Method, device and system for identifying abnormal operation of kitchen

Abstract

The invention discloses a method, device and system for identifying abnormal operation of a kitchen, and the method comprises the steps: obtaining a target image of a target environment region in the kitchen, the target environment region at least comprising a detection region; identifying a first article and a second article in the detection area, and determining whether the first article and the second article intersect according to the position information of the first article and the second article; wherein the first article is a tool used for processing food materials, the second article is a movable article, and the second article is forbidden to be placed on the first article; and if the first article intersects with the second article, outputting an alarm instruction. According to the invention, the efficiency of identifying illegal operation behaviors of kitchen workers can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer vision technology, and in particular to a method, device and system for identifying abnormal operations in a kitchen. Background Art

[0002] In the back kitchen of a restaurant, there are strict hygiene requirements for kitchen staff, one of which is that they are not allowed to wash brooms, mops, dustpans and other cleaning utensils in the food sink. Currently, such illegal operations can only be avoided through corporate management methods, which increases costs, increases management difficulties and reduces efficiency. Summary of the invention

[0003] Based on this, it is necessary to provide a method, device and system for identifying abnormal operations in the back kitchen in response to the above technical problems, which can solve the management difficulties of illegal operations in the back kitchen, improve the efficiency of identifying illegal operations in the back kitchen, and reduce management costs and difficulties.

[0004] In a first aspect, the present application provides a method for identifying abnormal operation in a back kitchen, the method comprising: Acquire a target image of a target environment area in the kitchen, where the target environment area includes at least one detection area; Identify a first object and a second object in a detection area, and determine whether the first object and the second object intersect according to position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; If the first object intersects with the second object, an alarm instruction is output.

[0005] In one embodiment, the first image acquisition device is configured to be able to identify the first object and the second object, the second image acquisition device is configured to be able to identify the first object and the second object, the first image acquisition device and the second image acquisition device are installed at different positions and at different shooting angles, and the method further includes: The first image acquisition device acquires and obtains a first target image of the target environment area, identifies a first object and a second object in a detection area of ​​the first target image, and determines whether the first object and the second object intersect; The second image acquisition device acquires and obtains a second target image of the target environment area, identifies the first object and the second object in the detection area of ​​the second target image, and determines whether the first object and the second object intersect; If the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, an alarm instruction is output.

[0006] In one embodiment, the first image acquisition device is a first camera, the second image acquisition device is a second camera, and the method further includes: When the first camera does not determine that the first object and the second object intersect, the second camera captures a second target image at a first time interval; When the first camera determines that the first object and the second object intersect, the second camera captures a second target image at a second time interval, wherein the second time interval is smaller than the first time interval.

[0007] In one embodiment, if the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, outputting an alarm instruction includes: The first image acquisition device determines that the first object and the second object intersect, and outputs a first recognition result to the server; The second image acquisition device determines that the first object and the second object intersect, and outputs a second recognition result to the server; The server outputs an alarm instruction in response to the first recognition result and the second recognition result.

[0008] In one embodiment, the method further comprises: Acquire false alarm information fed back by users in response to alarm instructions within a preset time period, determine movable objects with a false alarm rate higher than a preset false alarm rate threshold, and divide the movable objects into at least a first part and a second part, wherein, in the false alarm information fed back by users: the number of times the first part is determined to intersect with the first object is greater than a first threshold, or the proportion of the number of images of the first part determined to intersect with the first object in the number of false alarm images fed back by users of the intersection of the first object and the second object is greater than a second threshold; If it is determined that the first object intersects with the first part of the movable object, no warning instruction is output.

[0009] In one embodiment, the method further comprises: Acquire a plurality of specific false alarm images, where the specific false alarm images are false alarm images whose first portion is determined to intersect with the first object; marking intersection positions of the first part and the first object in a plurality of specific false alarm images respectively; Dividing the first part into at least a third subpart and a fourth subpart according to the intersection position of the marked first part and the first object, wherein the number of times the third subpart is determined to intersect with the first object is greater than a third threshold, or the proportion of the number of images of the third subpart that are determined to intersect with the first object in the number of specific false positive images is greater than a fourth threshold; If, in the acquired target image, it is determined that the third sub-portion intersects with the first object, no warning instruction is output; If it is determined in the acquired target image that the fourth sub-portion intersects with the first object, an alarm instruction is output.

[0010] In one embodiment, the first object is a first cleaning container for cleaning food, and the second object is a floor cleaning tool, and the method further includes: Determine a first detection area in the target image, and determine whether a first cleaning container and a floor cleaning tool exist in the first detection area at the same time, wherein the first detection area includes at least one first cleaning container; If the first cleaning container and the floor cleaning tool exist in the first detection area at the same time, determine whether the first cleaning container and the floor cleaning tool intersect, or whether the first cleaning container contains the floor cleaning tool, and if they intersect or contain, output a first alarm instruction; and / or, The first object includes a second cleaning container and a third cleaning container, the second object includes tableware and food, the second cleaning container is used to clean the tableware, and the third cleaning container is used to clean the food, and the method further includes: Determine a second detection area in the target image, wherein the second detection area includes a second cleaning container and a third cleaning container; If it is identified that the second cleaning container intersects with the food in the second detection area, or the third cleaning container intersects with the tableware, a second alarm instruction is output; and / or, The first object includes a first placement board, the second object includes a first type of food and a second type of food, the first placement board is used to process the first type of food, and the method further includes: Determine a third detection area in the target image, wherein the third detection area at least includes the first placement board; If the first placement plate and the second type of food are simultaneously identified in the third detection area, and the first placement plate and the second type of food intersect, outputting a third alarm instruction; and / or, The first object includes a second placement board, the second object includes a first type of food and a second type of food, the second placement board is used to process the second type of food, and the method further includes: Determine a fourth detection area in the target image, wherein the fourth detection area at least includes the second placement board; If the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, a fourth alarm instruction is output; and / or, The first object is a food processing board, the second object is a cleaning cloth, and the method further comprises: Determining a fifth detection area in the target image, wherein the fifth detection area includes a food processing board; If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output.

[0011] In one embodiment, determining whether the first cleaning container and the floor cleaning tool intersect, or the first cleaning container contains the floor cleaning tool, comprises: Acquire an identification frame of a first cleaning container in a first detection area; Acquire an identification frame of the floor cleaning tool in the first detection area; Determine whether there is an overlapping area between the identification frame of the first cleaning container and the identification frame of the floor cleaning tool according to the position coordinates of the identification frame of the first cleaning container in the target image and the position coordinates of the identification frame of the floor cleaning tool in the target image; If the area ratio of the overlapping region to the identification frame of the first cleaning container is greater than or equal to a preset first ratio threshold, it is determined that the first cleaning container contains the floor cleaning tool; otherwise, it is determined that the first cleaning container and the floor cleaning tool intersect; and / or, The first placement plate and the second type of food are simultaneously identified in the third detection area, and the first placement plate and the second type of food intersect, including: Acquire an identification frame of the first placement board in the third detection area; Obtaining an identification frame of the second type of food in the third detection area; Determine whether there is an overlapping area between the identification frame of the first placement plate and the identification frame of the second type of food according to the position coordinates of the identification frame of the first placement plate in the target image and the position coordinates of the identification frame of the second type of food in the target image; If the area ratio of the overlapping region to the identification frame of the first placement board is greater than or equal to a preset second ratio threshold, it is determined that the first placement board and the second type of food intersect; and / or, If the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, including: Acquire an identification frame of the second placement board in the fourth detection area; Obtaining an identification frame of the first type of food in the fourth detection area; Determine whether there is an overlapping area between the identification frame of the second placement plate and the identification frame of the first category of ingredients according to the position coordinates of the identification frame of the second placement plate in the target image and the position coordinates of the identification frame of the first category of ingredients in the target image; If the area ratio of the overlapping region to the identification frame of the second placement board is greater than or equal to a preset third ratio threshold, it is determined that the second placement board and the first type of food intersect; and / or, The food processing board is configured as a first color, and the cleaning cloth is configured as a second color, and the first color is different from the second color. If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output, including: Performing color recognition on the target object in the fifth detection area, recognizing that the target object with the first color is a food processing board, and recognizing that the target object with the second color is a cleaning cloth; If the food processing plate and the cleaning cloth intersect, the fifth alarm instruction is output.

[0012] In a second aspect, the present application provides a device for identifying abnormal operations in a back kitchen, the device comprising: An acquisition module is used to acquire a target image of a target environment area in the kitchen, where the target environment area includes at least one detection area; an identification module, used to identify a first object and a second object in a detection area, and determine whether the first object and the second object intersect according to position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The alarm module is used to output an alarm instruction if the first object intersects with the second object.

[0013] In a third aspect, the present application provides a system for identifying abnormal operations in a back kitchen, the system comprising: An image acquisition device is used to obtain a target image of a target environment area in a back kitchen, the target environment area includes at least one detection area, identify a first object and a second object in the detection area, determine whether the first object and the second object intersect according to position information of the first object and the second object, and output a recognition result if they intersect; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The server is used to output an alarm instruction in response to the recognition result.

[0014] In a fourth aspect, the present application provides a computer device including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method for identifying abnormal operations in the back kitchen as described above are implemented.

[0015] In a fifth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for identifying abnormal operations in the back kitchen as described above.

[0016] The present application obtains a target image of a target environment area in a back kitchen, identifies a detection area in the target image, identifies a second object in the detection area, wherein the first object is a tool for processing food materials, the second object is a movable object, and the second object is prohibited from being placed on the first object, and determines whether the first object and the second object intersect. If they intersect, it is determined that the back kitchen staff has violated the regulations, such as washing a mop in a sink for washing food materials. This solves the problem of difficulty in managing the illegal operations of the back kitchen staff, can automatically and effectively identify illegal operations, promptly discover the illegal operations of the staff, improve the efficiency of store management, and reduce the difficulty and cost of managing the staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A first flow chart of a method for identifying abnormal kitchen operations provided in an embodiment of the present application; Figure 2 A second flow chart of the method for identifying abnormal kitchen operations provided in an embodiment of the present application; Figure 3 A third flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Figure 4 A fourth flow chart of the method for identifying abnormal kitchen operation provided by an embodiment of the present application; Figure 5 A fifth flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Figure 6 A sixth flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Figure 7 A seventh flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Figure 8 An eighth flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Fig. 9 A ninth flow chart of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Fig.10 A tenth flow chart of a method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Fig.11 A schematic diagram of the eleventh process of the method for identifying abnormal operation in a back kitchen provided in an embodiment of the present application; Fig.12 A twelfth flow chart of the method for identifying abnormal operation in a kitchen provided in an embodiment of the present application; Fig.13A system block diagram of a device for identifying abnormal kitchen operations provided in an embodiment of the present application; Fig.14 A system block diagram of a system for identifying abnormal kitchen operations provided by an embodiment of the present application; Fig.15 A system block diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The present application will be described in detail below in conjunction with the specific implementation modes shown in the accompanying drawings, but these implementation modes do not limit the present application. Structural, methodological, or functional changes made by ordinary technicians in the field based on these implementation modes are included in the protection scope of the present application.

[0019] The embodiment of the present application provides a method for identifying abnormal operation in the back kitchen, and the execution subject of the method can be a camera or a camera, and the camera or the camera can be connected to a server. The camera or the camera can be provided with a processor, a transceiver, and a camera device, wherein the camera device can be used to capture an image, that is, to obtain a target image, the processor recognizes the obtained target image to identify whether the first object and the second object intersect, and the transceiver can be used to transmit data with the server, that is, when it is determined that the first object and the second object intersect, the recognition result is output to the server so that the server outputs an alarm instruction.

[0020] The execution subject of the method for identifying abnormal operation of the back kitchen in the embodiment of the present application may be an electronic device. The camera is connected to the electronic device. The camera is used to capture a target image. The electronic device obtains the target image captured by the camera and performs image processing to identify whether the first object and the second object intersect, and outputs an alarm instruction when it is determined that the first object and the second object intersect.

[0021] Exemplarily, the electronic device may be a terminal or other device, including but not limited to a mobile terminal and a fixed terminal. For example, a mobile terminal includes but is not limited to a smart phone, a smart watch, a tablet computer, a laptop computer, a smart vehicle, a smart car-mounted device, etc., wherein a fixed terminal includes but is not limited to a desktop computer, a smart TV, etc.

[0022] Exemplarily, the electronic device may also be a device such as a server, which may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It may also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, as well as big data and artificial intelligence platforms, but is not limited to these.

[0023] The following describes in detail the implementation of the method for identifying abnormal operations in the kitchen according to an embodiment of the present application.

[0024] Please refer to Figure 1 , an embodiment of the present application provides a method for identifying abnormal operations in a back kitchen, the method comprising steps S101-S103.

[0025] Step S101, obtaining a target image of a target environment area in a kitchen, where the target environment area includes at least one detection area.

[0026] Exemplarily, the target environmental area may be an environmental area in a back kitchen of a restaurant.

[0027] Exemplarily, the detection area may be an area where illegal operation occurs, such as an area of ​​a sink for washing food, or an area of ​​a food processing board.

[0028] For example, a target image corresponding to the target environment area is collected and acquired by a camera or camera installed in the back kitchen. The target image can be captured in real time, or it can be an image frame extracted from a video recording, or it can be an image captured at periodic intervals. The installation position of the camera can be determined according to the site conditions of the actual use site so as to be able to fully capture the behavior of the staff in the back kitchen, and this application does not limit this.

[0029] Based on the acquired target image, determine the detection area in the target image. Exemplarily, based on the environmental layout in the back kitchen, the target environmental area can be divided into regions to obtain the required detection area. The camera is generally installed in a fixed position, and based on the pre-divided area, the detection area can be directly determined in the acquired target image. The detection area under the camera's viewing angle can also be determined by image segmentation, edge detection, etc., and the detection area can also be defined in the camera by manual annotation to generate the coordinates of the detection area. The pre-generated coordinates of each detection area can be used to simply and efficiently locate the detection area in the target image, which is conducive to effectively and accurately identifying the target object in the detection area.

[0030] Step S102, identifying a first object and a second object in a detection area, and determining whether the first object and the second object intersect based on position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object.

[0031] The detection area is subjected to target object recognition, and a first object and a second object in the detection area are identified, wherein the first object is a tool for processing food, and the second object is a movable object. Exemplarily, the detection area can be subjected to image recognition by a computer vision algorithm or a deep learning algorithm to identify the first object and the second object in the image. According to the position coordinates of the first object in the target image and the position coordinates of the second object in the target image, it is determined whether the first object and the second object intersect.

[0032] The first item is a tool for processing food, and the second item is a movable item. The second item is not allowed to be placed on the first item. For example, the first item is a sink for washing food, and the second item is a floor cleaning tool, such as a broom, mop, dustpan, etc. In the hygienic operation specifications of the back kitchen, floor cleaning tools are not allowed to be placed in the sink. For example, the first item is a meat processing board, and the second item is a fruit or vegetable food. Fruit or vegetable food is not allowed to be processed on the meat processing board.

[0033] Step S103: if the first object intersects with the second object, output an alarm instruction.

[0034] If it is determined that the first object intersects with the second object, it indicates that the staff has violated the regulations and an alarm instruction is output. The alarm instruction is used to indicate that the staff has violated the regulations in the kitchen.

[0035] In this embodiment, by acquiring a target image of a target environment area in the back kitchen, identifying a detection area in the target image, identifying a second object in the detection area, determining whether the first object and the second object intersect, and determining whether a staff member has violated regulations, for example, washing a mop in a sink for washing ingredients, or placing fruit or vegetable ingredients on a meat ingredient processing board, etc., the problem of difficulty in managing the illegal operations of staff in the back kitchen is solved, management costs are reduced, illegal operations can be automatically and effectively identified, illegal operations of staff can be discovered in a timely manner, the efficiency of store management is improved, the difficulty and cost of managing staff are reduced, and the supervision effect is improved.

[0036] In some embodiments, Figure 2 As shown, the first image acquisition device is configured to be able to identify the first object and the second object, the second image acquisition device is configured to be able to identify the first object and the second object, the installation positions and shooting angles of the first image acquisition device and the second image acquisition device are different, and the method also includes steps S201-S203.

[0037] Step S201: A first image acquisition device acquires and obtains a first target image of a target environment area, identifies a first object and a second object in a detection area of ​​the first target image, and determines whether the first object and the second object intersect.

[0038] In order to improve the accuracy of identifying illegal operations in the kitchen, a first image acquisition device and a second image acquisition device are installed in the kitchen. The first image acquisition device and the second image acquisition device are installed at different positions and shooting angles to shoot the target environment area from different angles and obtain target images at different shooting angles. The first image acquisition device can be a camera or a camera, and the second image acquisition device can be a camera or a camera. The camera can be an edge camera with an image recognition function, and the recognition output result is sent to the server, which can effectively reduce the data traffic transmitted from the camera to the server.

[0039] The first image acquisition device acquires and obtains a first target image of the target environment area, determines a detection area in the first target image, performs image recognition on the detection area, identifies a first object and a second object in the detection area, and determines whether the first object and the second object intersect based on the position coordinates of the first object in the first target image and based on the position coordinates of the second object in the first target image.

[0040] Step S202: The second image acquisition device acquires and obtains a second target image of the target environment area, identifies a first object and a second object in a detection area of ​​the second target image, and determines whether the first object and the second object intersect.

[0041] The second image acquisition device acquires and obtains a second target image of the target environment area, determines a detection area in the second target image, performs image recognition on the detection area, identifies a first object and a second object in the detection area, and determines whether the first object and the second object intersect based on the position coordinates of the first object in the second target image and the position coordinates of the second object in the second target image. It should be noted that the detection area in the first target image is consistent with the detection area in the second target image.

[0042] Step S203: the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, and outputs an alarm instruction.

[0043] When the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, it indicates that a staff member has violated the operating rules, and an alarm instruction is output.

[0044] In this embodiment, two image acquisition devices are installed to photograph the target environment area from different shooting angles, and object recognition is performed on the same detection area in the respectively acquired target images. When the first object and the second object are both recognized to intersect, it is determined that a staff member has violated the regulations. This can improve the accuracy of identifying illegal operations, reduce the false alarm rate of recognition errors, and further improve the efficiency of store management.

[0045] In some embodiments, the first image acquisition device is a first camera, the second image acquisition device is a second camera, and the method further includes: when the first camera has not determined that the first object and the second object intersect, the second camera captures the second target image at a first time interval; when the first camera determines that the first object and the second object intersect, the second camera captures the second target image at a second time interval, wherein the second time interval is smaller than the first time interval.

[0046] When the first camera does not determine that the first object and the second object intersect, it indicates that there is no illegal operation behavior at present, and there is no need to identify whether there is illegal operation behavior at this time. Therefore, the second camera captures the second target image at a longer first time interval to reduce equipment consumption. When the first camera determines that the first object and the second object intersect, it indicates that there is illegal operation behavior at present, which is a peak time period for illegal operation behavior. Therefore, the second camera captures the second target image at a shorter second time interval to ensure that the complete behavior monitoring process can be accurately monitored and improve the accuracy of recognition. The first time interval and the second time interval can be set according to empirical values.

[0047] In some embodiments, if the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, outputting an alarm instruction includes: the first image acquisition device determines that the first object and the second object intersect, and outputs a first recognition result to the server; the second image acquisition device determines that the first object and the second object intersect, and outputs a second recognition result to the server; the server outputs an alarm instruction in response to the first recognition result and the second recognition result. The first recognition result is used to indicate that the first object and the second object intersect, and the second recognition result is used to indicate that the first object and the second object intersect. In this embodiment, when the server receives the first recognition result sent by the first image acquisition device and simultaneously receives the second recognition result sent by the second image acquisition device, it indicates that the image acquisition device detects illegal operation behavior and outputs an alarm instruction to remind the staff that they are operating in violation of the regulations and reduce the false alarm rate of the alarm information.

[0048] In some embodiments, Figure 3 As shown, the method also includes the steps of: Step S301, obtaining false alarm information fed back by users in response to alarm instructions within a preset time period, determining movable objects with a false alarm rate higher than a preset false alarm rate threshold, and dividing the movable objects into at least a first part and a second part, wherein, in the false alarm information fed back by the user: the number of times the first part is determined to intersect with the first object is greater than a first threshold, or the proportion of the number of images in the first part determined to intersect with the first object in the number of false alarm images fed back by the user of the intersection of the first object and the second object is greater than a second threshold; Step S302: if it is determined that the first object intersects with the first part of the movable object, no alarm instruction is output.

[0049] In one of the application scenarios of this embodiment, the second object is a movable object. When the staff moves the object, they may accidentally touch the first object, resulting in a false alarm. For example, the movable object is a mop. When the staff is mopping the floor, the top of the mop rod may often touch the sink, resulting in it being judged as intersecting with the sink, which is prone to false alarms. Therefore, for objects that are prone to false alarms, the objects are segmented for identification to reduce the false alarm rate.

[0050] Collect the false alarm information of the alarm command fed back by the user, obtain the false alarm information within the preset time period, analyze the obtained false alarm information, calculate the false alarm rate of each movable object, and thus determine the movable objects whose false alarm rate is higher than the preset false alarm rate threshold. The preset time period can be set according to the actual situation. The false alarm rate threshold can be set according to the experience value.

[0051] Based on the false alarm information fed back by the user, the movable object is divided into at least a first part and a second part. The number of times the first part is determined to intersect with the second object is greater than a first threshold, or the proportion of the number of images of the first part determined to intersect with the second object in the number of false alarm images fed back by the user of the first object intersecting with the second object is greater than a second threshold.

[0052] For example, in the feedback false alarm information, if the number of times the first part of the movable object intersects with the first object is greater than the first threshold, it indicates that the first part of the movable object is prone to false alarms, and in subsequent processing, when it is determined that the first object intersects with the first part of the movable object, no alarm instruction is output. The first threshold can be set according to an empirical value.

[0053] For example, the number of false alarm images of the first object intersecting with the first object fed back by the user is obtained, and among these false alarm images, the number of images in which the first part of the movable object is determined to intersect with the first object accounts for a proportion greater than a second threshold value in the number of false alarm images, indicating that the first part of the movable object is prone to false alarms, and in subsequent processing, when it is determined that the first object intersects with the first part of the movable object, no alarm instruction is output. The second threshold value can be set according to an empirical value.

[0054] In this embodiment, movable objects that are prone to false alarms are identified in sections or parts, and a part of the movable object that is prone to false alarms is determined. When the part intersects with the first object, no alarm is generated. For example, in the above-mentioned mop, the top of the mop rod is prone to false alarms. The mop is divided into a mop rod and a mop head. There is no alarm when the mop rod intersects with the sink, but an alarm is generated when the mop head intersects with the sink. This can improve the accuracy of identifying illegal operating behaviors, reduce the false alarm rate, and further improve the efficiency of store management.

[0055] In some embodiments, Figure 4 As shown, the method also includes the steps of: Step S401, obtaining a plurality of specific false alarm images, where the specific false alarm images are false alarm images whose first portion is determined to intersect with the first object; Step S402, marking the intersection positions of the first part and the first object in a plurality of specific false alarm images respectively; Step S403, dividing the first part into at least a third sub-part and a fourth sub-part according to the intersection positions of the marked first part and the first object, wherein the third sub-part is determined as the number of times the first object intersects is greater than a third threshold, or the third sub-part is determined as the number of images intersecting with the first object accounts for a proportion of the number of specific false positive images greater than a fourth threshold; Step S404: if it is determined that the third sub-portion intersects with the first object in the acquired target image, no alarm instruction is output; Step S405: if it is determined that the fourth sub-portion intersects with the first object in the acquired target image, an alarm instruction is output.

[0056] Acquire multiple specific false alarm images, where the first part of the specific false alarm image is determined to intersect with the second part, and perform image analysis on the specific false alarm image to more accurately identify the position where the first part of the movable object intersects with the first object. Annotate the specific false alarm image, and annotate the intersection position manually or automatically. According to the annotated intersection position, divide the first part into at least a third sub-part and a fourth sub-part manually or automatically. Perform image recognition on the acquired target image, and if it is identified that the third sub-part in the detection area intersects with the first object, no alarm instruction is output. Perform image recognition on the acquired target image, and if it is identified that the fourth sub-part in the detection area intersects with the first object, an alarm instruction is output.

[0057] For example, if the number of times the third sub-portion of the movable object intersects with the first object in the marked position information is greater than the third threshold, it indicates that the third sub-portion of the movable object is prone to false alarms, and in subsequent processing, when it is determined that the first object intersects with the third sub-portion of the movable object, no alarm instruction is output. The third threshold can be set according to an empirical value.

[0058] Exemplarily, the number of specific false alarm images is obtained, and in these specific false alarm images, the number of images in which the third part of the movable object is determined to intersect with the first object accounts for a proportion greater than a fourth threshold value in the number of specific false alarm images, indicating that the third sub-part of the movable object is prone to false alarms, and in subsequent processing, when it is determined that the first object intersects with the third sub-part of the movable object, no alarm instruction is output. The fourth threshold value can be set according to an empirical value.

[0059] In this embodiment, by marking the intersection position of the movable object and the first object in the false alarm image, the first part of the movable object that is prone to false alarms is further segmented or partially identified. When the third sub-part intersects with the first object, no alarm is generated, and when the fourth sub-part intersects with the first object, an alarm is generated. For example, the above-mentioned mop rod is divided into two parts, the top of the mop rod to 1 / 3 of the mop rod is the third sub-part, and the remaining part of the mop rod is the fourth sub-part, so that illegal operations can be identified more accurately, which can reduce both false alarms and missed alarms.

[0060] In some embodiments, Figure 5 As shown, the first object is a first cleaning container for cleaning food, and the second object is a floor cleaning tool. The method further includes the steps of: Step S501, determining a first detection area in a target image, and determining whether a first cleaning container and a floor cleaning tool exist in the first detection area at the same time, wherein the first detection area includes at least one first cleaning container; Step S502: if the first cleaning container and the floor cleaning tool exist in the first detection area at the same time, determine whether the first cleaning container and the floor cleaning tool intersect, or whether the first cleaning container contains the floor cleaning tool; if so, output a first alarm instruction.

[0061] One of the application scenarios of this embodiment is that the first item is a first cleaning container for cleaning food, for example, the first cleaning container may be a sink or a pool, and the second item is a floor cleaning tool, for example, the floor cleaning tool may be a mop or a broom, and the floor cleaning tool is prohibited from being placed in the first cleaning container.

[0062] Obtain a target image of the target environment area of ​​the back kitchen, determine the first detection area in the target image, perform image recognition on the first detection area, and identify whether the first cleaning container and the floor cleaning tool exist in the first detection area at the same time, and the first detection area includes at least one cleaning container. A convolutional neural network for target detection can be used to identify the target object in the image. If it is identified that the first cleaning container and the floor cleaning tool exist in the first detection area at the same time, based on the position coordinates of the first cleaning container in the target image and the position coordinates of the floor cleaning tool in the target image, determine whether the first cleaning container and the floor cleaning tool intersect, or whether the first cleaning container contains the floor cleaning tool. If they intersect or contain, it indicates that the floor cleaning tool is partially or completely placed in the first cleaning container, and it is determined that there is currently an illegal operation behavior, and an alarm instruction is output to remind the staff.

[0063] In some embodiments, Figure 6 As shown, determining whether the first cleaning container and the floor cleaning tool intersect, or the first cleaning container contains the floor cleaning tool, comprises the steps of: Step S601, obtaining an identification frame of a first cleaning container in a first detection area; Step S602, obtaining an identification frame of the floor cleaning tool in the first detection area; Step S603, determining whether there is an overlapping area between the identification frame of the first cleaning container and the identification frame of the floor cleaning tool according to the position coordinates of the identification frame of the first cleaning container in the target image and the position coordinates of the identification frame of the floor cleaning tool in the target image; Step S604: if the area ratio of the overlapping region to the identification frame of the first cleaning container is greater than or equal to a preset first ratio threshold, it is determined that the first cleaning container contains the floor cleaning tool; otherwise, it is determined that the first cleaning container and the floor cleaning tool intersect.

[0064] Exemplarily, a convolutional neural network for target detection, such as a YOLO network, can be used to detect items in the first detection area to obtain an item target rectangular frame, each item corresponds to an item target rectangular frame, and the corresponding target object is determined based on the item target rectangular frame, that is, the identification frame of the first cleaning container and the identification frame of the floor cleaning tool in the first detection area are determined. Alternatively, other target detection algorithms can also be used to determine the target object in the image. The specific detection method is not limited. According to the position coordinates of the identification frame of the first cleaning container in the target image, and the position coordinates of the identification frame of the floor cleaning tool in the target image, it is determined whether there is an overlapping area between the identification frame of the first cleaning container and the identification frame of the floor cleaning tool. If the overlapping area accounts for a proportion of the area of ​​the identification frame of the first cleaning container that is greater than or equal to the first proportion threshold, it is determined that the first cleaning container contains a floor cleaning tool, otherwise it is determined that the first cleaning container and the floor cleaning tool intersect, and based on this implementation, it is determined whether the floor cleaning tool is placed in the first cleaning container to provide a judgment basis for subsequent processing. The first proportion threshold can be set according to an empirical value.

[0065] In some embodiments, Figure 7 As shown, the first item includes a second cleaning container and a third cleaning container, the second item includes tableware and food, the second cleaning container is used to clean the tableware, and the third cleaning container is used to clean the food, and the method further includes the steps of: Step S701, determining a second detection area in the target image, wherein the second detection area includes a second cleaning container and a third cleaning container; Step S702: If it is identified that the second cleaning container intersects with the food in the second detection area, or the third cleaning container intersects with the tableware, a second alarm instruction is output.

[0066] One of the application scenarios of this embodiment is that the first item is the second cleaning container and the third cleaning container, the second item includes tableware and food, the second cleaning container is used to clean tableware, the third cleaning container is used to clean food, the second cleaning container is prohibited from cleaning food, the third cleaning container is prohibited from cleaning tableware, and the containers for cleaning tableware and cleaning food are separated. Based on this implementation, when the cleaning container for cleaning tableware cleans food, or the cleaning container for cleaning food cleans tableware, it is identified that there is currently an illegal operation behavior and an alarm message is output. The convolutional neural network of target detection can be used to identify the second cleaning container and the third cleaning container in the second detection area, as well as the identification of food and tableware, which will not be repeated here.

[0067] In some embodiments, Figure 8 As shown, the first object includes a first placement board, the second object includes a first type of food and a second type of food, the first placement board is used to process the first type of food, and the method further includes the steps of: Step S801, determining a third detection area in the target image, wherein the third detection area at least includes a first placement board; Step S802: If the first placement board and the second type of food are simultaneously identified in the third detection area, and the first placement board and the second type of food intersect, a third alarm instruction is output.

[0068] In one of the application scenarios of this embodiment, the first item is a first placement board, the second item includes a first type of food and a second type of food, and the first placement board is used to process the first type of food, for example, the first placement board is a meat processing board, the first food is meat or aquatic products, and the second type of food is fruit or vegetables. Fruit or vegetables are prohibited from being processed on the meat processing board, and the food processing boards need to be separated. Based on this implementation, when the first placement board that processes the first type of food processes the second type of food, it is identified that there is currently an illegal operation behavior and an alarm message is output.

[0069] In some embodiments, Fig. 9 As shown, in the third detection area, the first placement plate and the second type of food are simultaneously identified, and the first placement plate and the second type of food intersect, including the steps of: Step S901, obtaining an identification frame of the first placement board in the third detection area; Step S902, obtaining an identification frame of the second type of food in the third detection area; Step S903, determining whether there is an overlapping area between the identification frame of the first placement board and the identification frame of the second type of food according to the position coordinates of the identification frame of the first placement board in the target image and the position coordinates of the identification frame of the second type of food in the target image; Step S904: if the area ratio of the overlapping region to the identification frame of the first placement board is greater than or equal to a preset second ratio threshold, it is determined that the first placement board and the second type of food intersect.

[0070] Exemplarily, a convolutional neural network for target detection can be used to perform object detection on the third detection area to determine the identification frame of the first placement board and the identification frame of the second type of food in the third detection area. According to the position coordinates of the identification frame of the first placement board in the target image, and the position coordinates of the identification frame of the second type of food in the target image, it is determined whether there is an overlapping area between the identification frame of the first placement board and the identification frame of the second type of food. If the proportion of the area of ​​the identification frame of the first placement board occupied by the overlapping area is greater than or equal to the second proportion threshold, it is determined that the first placement board and the second type of food intersect, and based on this implementation, it is determined whether the second type of food is placed on the first placement board to provide a judgment basis for subsequent processing. The second proportion threshold can be set according to an empirical value.

[0071] In some embodiments, Fig.10As shown, the first object includes a second placement board, the second object includes a first type of food and a second type of food, the second placement board is used to process the second type of food, and the method further includes the steps of: Step S1001, determining a fourth detection area in the target image, wherein the fourth detection area at least includes a second placement board; Step S1002: If the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, a fourth alarm instruction is output.

[0072] In one of the application scenarios of this embodiment, the first item is the second placement board, the second item includes the first category of food and the second category of food, and the second placement board is used to process the second category of food. For example, the second placement board is a vegetable processing board, the first food is meat or aquatic products, and the second category of food is fruit or vegetables. Meat or aquatic products are prohibited from being processed on the vegetable processing board, and the boards for processing food need to be separated. Based on this implementation, when the second placement board for processing the second category of food processes the first category of food, it is identified that there is currently an illegal operation behavior and an alarm message is output.

[0073] In some embodiments, Fig.11 As shown, if the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, the steps include: Step S1101, obtaining an identification frame of the second placement board in the fourth detection area; Step S1102, obtaining an identification frame of the first type of food in the fourth detection area; Step S1103, determining whether there is an overlapping area between the identification frame of the second placement board and the identification frame of the first category of ingredients according to the position coordinates of the identification frame of the second placement board in the target image and the position coordinates of the identification frame of the first category of ingredients in the target image; Step S1104: if the area ratio of the overlapping region to the identification frame of the second placement board is greater than or equal to a preset third ratio threshold, it is determined that the second placement board and the first type of food intersect.

[0074] Exemplarily, a convolutional neural network for target detection can be used to perform object detection on the fourth detection area to determine the identification frame of the second placement board and the identification frame of the first category of food in the fourth detection area. According to the position coordinates of the identification frame of the second placement board in the target image, and the position coordinates of the identification frame of the first category of food in the target image, it is determined whether there is an overlapping area between the identification frame of the second placement board and the identification frame of the first category of food. If the area ratio of the overlapping area to the area of ​​the identification frame of the second placement board is greater than or equal to the third ratio threshold, it is determined that the second placement board and the first category of food intersect, and based on this implementation, it is determined whether the first category of food is placed on the second placement board to provide a judgment basis for subsequent processing. The third ratio threshold can be set according to an empirical value. The third detection area and the fourth detection area may be the same or different.

[0075] In some embodiments, Fig.12 As shown, the first object is a food processing board, and the second object is a cleaning cloth. The method further comprises the steps of: Step S1201, determining a fifth detection area in the target image, wherein the fifth detection area includes a food processing board; Step S1202: If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output.

[0076] In one application scenario of this embodiment, the first item is a food processing board, and the second item includes a cleaning cloth, which is not allowed to be placed on the food processing board. Based on this implementation, when it is identified that a cleaning cloth is placed on the food processing board, it is determined that there is currently an illegal operation behavior, and an alarm message is output.

[0077] In some embodiments, the food processing board is configured as a first color, and the cleaning cloth is configured as a second color, and the first color is different from the second color. If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output, including: performing color recognition on the target object in the fifth detection area, identifying the target object of the first color as the food processing board, and identifying the target object of the second color as the cleaning cloth; if the food processing board and the cleaning cloth intersect, outputting the fifth alarm instruction.

[0078] In this embodiment, by configuring different colors for the food processing board and the cleaning cloth, it is possible to identify whether the cleaning cloth is placed on the food processing board, so as to provide a judgment basis for subsequent processing.

[0079] Based on the same inventive concept, the embodiment of the present application also provides a device for realizing the above-mentioned identification of abnormal operation in the back kitchen. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in the embodiment of the device for identifying abnormal operation in the back kitchen provided below can refer to the limitations of the method for identifying abnormal operation in the back kitchen above, and will not be repeated here.

[0080] Please refer to Fig.13 The present application provides a device for identifying abnormal operation in a back kitchen, the device comprising: An acquisition module 1301 is used to acquire a target image of a target environment area in a back kitchen, where the target environment area includes at least one detection area; The identification module 1302 is used to identify a first object and a second object in the detection area, and determine whether the first object and the second object intersect according to the position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The alarm module 1303 is used to output an alarm instruction if the first object intersects with the second object.

[0081] In this embodiment, by acquiring a target image of a target environment area in the back kitchen, identifying the detection area in the target image, identifying the second object in the detection area, determining whether the first object and the second object intersect, and determining whether the staff has violated the regulations, the problem of difficult management of the staff's illegal operations in the back kitchen is solved, management costs are reduced, and illegal operations can be automatically and effectively identified, illegal operations of staff can be discovered in a timely manner, the efficiency of store management is improved, the difficulty and cost of managing staff are reduced, and the supervision effect is improved.

[0082] Please refer to Fig.14 The present application provides a system for identifying abnormal operations in a back kitchen, the system comprising: The image acquisition device 1401 is used to obtain a target image of a target environment area in a back kitchen, the target environment area includes at least one detection area, identify a first object and a second object in the detection area, determine whether the first object and the second object intersect according to the position information of the first object and the second object, and if they intersect, output a recognition result; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The server 1402 is used to output an alarm instruction in response to the recognition result.

[0083] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, any one of the above-mentioned methods for identifying abnormal operations in the back kitchen is implemented.

[0084] Fig.15 It is a schematic diagram of the hardware structure of a computer device provided in an embodiment of the present application. Fig.15 The computer device shown includes: a processor 1501, a communication interface 1502, a memory 1503 and a communication bus 1504. The processor 1501, the communication interface 1502 and the memory 1503 communicate with each other via the communication bus 1504. Fig.15 The connection method between the processor 1501, the communication interface 1502, and the memory 1503 shown is merely exemplary. During implementation, the processor 1501, the communication interface 1502, and the memory 1503 may also be connected to each other in communication with each other using other connection methods besides the communication bus 1504.

[0085] The memory 1503 can be used to store a computer program, which may include instructions and data to implement the steps of any of the above methods for identifying abnormal operations in the kitchen. In an embodiment of the present application, the memory 1503 may be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), flash memory, optical storage, and registers. The memory 1503 may include a hard disk and / or a memory.

[0086] Processor 1501 may be a general-purpose processor, which may be a processor that performs specific steps and / or operations by reading and executing a computer program (e.g., a computer program) stored in a memory (e.g., memory 1503). The general-purpose processor may use data stored in the memory (e.g., memory 1503) in the process of performing the steps and / or operations. A general-purpose processor may be, for example, but not limited to, a central processing unit (CPU). In addition, processor 1501 may also be a dedicated processor, which may be a processor specially designed to perform specific steps and / or operations. A dedicated processor may be, for example, but not limited to, an ASIC and an FPGA. In addition, processor 1501 may also be a combination of multiple processors, such as a multi-core processor.

[0087] The communication interface 1502 may include an input / output (I / O) interface, a physical interface, and a logical interface for interconnecting devices within the network device, as well as an interface for interconnecting the network device with other devices (e.g., network devices). The communication network may be Ethernet, a radio access network (RAN), a wireless local area network (WLAN), etc. The communication interface 1502 may be a module, a circuit, a transceiver, or any device capable of implementing communication.

[0088] In the implementation process, each step of the above method can be completed by an integrated logic circuit of the hardware in the processor 1501 or an instruction in the form of software. The method for identifying abnormal operation of the back kitchen in combination with the embodiment of the present application can be directly embodied as a hardware processor, or a combination of hardware and software modules in the processor. The software module can be located in a mature storage medium in the field such as a random access memory flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 1503, and the processor 1501 reads the information in the memory 1503 and completes the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

[0089] Although the preferred embodiments of the present application have been disclosed for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the application as disclosed in the accompanying claims.

Claims

1. A method for identifying abnormal operation in a kitchen, characterized in that: The method comprises: Acquire a target image of a target environment area in a back kitchen, wherein the target environment area includes at least one detection area; Identify a first object and a second object in the detection area, and determine whether the first object and the second object intersect according to position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; If the first object intersects with the second object, an alarm instruction is output.

2. The method for identifying abnormal kitchen operation according to claim 1, characterized in that: The first image acquisition device is configured to be able to identify the first item and the second item, and the second image acquisition device is configured to be able to identify the first item and the second item, and the first image acquisition device and the second image acquisition device are installed at different positions and at different shooting angles, and the method further includes: The first image acquisition device acquires and obtains a first target image of the target environment area, identifies a first object and a second object in the detection area of ​​the first target image, and determines whether the first object and the second object intersect; The second image acquisition device acquires and obtains a second target image of the target environment area, identifies a first object and a second object in the detection area of ​​the second target image, and determines whether the first object and the second object intersect; If the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, the alarm instruction is output.

3. The method for identifying abnormal kitchen operation according to claim 2, characterized in that: The first image acquisition device is a first camera, the second image acquisition device is a second camera, and the method further includes: When the first camera does not determine that the first object and the second object intersect, the second camera captures the second target image at a first time interval; When the first camera determines that the first object and the second object intersect, the second camera captures the second target image at a second time interval, wherein the second time interval is smaller than the first time interval.

4. The method for identifying abnormal kitchen operation according to claim 2, characterized in that: If the first image acquisition device determines that the first object and the second object intersect, and the second image acquisition device determines that the first object and the second object intersect, outputting the warning instruction includes: The first image acquisition device determines that the first object and the second object intersect, and outputs a first recognition result to a server; The second image acquisition device determines that the first object and the second object intersect, and outputs a second recognition result to the server; The server outputs an alarm instruction in response to the first recognition result and the second recognition result.

5. The method for identifying abnormal kitchen operation according to claim 1, characterized in that: The method further comprises: Acquire false alarm information fed back by the user in response to the alarm instruction within a preset time period, determine movable objects with a false alarm rate higher than a preset false alarm rate threshold, and divide the movable objects into at least a first part and a second part, wherein, in the false alarm information fed back by the user: the number of times the first part is determined to intersect with the first object is greater than a first threshold, or the proportion of the number of images of the first part determined to intersect with the first object in the number of false alarm images of the first object intersecting with the second object fed back by the user is greater than a second threshold; If it is determined that the first object intersects with the first part of the movable object, no alarm instruction is output.

6. The method for identifying abnormal kitchen operation according to claim 5, characterized in that: The method further comprises: Acquire a plurality of specific false alarm images, where the specific false alarm images are false alarm images in which the first portion is determined to intersect with the first object; Marking intersection positions of the first part and the first object in a plurality of the specific false alarm images respectively; Dividing the first part into at least a third subpart and a fourth subpart according to the intersection position of the marked first part and the first object, wherein the number of times the third subpart is determined to intersect with the first object is greater than a third threshold, or the proportion of the number of images of the third subpart that are determined to intersect with the first object in the number of specific false positive images is greater than a fourth threshold; If it is determined in the acquired target image that the third sub-portion intersects with the first object, no warning instruction is output; If it is determined in the acquired target image that the fourth sub-portion intersects with the first object, an alarm instruction is output.

7. The method for identifying abnormal kitchen operation according to claim 1, characterized in that: The first object is a first cleaning container for cleaning food, the second object is a floor cleaning tool, and the method further includes: Determine a first detection area in the target image, and determine whether a first cleaning container and a floor cleaning tool exist in the first detection area at the same time, wherein the first detection area includes at least one first cleaning container; If the first cleaning container and the floor cleaning tool exist in the first detection area at the same time, determine whether the first cleaning container and the floor cleaning tool intersect, or whether the first cleaning container contains the floor cleaning tool, and if they intersect or contain, output a first alarm instruction; and / or, The first article includes a second cleaning container and a third cleaning container, the second article includes tableware and food, the second cleaning container is used to clean tableware, and the third cleaning container is used to clean food, and the method further includes: Determining a second detection area in the target image, wherein the second detection area includes the second cleaning container and a third cleaning container; If it is identified that the second cleaning container intersects with the food in the second detection area, or that the third cleaning container intersects with the tableware, a second alarm instruction is output; and / or, The first object includes a first placement board, the second object includes a first type of food and a second type of food, the first placement board is used to process the first type of food, and the method further includes: Determine a third detection area in the target image, wherein the third detection area at least includes the first placement board; If the first placement plate and the second type of food are simultaneously identified in the third detection area, and the first placement plate and the second type of food intersect, outputting a third alarm instruction; and / or, The first object includes a second placement board, the second object includes a first type of food and a second type of food, the second placement board is used to process the second type of food, and the method further includes: Determine a fourth detection area in the target image, wherein the fourth detection area at least includes the second placement board; If the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, outputting a fourth alarm instruction; and / or, The first object is a food processing plate, the second object is a cleaning cloth, and the method further comprises: Determining a fifth detection area in the target image, wherein the fifth detection area includes the food processing board; If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output.

8. The method for identifying abnormal kitchen operation according to claim 7, characterized in that: The determining whether the first cleaning container and the floor cleaning tool intersect, or whether the first cleaning container contains the floor cleaning tool, comprises: Acquire an identification frame of a first cleaning container in the first detection area; Acquire an identification frame of the floor cleaning tool in the first detection area; Determine whether there is an overlapping area between the identification frame of the first cleaning container and the identification frame of the floor cleaning tool according to the position coordinates of the identification frame of the first cleaning container in the target image and the position coordinates of the identification frame of the floor cleaning tool in the target image; If the area ratio of the overlapping region to the identification frame of the first cleaning container is greater than or equal to a preset first ratio threshold, it is determined that the first cleaning container contains the floor cleaning tool; otherwise, it is determined that the first cleaning container and the floor cleaning tool intersect; and / or, If the first placement plate and the second type of food are simultaneously identified in the third detection area, and the first placement plate and the second type of food intersect, the method includes: Acquire an identification frame of the first placement board in the third detection area; Acquire an identification frame of the second type of food in the third detection area; Determine whether there is an overlapping area between the identification frame of the first placement board and the identification frame of the second category of food according to the position coordinates of the identification frame of the first placement board in the target image and the position coordinates of the identification frame of the second category of food in the target image; If the area ratio of the overlapping region to the identification frame of the first placement plate is greater than or equal to a preset second ratio threshold, it is determined that the first placement plate and the second type of food intersect; and / or, If the second placement plate and the first type of food are simultaneously identified in the fourth detection area, and the second placement plate and the first type of food intersect, the method includes: Acquire an identification frame of the second placement board in the fourth detection area; Acquire an identification frame of the first type of food in the fourth detection area; Determine whether there is an overlapping area between the identification frame of the second placement plate and the identification frame of the first category of food according to the position coordinates of the identification frame of the second placement plate in the target image and the position coordinates of the identification frame of the first category of food in the target image; If the area ratio of the overlapping region to the identification frame of the second placement plate is greater than or equal to a preset second ratio threshold, it is determined that the second placement plate and the first type of food intersect; and / or, The food processing board is configured as a first color, and the cleaning cloth is configured as a second color, and the first color is different from the second color. If the food processing board and the cleaning cloth are simultaneously identified in the fifth detection area, and the food processing board and the cleaning cloth intersect, a fifth alarm instruction is output, including: Performing color recognition on the target object in the fifth detection area, recognizing that the target object of the first color is a food processing board, and recognizing that the target object of the second color is a cleaning cloth; If the food processing plate and the cleaning cloth intersect, a fifth alarm instruction is output.

9. A device for identifying abnormal operation in the kitchen, characterized in that: The device: An acquisition module, used to acquire a target image of a target environment area in a back kitchen, wherein the target environment area includes at least one detection area; An identification module, used for identifying a first object and a second object in the detection area, and determining whether the first object and the second object intersect according to position information of the first object and the second object; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The alarm module is used to output an alarm instruction if the first object intersects with the second object.

10. A system for identifying abnormal kitchen operations, characterized in that: The system comprises: An image acquisition device is used to obtain a target image of a target environment area in a back kitchen, wherein the target environment area includes at least one detection area, identify a first object and a second object in the detection area, determine whether the first object and the second object intersect according to position information of the first object and the second object, and output a recognition result if they intersect; wherein the first object is a tool for processing food, the second object is a movable object, and the second object is prohibited from being placed on the first object; The server is used to output an alarm instruction in response to the recognition result.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for identifying abnormal operation in the back kitchen as described in any one of claims 1 to 8 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying abnormal operation in the back kitchen as described in any one of claims 1 to 8 are implemented.

Images