinning boxes, and an inspection order of the second spinning box is later than an inspection order of the first spinning box; When the patrol inspection robot patrol-inspects a second position in the patrol inspection passage corresponding to the second spinning box, second image information of the patrol inspection passage is acquired, the second image information being obtained by photographing the patrol inspection robot, and the second image information being an image including a position where the first inspection target is located in the first image information; determining state information of the first inspection object based on the first image information and the second image information; Patrol inspection method for spinning workshop.

2. When the first image information includes the first inspection object, determining a second spinning box from among the plurality of spinning boxes includes: When the first image information includes the first inspection object, determining a target inspection grade of the first inspection object based on posture information of the first inspection object in the first image information; determining the second spinning box based on the target inspection grade; The method for inspecting a spinning workshop according to claim 1.

3. Determining a target inspection grade of the first inspection object based on posture information of the first inspection object in the first image information, predicting the first image information using a human body pose prediction model to obtain the pose information of the first test object and a confidence level of the pose information; determining a test score for the first test object based on the posture information and a reliability of the posture information; determining the target inspection grade based on the inspection score and a correspondence between the inspection score and the inspection grade; The method for inspecting a spinning workshop according to claim 2.

4. determining the second spinning box based on the target inspection grade, determining a target inspection time interval corresponding to the target inspection class based on the target inspection class; determining the second spinning box based on the target inspection time interval; The method for inspecting a spinning workshop according to claim 2.

5. The method for inspecting the spinning workshop comprises: Segmenting the first image information using an image segmentation model to obtain a first segmentation result; When the first division result includes an inspection object of a target category, determining the inspection object of the target category in the first image information as a first inspection object; and segmenting the second image information using the image segmentation model to obtain a second segmentation result. Determining state information of the first inspection object based on the first image information and the second image information includes: When the second division result includes an inspection object of the target category, determining the inspection object of the target category in the second image information as a second inspection object; comparing the first test object with the second test object to obtain a comparison result indicating whether the second test object and the first test object are the same object; and determining status information of the first test object based on a comparison result. The method for inspecting a spinning workshop according to claim 1.

6. Comparing the first test object with the second test object to obtain a comparison result includes: Obtaining first identification information of the first test object; Obtaining second identification information of the second test object; comparing the first identification information with the second identification information to obtain the comparison result. The method for inspecting a spinning workshop according to claim 5.

7. the first identification information includes a first human face image of the first test subject; the second identification information includes a second human face image of the second test subject; Obtaining first identification information of the first test object includes: determining, when the first segmentation result includes a first human face image whose category is a human face, the first human face image as the first identification information; Obtaining second identification information of the second test object includes: determining, when the second segmentation result includes a second human face image whose category is a human face, the second human face image as the second identification information. The method for inspecting a spinning workshop according to claim 6.

8. Obtaining first identification information of the first test object includes: When the first division result does not include a first human face image whose category is a human face, determining a first photographing time of the first image information and target position information of a first inspection object, the target position information being number information of a third spinning box that is closest to the first inspection object, the target position information being obtained based on the first division result, and the third spinning box being one of the plurality of spinning boxes; determining a first imaging device capable of photographing a position in the spinning workshop according to the target position information based on the target position information; acquiring a first auxiliary image captured at the first capturing time by the first imaging device; determining the first human face image based on the first auxiliary image. The method for inspecting a spinning workshop according to claim 7.

9. Comparing the first test object with the second test object to obtain a comparison result includes: acquiring a first character on clothing of the first test subject and a second character on clothing of the second test subject when a first human face image of the first test subject or a second human face image of the second test subject cannot be acquired; obtaining first identification information of the first inspection object includes determining the first character as the first identification information; obtaining second identification information of the second test object includes determining the second character as the second identification information; The method for inspecting a spinning workshop according to claim 6.

10. Acquiring a first character on clothing of the first test subject includes: determining a first region in the first image information based on the first division result, the first region being a region in which the first inspection object is located in the first image information; and and when the first segmentation result indicates that the first region includes a first local image whose category is a character, determining the first character based on the first local image. The method for inspecting a spinning workshop according to claim 9.

11. Acquiring a first character on clothing of the first test subject includes: determining a first photographing time of the first image information and target position information of the first inspection object when the first division result indicates that the first region does not include a first local image whose category is character, the target position information being number information of a third spinning box closest to the first inspection object, the target position information being obtained based on the first division result, and the third spinning box being one of the plurality of spinning boxes; determining a first imaging device capable of photographing a position in the spinning workshop according to the target position information based on the target position information; acquiring a first auxiliary image captured at the first capturing time by the first imaging device; determining the first character based on the first auxiliary image. The method for inspecting a spinning workshop according to claim 10.

12. Determining status information of the first test object based on a comparison result includes: determining a first region in which the first inspection object is located in the first image information and a second region in which the second inspection object is located in the second image information when the comparison result indicates that the second inspection object and the first inspection object are the same object; determining a similarity between the first region and the second region; determining state information of the first test object based on the similarity. The method for inspecting a spinning workshop according to claim 5.

13. Determining state information of the first inspection object based on the similarity When the similarity exceeds a similarity threshold, determining that the state information of the first inspection object is a dangerous state and issuing first alarm information; and determining that the state information of the first test object is a safe state if the similarity does not exceed the similarity threshold. The method for inspecting a spinning workshop according to claim 12.

14. Determining that the state information of the first inspection object is a dangerous state when the similarity exceeds a similarity threshold, determining first object information of the first test object if the similarity exceeds a similarity threshold; Determining a wearable device worn by the first test subject based on the first subject information; acquiring current health data of the wearable device representative of a current vital sign indicator of the first test subject; and determining that the condition information of the first test subject is a dangerous condition when the current health data is outside a preset health range. The method for inspecting a spinning workshop according to claim 13.

15. Determining a similarity between the first region and the second region includes: obtaining a first set of position information for each key point of the first test object in the first image information using a body pose prediction model; obtaining a second set of position information for each key point of the second test object in the second image information using the human pose prediction model; determining the similarity based on the first set of location information and the second set of location information. The method for inspecting a spinning workshop according to claim 12.

16. Determining a similarity between the first region and the second region includes: adjusting the size of the first region such that the size of a minimum bounding rectangle of the first region after the adjustment matches the size of a minimum bounding rectangle of the second region; determining a similarity between a contour of the second region and a contour of the first region after adjustment; The method for inspecting a spinning workshop according to claim 13.

17. Determining status information of the first test object based on a comparison result includes: and determining, when the comparison result indicates that the second test object and the first test object are not the same object, that the status information of the first test object is a safe status. The method for inspecting a spinning workshop according to claim 5.

18. Determining state information of the first inspection object based on the first image information and the second image information includes: and determining, when the second segmentation result does not include an inspection object of the target category, that the state information of the first inspection object is a safe state. The method for inspecting a spinning workshop according to claim 5.

19. A patrol inspection device for a spinning workshop, comprising: The spinning workshop includes a plurality of spinning boxes arranged along a first direction, and has a patrol inspection passage on one side of the plurality of spinning boxes, and a patrol inspection robot is movable along the patrol inspection passage to sequentially inspect the operation states of the plurality of spinning boxes; The spinning workshop inspection device includes: a first acquisition unit for acquiring first image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a first position corresponding to a first spinning box in the patrol inspection passage, the first spinning box being one of the plurality of spinning boxes, and the first image information being obtained by photographing the patrol inspection robot; A first determination unit for determining a second spinning box from among the plurality of spinning boxes when the first image information includes a first inspection object, and the inspection order of the second spinning box is later than the inspection order of the first spinning box; a second acquisition unit for acquiring second image information of the patrol inspection passage when the patrol inspection robot patrols and inspects a second position corresponding to the second spinning box in the patrol inspection passage, the second image information being acquired by photographing the patrol inspection robot, and the second image information being an image including a position where the first inspection target is located in the first image information; a second determination unit for determining state information of the first inspection object based on the first image information and the second image information. Patrol inspection equipment for spinning workshop.

20. At least one processor; a memory in communication with the at least one processor; The memory stores instructions executable by the at least one processor, the instructions, when executed by the at least one processor, causing the at least one processor to perform a method according to any one of claims 1 to 18. Electronic devices.

21. A non-transitory computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any one of claims 1 to 18.

22. In a computer, a processor executes A program for implementing the method according to any one of claims 1 to 18.

Images