An attendance time determination method, device and equipment

Abstract

The application provides an attendance time determination method, device and equipment. The method comprises the following steps: acquiring a first candidate image of a target scene collected by a first type of camera, determining a first collection time and a first position corresponding to a target object in the first candidate image; acquiring an identity recognition result corresponding to the target object; if the identity recognition result indicates that there is identity information corresponding to the target object, and the identity information does not correspond to an attendance time, determining whether the target object and an unidentified object are the same object based on the first position corresponding to the target object and a second position corresponding to the unidentified object; a second collection time of a second candidate image corresponding to the unidentified object is earlier than the first collection time, and the unidentified object does not correspond to the identity information; if yes, determining the attendance time of the identity information based on the second collection time. Through the technical scheme of the application, an accurate and reliable attendance time can be obtained, and the actual attendance time of the target object can be accurately reflected.

Description

Technical Field

[0001] This application relates to the field of electronic information, and in particular to a method, apparatus and equipment for determining attendance time. Background Technology

[0002] Attendance time refers to the recording of when someone is present at work. It involves obtaining information about an individual's (such as a student or employee) attendance at a specific location and during a specific time period, including arrival and departure, lateness, early departure, and overtime. To obtain attendance time, images of a specific location can be captured by a camera, and attendance time can be analyzed based on these images. However, image-based attendance time analysis may suffer from inaccuracies and may not accurately reflect the individual's actual attendance time. Summary of the Invention

[0003] In view of this, this application provides a method, apparatus, and equipment for determining attendance time, which can accurately analyze the actual attendance time of the subject. The technical solution adopted by this application is as follows:

[0004] This application provides a method for determining attendance time, the method comprising:

[0005] Acquire a first candidate image of the target scene captured by a first type of camera, and determine the first acquisition time and first position of the target object in the first candidate image;

[0006] Obtain the identity recognition result corresponding to the target object. If the identity recognition result indicates that there is identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then determine whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object; wherein, the second acquisition time of the second candidate image corresponding to the unidentified object is earlier than the first acquisition time, and the unidentified object does not correspond to identity information;

[0007] If so, the attendance time of the identity information is determined based on the second collection time.

[0008] For example, after determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, the method further includes: if not, determining the attendance time of the identity information based on the first collection time, and updating the target object to an identified object, wherein the identified object is an object with identity information.

[0009] For example, before determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, the method further includes:

[0010] If the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information corresponds to an attendance time, then the first collection time and the first location corresponding to the target object are ignored.

[0011] For example, determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object includes:

[0012] Based on the first location, a first region of the target object in the first candidate image is determined;

[0013] Based on the second location, a second region of the unidentified object in the second candidate image is determined;

[0014] Based on the overlap information between the first region and the second region, it is determined whether the target object and the unidentified object are the same object.

[0015] For example, determining whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region includes:

[0016] Determine the overlapping area between the first region and the second region;

[0017] If the area of ​​the overlapping region is greater than a preset area threshold, then the target object and the unidentified object are determined to be the same object; otherwise, the target object and the unidentified object are determined to be different objects.

[0018] For example, after obtaining the identity recognition result corresponding to the target object, if the identity recognition result indicates that there is no identity information corresponding to the target object, the method further includes: if it is determined that the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or it is determined that the target object and the identified object are the same object based on the first position corresponding to the target object and the position corresponding to the identified object, then the first collection time and the first position corresponding to the target object are ignored; if the target object and the unidentified object are different objects, and the target object and the identified object are different objects, then the target object is determined as an unidentified object, and the first collection time and the first position corresponding to the target object are recorded.

[0019] After determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, the method further includes: if so, updating the unidentified object to an identified object and ignoring the first acquisition time and the first position corresponding to the target object.

[0020] For example, the first type of camera includes a panoramic camera, and the method further includes: acquiring a PTZ image of the target scene captured by a translation tilt zoom PTZ camera, and determining a third acquisition time corresponding to a candidate object in the PTZ image;

[0021] The step of determining the attendance time based on the first collection time includes: if the candidate object and the target object are determined to be the same object, if the first collection time is earlier than the third collection time, the first collection time is determined as the attendance time; if the first collection time is later than the third collection time, the third collection time is determined as the attendance time.

[0022] For example, determining the attendance time of the identity information based on the second collection time includes: if it is determined that the candidate object and the target object are the same object, when the second collection time is earlier than the third collection time, the second collection time is determined as the attendance time; when the second collection time is later than the third collection time, the third collection time is determined as the attendance time.

[0023] For example, determining that the candidate object and the target object are the same object includes:

[0024] The similarity between the facial features corresponding to the candidate object and the facial features corresponding to the target object is obtained. If the similarity is greater than a preset similarity threshold, the candidate object and the target object are determined to be the same object. The facial features corresponding to the candidate object are determined based on the PTZ image, and the facial features corresponding to the target object are determined based on the first candidate image.

[0025] This application provides an attendance time determination device, the device comprising:

[0026] The acquisition module is used to acquire a first candidate image of a target scene captured by a first type of camera, determine a first acquisition time and a first position corresponding to a target object in the first candidate image; and acquire an identity recognition result corresponding to the target object. If the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then the module determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object; wherein, the second acquisition time of the second candidate image corresponding to the unidentified object is earlier than the first acquisition time, and the unidentified object does not correspond to identity information.

[0027] The determination module is used to determine the attendance time of the identity information based on the second collection time if the target object and the unidentified object are the same object.

[0028] For example, the determining module is further configured to determine the attendance time of the identity information based on the first collection time if the target object and the unidentified object are different objects, and update the target object to an identified object, wherein the identified object is an object with identity information.

[0029] For example, the determining module is further configured to ignore the first collection time and the first location corresponding to the target object if the identity recognition result indicates that there is identity information corresponding to the target object and the identity information corresponds to an attendance time.

[0030] For example, when the acquisition module determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, it is specifically used to: determine the first region of the target object in the first candidate image based on the first position; determine the second region of the unidentified object in the second candidate image based on the second position; and determine whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region.

[0031] For example, when the acquisition module determines whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region, it is specifically used to determine the overlapping area between the first region and the second region; if the area of ​​the overlapping area is greater than a preset area threshold, then the target object and the unidentified object are determined to be the same object; otherwise, the target object and the unidentified object are determined to be different objects.

[0032] For example, the determining module is further configured to: if the identity recognition result indicates that there is no identity information corresponding to the target object, and the target object and the unidentified object are determined to be the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or the target object and the identified object are determined to be the same object based on the first position corresponding to the target object and the position corresponding to the identified object, then ignore the first collection time and the first position corresponding to the target object; if the target object and the unidentified object are different objects, and the target object and the identified object are different objects, then determine the target object as an unidentified object, and record the first collection time and the first position corresponding to the target object; the determining module is further configured to: after determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, if so, update the unidentified object to an identified object, and ignore the first collection time and the first position corresponding to the target object.

[0033] For example, the first type of camera includes a panoramic camera. The acquisition module is further configured to acquire a PTZ image of the target scene captured by a panning, tilting, zooming PTZ camera, and determine a third acquisition time corresponding to a candidate object in the PTZ image. When the determining module determines the attendance time of the identity information based on the first acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, if the first acquisition time is earlier than the third acquisition time, determine the first acquisition time as the attendance time; if the first acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time. When the determining module determines the attendance time of the identity information based on the second acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, if the second acquisition time is earlier than the third acquisition time, determine the second acquisition time as the attendance time; if the second acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time.

[0034] For example, when the determining module determines that the candidate object and the target object are the same object, it is specifically used to: obtain the similarity between the facial features corresponding to the candidate object and the facial features corresponding to the target object; if the similarity is greater than a preset similarity threshold, then the candidate object and the target object are determined to be the same object; wherein, the facial features corresponding to the candidate object are determined based on the PTZ image, and the facial features corresponding to the target object are determined based on the first candidate image.

[0035] This application provides an electronic device, including: a processor and a machine-readable storage medium, the machine-readable storage medium storing machine-executable instructions that can be executed by the processor; the processor is used to execute the machine-executable instructions to implement the attendance time determination method disclosed in the above example.

[0036] This application provides a machine-readable storage medium storing machine-executable instructions that can be executed by a processor; wherein the processor is used to execute the machine-executable instructions to implement the attendance time determination method disclosed in the above example.

[0037] This application provides a computer program stored in a machine-readable storage medium, which, when executed by a processor, causes the processor to implement the above-described attendance time determination method.

[0038] As can be seen from the above technical solutions, in the embodiments of this application, if it is determined that the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, then the attendance time of the target object is determined based on the second collection time corresponding to the unidentified object, rather than based on the first collection time corresponding to the target object. This results in an accurate and reliable attendance time that can accurately reflect the actual attendance time of the target object. In other words, although the identity information of the target object is not identified when the target object is first detected, the time when the target object is first detected can still be used as the attendance time by identifying the target object and the unidentified object as the same object. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating a method for determining attendance time in one embodiment of this application;

[0040] Figure 2 This is a schematic diagram of the overlapping area in one embodiment of this application;

[0041] Figure 3 This is a flowchart illustrating a method for determining attendance time in one embodiment of this application;

[0042] Figure 4 This is a flowchart illustrating a method for determining attendance time in one embodiment of this application;

[0043] Figure 5 This is a schematic diagram of the attendance time determination device in one embodiment of this application;

[0044] Figure 6 This is a hardware structure diagram of an electronic device according to one embodiment of this application. Detailed Implementation

[0045] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the application. The singular forms “a,” “the,” and “the” as used in this application and claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to any and all possible combinations comprising one or more of the associated listed items.

[0046] It should be understood that although the terms first, second, third, etc., may be used to describe various information in embodiments of this application, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" may also be interpreted as "when," "when," or "in response to a determination."

[0047] This application proposes a method for determining attendance time, used to determine the attendance time of objects within a target scenario. The target scenario can be a specific location, such as a classroom or office, without limitation. The object can be a student, employee, etc., without limitation. In this application, attendance time can refer to the attendance start time, i.e., the time when the object enters the specific location. It should be noted that the object in the target scenario of this embodiment is in a fixed position (e.g., a student remaining stationary in their seat), or the object remains stationary for a certain period of time. This embodiment is applicable to determining attendance time in such scenarios.

[0048] In this embodiment, an image acquisition device can be deployed in the target scene. This device can be a first-type camera, or it can simultaneously deploy both a first-type camera and a second-type camera. For example, the first-type camera can be a panoramic camera, and the second-type camera can be a PTZ (Pan Tilt Zoom) camera; or, the first-type camera can be a PTZ camera, and the second-type camera can be a panoramic camera; there are no limitations on this. For ease of description, in subsequent embodiments, we will use a panoramic camera as the first-type camera and a PTZ camera as the second-type camera as an example. The panoramic camera is used to acquire images of the target scene; for ease of distinction, the images acquired by the panoramic camera can be called panoramic images. The PTZ camera is used to acquire images of the target scene; for ease of distinction, the images acquired by the PTZ camera can be called PTZ images.

[0049] For example, a panoramic camera can also be called a panoramic camera. A panoramic camera is positioned in a fixed location and can capture images within its field of view. The field of view of a panoramic camera does not change; that is, the panoramic camera always captures images within the same field of view. A PTZ camera can also be called a PTZ camera. A PTZ camera can rotate to capture images from all directions. It can rotate flexibly in all directions to achieve omnidirectional image acquisition and can use zoom control to obtain more targeted images as needed.

[0050] See Figure 1 The diagram shown illustrates a method for determining attendance time, which may include:

[0051] Step 101: Obtain the first candidate image of the target scene captured by the first type of camera, and determine the first acquisition time and first position of the target object in the first candidate image.

[0052] For example, the first type of camera can be used as a panoramic camera for illustration. The first candidate image of the target scene captured by the first type of camera can also be called the first panoramic image.

[0053] For example, a panoramic camera can periodically acquire panoramic images of the target scene. Since the processing method for each panoramic image is the same, the following discussion will use the processing of one panoramic image as an example. For ease of distinction, the panoramic image currently acquired by the panoramic camera will be referred to as the first panoramic image. If the first panoramic image does not contain any objects, it will be discarded. If the first panoramic image contains at least one object, for ease of distinction, each object in the first panoramic image will be referred to as the target object. Since the processing method for each target object is the same, the following discussion will use the processing of one target object as an example.

[0054] In one possible implementation, for a target object in the first panoramic image, the first acquisition time and the first position corresponding to the target object can be determined.

[0055] The first acquisition time corresponding to the target object can be the acquisition time of the first panoramic image. For example, when the first panoramic image is acquired at time A, the first acquisition time corresponding to the target object is time A.

[0056] The first position corresponding to the target object can be its position information in the first panoramic image, such as the position information of the target object's minimum bounding rectangle. This position information can be the coordinates of the four vertices of the minimum bounding rectangle (e.g., the coordinates of the top-left, top-right, bottom-right, and bottom-left vertices), or it can be the coordinates of a single vertex of the minimum bounding rectangle, its length, and its width. Of course, the above are just examples of position information and are not limited thereto, as long as the minimum bounding rectangle of the target object can be determined based on this position information. Naturally, the first position corresponding to the target object can also be other shapes, such as a circle, etc., without restriction.

[0057] Step 102: Obtain the identity recognition result corresponding to the target object. The identity recognition result indicates whether the target object has corresponding identity information or not.

[0058] In one possible implementation, after acquiring a first panoramic image of the target scene, the image acquisition device can obtain first feature information corresponding to the target object from the first panoramic image, and determine the identity recognition result corresponding to the target object based on the first feature information. Alternatively, the image acquisition device can send the first panoramic image to a server, which can then obtain the first feature information corresponding to the target object from the first panoramic image, determine the identity recognition result corresponding to the target object based on the first feature information, and send the identity recognition result corresponding to the target object to the image acquisition device. In this embodiment, this identity recognition process is not limited, as long as the identity recognition result corresponding to the target object can be obtained.

[0059] The first feature information corresponding to the target object can be either the target object's facial features or its body features; there are no restrictions on this, as long as it can uniquely distinguish the target object.

[0060] Step 103: If the identity recognition result indicates that there is identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then determine whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object.

[0061] If not, proceed to step 104; if yes, proceed to step 105.

[0062] For example, the second acquisition time of the second candidate image (such as the second panoramic image) corresponding to the unidentified object is earlier than the first acquisition time of the first panoramic image, and the unidentified object does not correspond to identity information. That is, the unidentified object is an object in the second panoramic image, and when processing the second panoramic image (the processing method is the same as the first panoramic image), the identity information of the unidentified object is not determined based on the second panoramic image.

[0063] For example, since the panoramic camera periodically collects panoramic images of the target scene, a large number of panoramic images have been collected before the panoramic camera collects the first panoramic image. For each panoramic image, if the identity information of the object in the panoramic image is not obtained, the panoramic image is called the second panoramic image, the object is called the unidentified object, the collection time corresponding to the unidentified object is called the second collection time, the position corresponding to the unidentified object is called the second position, and the feature information corresponding to the unidentified object is called the second feature information.

[0064] Step 104: Determine the attendance time of the identity information based on the first collection time corresponding to the target object, and update the target object to the identified object. The identified object is an object with identity information.

[0065] Step 105: Determine the attendance time of the identity information based on the second acquisition time corresponding to the unidentified object, update the unidentified object to the identified object, and ignore the first acquisition time and first position corresponding to the target object, that is, no longer determine the attendance time based on the first panoramic image, and discard the first panoramic image.

[0066] For example, if the identity recognition result indicates that there is identity information corresponding to the target object, and the identity information corresponds to an attendance time, then steps 103-105 can be skipped, and the first acquisition time and first location corresponding to the target object can be directly ignored, that is, the attendance time is no longer determined based on the first panoramic image.

[0067] For example, after obtaining the identity recognition result corresponding to the target object, if the identity recognition result indicates that there is no identity information corresponding to the target object, then: if it is determined that the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or if it is determined that the target object and the identified object are the same object based on the first position corresponding to the target object and the position corresponding to the identified object, then the first acquisition time and the first position corresponding to the target object are ignored, that is, the attendance time is no longer determined based on the first panoramic image. If the target object and the unidentified object are different objects, and the target object and the identified object are different objects, then the target object is determined as an unidentified object, and the first acquisition time and the first position corresponding to the target object are recorded, that is, the first panoramic image becomes the second panoramic image, the target object becomes an unidentified object, the first acquisition time becomes the second acquisition time, and the first position becomes the second position. In this way, the attendance time can be determined by referring to these data in the subsequent attendance time determination process.

[0068] In one possible implementation, after determining the first acquisition time and first position corresponding to the target object in the first panoramic image, the processing can be divided into the following cases:

[0069] Scenario 1: If the identity information of the target object is obtained based on the first panoramic image, and the identity information already corresponds to the attendance time, it means that the target object was already in the target scene before the first acquisition time, and the identity information of the target object has been identified and the attendance time corresponding to the identity information has been recorded. That is, the attendance time of the identity information has been known. Therefore, the first panoramic image is ignored, the first acquisition time and the first position corresponding to the target object are ignored, and the processing flow of the target object ends.

[0070] For example, after obtaining the identity information of the target object, it can be determined whether that identity information corresponds to an attendance time. For instance, since identity information is unique, if the attendance time corresponding to that identity information has already been recorded before the first data collection time, then after obtaining the target object's identity information, it can be known that the identity information already corresponds to an attendance time, and case 1 can be used for processing. Conversely, if the attendance time corresponding to that identity information has not been recorded before the first data collection time, it can be known that the identity information does not correspond to an attendance time, and case 2 or case 3 can be used for processing.

[0071] Scenario 2: If the target object's identity information is obtained, but this identity information does not correspond to an attendance time, and the target object and the unidentified object are determined to be the same object based on the first location and the second location corresponding to the unidentified object, it means that the target object was already in the target scene before the first collection time. However, when the target object was in the target scene, its identity information was not identified, and the attendance time corresponding to this identity information was not recorded. For example, due to reasons such as the collection angle of the target object, accurate and reliable feature information was not obtained, and thus the target object's identity information could not be identified. Therefore, the attendance time corresponding to this identity information can be determined based on the second collection time corresponding to the unidentified object, which is the target object's attendance time. The correspondence between this identity information and this attendance time can be recorded. For example, the second collection time can be used as the target object's attendance time, instead of the first collection time corresponding to the target object. This allows the time when the target object was first detected to be used as the attendance time, resulting in an accurate and reliable attendance time that accurately reflects the target object's actual attendance time.

[0072] For example, if the target object belongs to the same object as only one unidentified object, then the second collection time corresponding to that unidentified object is taken as the target object's attendance time. Alternatively, if the target object belongs to the same object as at least two unidentified objects, then based on the second collection times corresponding to the at least two unidentified objects, the earliest second collection time is taken as the target object's attendance time.

[0073] For example, after identifying the target object and the unidentified object as the same object, the identity information of the target object can be used as the identity information of the unidentified object. That is, the identity information of the unidentified object has been obtained, and the unidentified object is updated to an identified object. This object is no longer considered an unidentified object.

[0074] For example, a personnel information database can be maintained in advance. The personnel information feature database is used to record the mapping relationship between target feature information (such as facial features, body features, or facial images) and identity information. The identity information can be a name, ID number, mobile phone number, etc., and there are no restrictions on the identity information.

[0075] After identifying the first feature information corresponding to the target object, the similarity between the first feature information and the target feature information in the personnel information database can be calculated. If the similarity is greater than a preset similarity threshold (which can be configured empirically, such as 90% or 95%), then the identity information corresponding to the target feature information is taken as the identity information of the target object, i.e., the identity information of the target object is determined based on the first feature information. If the similarity is not greater than the preset similarity threshold, then the similarity between the first feature information and the next target feature information in the personnel information database is calculated, and so on, until the identity information of the target object is determined. Alternatively, if the similarity between the first feature information and all target feature information in the personnel information database is not greater than the preset similarity threshold, then the identity information of the target object has not been determined.

[0076] In summary, after obtaining the first feature information, the identity information of the target object can be determined based on the first feature information, or the identity information of the target object cannot be determined based on the first feature information.

[0077] For example, for each unidentified object, it can be determined whether the target object and the unidentified object are the same object. In this embodiment, the determination is based on the first position corresponding to the target object and the second position corresponding to the unidentified object. For instance, based on the first position, a first region (such as a rectangular region, a circular region, etc., hereinafter, the first rectangular region will be used as an example) of the target object in the first panoramic image is determined, and based on the second position, a second region (such as a rectangular region, a circular region, etc., hereinafter, the second rectangular region will be used as an example) of the unidentified object in the second panoramic image is determined. Based on the overlap information between the first rectangular region and the second rectangular region, it is determined whether the target object and the unidentified object are the same object.

[0078] The first position corresponding to the target object can be the position information of the target object in the first panoramic image, such as the position information of the minimum bounding rectangle of the target object. A first rectangular region can be determined based on the first position, and the first rectangular region can be the minimum bounding rectangle of the target object in the first panoramic image. The second position corresponding to the unidentified object can be the position information of the unidentified object in the second panoramic image, such as the position information of the minimum bounding rectangle of the unidentified object. A second rectangular region can be determined based on the second position, and the second rectangular region can be the minimum bounding rectangle of the unidentified object in the second panoramic image.

[0079] The overlap information between the first and second rectangular regions can be the area of ​​the overlapping region (e.g., the number of pixels in the overlapping region). Based on this, the overlapping region between the first and second rectangular regions can be determined, and its area can be determined. If the area of ​​the overlapping region is greater than a preset area threshold (which can be configured empirically), the target object and the unidentified object are determined to be the same object; otherwise, if the area of ​​the overlapping region is not greater than the preset area threshold, the target object and the unidentified object are determined to be different objects.

[0080] See Figure 2 As shown, the diagram illustrates a first rectangular region, a second rectangular region, and the overlapping area between the first and second rectangular regions. Figure 2 As can be seen, based on the known first rectangular region and the second rectangular region, the overlapping area between the first rectangular region and the second rectangular region can be determined, and then the area of ​​the overlapping area between the first rectangular region and the second rectangular region can be determined.

[0081] Of course, in addition to the area of ​​the overlapping region between the first rectangular region and the second rectangular region, the overlap information can also be other features, such as the area of ​​the intersection region between the first rectangular region and the second rectangular region, or the ratio between the areas of the union region between the first rectangular region and the second rectangular region, and there are no restrictions on this.

[0082] If the overlap information is the ratio between the area of ​​the intersection region and the area of ​​the union region, then if the ratio is greater than a preset ratio threshold, the target object and the unidentified object are determined to be the same object; otherwise, if the ratio is not greater than the preset ratio threshold, the target object and the unidentified object are determined to be different objects.

[0083] Of course, the above is just an example of determining whether the target object and the unidentified object are the same object. There are no restrictions on this, as long as it can be determined whether the target object and the unidentified object are the same object.

[0084] Scenario 3: If the identity information of the target object is obtained, and there is no corresponding attendance time for this identity information, and it is determined that the target object and the unidentified object are not the same object based on the first location and the second location corresponding to the unidentified object, it means that the target object was not in the target scene before the first collection time. That is, the target object was in the target scene at the first collection time. Therefore, the attendance time corresponding to the identity information can be determined based on the first collection time corresponding to the target object, which is the attendance time of the target object. The correspondence between the identity information and the attendance time can be recorded. For example, the first collection time can be used as the attendance time of the target object, so that the time when the target object is first detected can be used as the attendance time, so as to obtain an accurate and reliable attendance time, which can accurately reflect the actual attendance time of the target object.

[0085] For example, if the target object does not belong to the same object as each unidentified object, and the target object's identity information does not correspond to an attendance time, it means that the target object is detected for the first time. Therefore, the first collection time corresponding to the target object can be used as the attendance time corresponding to the identity information.

[0086] For example, for each unidentified object, regarding how to determine whether the target object and the unidentified object are the same object, please refer to Case 2, which will not be repeated here.

[0087] Scenario 4: If the target object's identity information is not obtained, meaning the target object is located in the target scene but its identity cannot be identified (e.g., due to factors such as the target object's acquisition angle, accurate and reliable feature information is not obtained, thus failing to identify the target object), then this target object can be designated as an unidentified object, and the first acquisition time and first position corresponding to the target object are recorded. It should be noted that in subsequent processes, the panoramic image acquired by the panoramic camera is referred to as the first panoramic image, while the panoramic image corresponding to this target object will be updated to the second panoramic image. Furthermore, as this target object is considered an unidentified object, its first acquisition time is updated to the second acquisition time, and its first position is updated to the second position. Based on this, scenarios 1, 2, or 3 are repeated; the process will not be elaborated further.

[0088] In summary, without obtaining the target's identity information, it is unnecessary to determine the target's attendance time, and it is impossible to record the correspondence between identity information and attendance time.

[0089] In one possible implementation, if the identity information of the target object is not obtained, the target object is directly identified as an unidentified object.

[0090] In another possible implementation, if the identity information of the target object is not obtained, it can be determined whether the target object and the identified object belong to the same object. The identified object is an object that already has identity information. The attendance time of the identified object has been recorded. The method of obtaining the identity information and attendance time corresponding to the identified object is described in Case 2 and Case 3, which will not be repeated here.

[0091] If the target object and the identified object belong to the same object, it means that the target object was already in the target scene before the first acquisition time. Furthermore, when the target object was in the target scene, its identity information was already identified, and the corresponding attendance time was recorded. Therefore, even if the target object's identity information is not obtained, it does not affect the target object's attendance. Thus, the first panoramic image, the first acquisition time, and the first position corresponding to the target object can be ignored, and the processing flow for that target object can be terminated. Since the first acquisition time and first position corresponding to the target object are not recorded, unnecessary repeated comparisons can be reduced in subsequent comparisons, thus reducing the computational workload.

[0092] If the target object is different from the identified objects (such as all identified objects), it means that the target object was not in the target scene before the first acquisition time, and the attendance time corresponding to the target object has not been recorded. Therefore, the first panoramic image will not be ignored, and the target object needs to be identified as an unidentified object and the first acquisition time and first position corresponding to the target object need to be recorded.

[0093] For example, to determine whether a target object and an identified object are the same object, the same object can be determined based on a first position corresponding to the target object and position A corresponding to the identified object. For instance, a first rectangular region of the target object in the first panoramic image is determined based on the first position, and a rectangular region A of the identified object in the corresponding panoramic image is determined based on position A. Based on the overlap information between the first rectangular region and rectangular region A, it is determined whether the target object and the identified object are the same object. The determination method can be found in Case 2, and will not be repeated here.

[0094] In another possible implementation, if the identity information of the target object is not obtained, it can be determined whether the target object and the identified object belong to the same object. The identified object is an object that already has identity information and the attendance time of the identified object has been recorded.

[0095] If the target object and the already identified object belong to the same object, it means that the target object was already in the target scene before the first acquisition time, its identity information has been identified, and the corresponding attendance time has been recorded. Therefore, the first panoramic image, the first acquisition time and the first position corresponding to the target object can be ignored, and the processing flow for that target object can be terminated. Since the first acquisition time and the first position corresponding to the target object are not recorded, unnecessary repeated comparisons can be reduced in subsequent comparisons, thus reducing the workload of computation.

[0096] If the target object and the identified objects (such as all identified objects) are different objects, it can also be determined whether the target object and the unidentified object belong to the same object. An unidentified object is an object without identity information, and its attendance time is not recorded; however, its second location and second acquisition time have been recorded. If the target object and the unidentified object belong to the same object, it means that the target object was already in the target scene before the first acquisition time. Although the target object's identity information has not yet been identified, and its corresponding attendance time has not been recorded, since the target object is already an unidentified object, the first panoramic image, the first acquisition time, and the first location corresponding to this target object can be ignored, and the processing flow for this target object can be terminated. Since the first acquisition time and first location corresponding to the target object are not recorded, unnecessary repeated comparisons can be reduced in subsequent comparisons, thus reducing computational workload.

[0097] If the target object and the unidentified object (such as all unidentified objects) are different objects, it means that the target object was not in the target scene before the first acquisition time, and the attendance time corresponding to the target object has not been recorded. Therefore, the first panoramic image will not be ignored. The target object needs to be identified as an unidentified object, and the first acquisition time and first position corresponding to the target object need to be recorded.

[0098] For example, in order to determine whether a target object is the same as an identified object (or an unidentified object), the first position corresponding to the target object and the position corresponding to the identified object (or the unidentified object) can be used to determine whether the target object is the same as the identified object (or the unidentified object).

[0099] As can be seen from the above technical solutions, in the embodiments of this application, if it is determined that the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, then the attendance time of the target object is determined based on the second collection time corresponding to the unidentified object, rather than based on the first collection time corresponding to the target object. This results in an accurate and reliable attendance time that can accurately reflect the actual attendance time of the target object. In other words, although the identity information of the target object is not identified when the target object is first detected, the time when the target object is first detected can still be used as the attendance time by identifying the target object and the unidentified object as the same object.

[0100] The above process of the embodiments of this application will be described below in conjunction with specific application scenarios.

[0101] Step S11: Obtain panoramic image a1 of the target scene (such as scene A) captured by the panoramic camera. Assume that panoramic image a1 includes target object a1-1 and target object a1-2. Determine the acquisition time b1 (i.e., the acquisition time of panoramic image a1), position a1-1-1 and feature information a1-1-2 corresponding to target object a1-1. Determine the acquisition time b1, position a1-2-1 and feature information a1-2-2 corresponding to target object a1-2.

[0102] Step S12: If the identity information of the target object a1-1 is determined to be "Li XX" based on the feature information a1-1-2, and the identity information does not correspond to an attendance time, and the target object a1-1 is not the same object as any of the unidentified objects, i.e., corresponding to case 3, then the collection time b1 is taken as the attendance time of the target object a1-1, and the correspondence between the identity information, the attendance time and the location a1-1-1 is recorded, as shown in Table 1.

[0103] If the identity information of the target object a1-2 cannot be determined based on the feature information a1-2-2, and the target object a1-2 is not the same object as any of the unidentified objects, i.e., corresponding to case 4, then the target object a1-2 is regarded as an unidentified object, and the collection time b1 and location a1-2-1 corresponding to this unidentified object are recorded, as shown in Table 1.

[0104] Table 1

[0105] Li XX Data collection time b1 Location a1-1-1 a1-2 Data collection time b1 Location a1-2-1

[0106] Step S13: Acquire panoramic image a2 of the target scene using a panoramic camera. Assume that panoramic image a2 includes target object a2-1, target object a2-2, target object a2-3, and target object a2-4. Determine the acquisition time b2 (i.e., the acquisition time of panoramic image a2), position a2-1-1, and feature information a2-1-2 corresponding to target object a2-1. Determine the acquisition time b2, position a2-2-1, and feature information a2-2-2 corresponding to target object a2-2. Determine the acquisition time b2, position a2-3-1, and feature information a2-3-2 corresponding to target object a2-3. Determine the acquisition time b2, position a2-4-1, and feature information a2-4-2 corresponding to target object a2-4.

[0107] Step S14: If the identity information of target object a2-1 is determined to be "Li XX" based on feature information a2-1-2, as shown in Table 1, this identity information corresponds to an attendance time (i.e., collection time b1), which corresponds to case 1. The attendance time corresponding to this identity information has already been recorded. Therefore, the collection time b2, location a2-1-1, and feature information a2-1-2 corresponding to target object a2-1 are ignored, and the processing flow of target object a2-1 ends.

[0108] If the identity information of target object a2-2 is determined to be "Wang XX" based on feature information a2-2-2, and this identity information does not correspond to an attendance time, and target object a2-2 and unidentified object a1-2 are the same object, i.e., corresponding to case 2, then the collection time b1 corresponding to unidentified object a1-2 (obtained from Table 1) is taken as the attendance time of target object a2-2, instead of taking collection time b2 as the attendance time of target object a2-2, and the correspondence between the identity information, the attendance time and the location a2-2-1 is recorded, as shown in Table 2.

[0109] Determining that target object a2-2 and unidentified object a1-2 are the same object can include: determining that target object a2-2 and unidentified object a1-2 are the same object based on the positions a2-2-1 of target object a2-2 and a1-2-1 of unidentified object a1-2. For example, a rectangular region of target object a2-2 in panoramic image a2 can be determined based on position a2-2-1, and a rectangular region of unidentified object a1-2 in panoramic image a1 can be determined based on position a1-2-1. If the area of ​​the overlapping region between the two rectangular regions is greater than a preset area threshold, then target object a2-2 and unidentified object a1-2 are determined to be the same object.

[0110] In this case, after identifying the target object a2-2 and the unidentified object a1-2 as the same object, the identity information of the target object a2-2 can also be used as the identity information of the unidentified object a1-2. That is, the identity information of the unidentified object a1-2 has been obtained, and the unidentified object a1-2 is no longer considered an unidentified object.

[0111] If the identity information of target object a2-3 is determined to be "Zhuge X" based on feature information a2-3-2, and this identity information does not correspond to an attendance time, and target object a2-3 is not the same object as any of the unidentified objects, i.e., corresponding to case 3, then the collection time b2 is taken as the attendance time of target object a2-3, and the correspondence between the identity information, the attendance time and the location a2-3-1 is recorded, as shown in Table 2.

[0112] If the identity information of the target object a2-4 cannot be determined based on the feature information a2-4-2, and the target object a2-4 is not the same object as any of the unidentified objects, i.e., corresponding to case 4, then the target object a2-4 is regarded as an unidentified object, and the collection time b2 and position a2-4-1 corresponding to this unidentified object are recorded, as shown in Table 2.

[0113] Table 2

[0114]

[0115] Step S15: Acquire panoramic image a3 of the target scene using a panoramic camera. Assume that panoramic image a3 includes target object a3-1, target object a3-2, target object a3-3, and target object a3-4. Determine the acquisition time b3 (i.e., the acquisition time of panoramic image a3), position a3-1-1, and feature information a3-1-2 corresponding to target object a3-1. Determine the acquisition time b3, position a3-2-1, and feature information a3-2-2 corresponding to target object a3-2. Determine the acquisition time b3, position a3-3-1, and feature information a3-3-2 corresponding to target object a3-3. Determine the acquisition time b3, position a3-4-1, and feature information a3-4-2 corresponding to target object a3-4.

[0116] Step S16: If the identity information of target object a3-1 is not determined based on feature information a3-1-2, and it is determined based on location a3-1-1 and location a1-1-1 that target object a3-1 and target object a1-1 (i.e., the identified object) belong to the same object, as shown in Table 2, since the attendance time (i.e., collection time b1) corresponding to the identity information of target object a1-1 has been recorded, the collection time b3, location a3-1-1, and feature information a3-1-2 corresponding to target object a3-1 are ignored, and the processing flow of target object a3-1 ends.

[0117] If the identity information of target object a3-2 is determined to be "Wang XX" based on feature information a3-2-2, as shown in Table 2, this identity information corresponds to attendance time. Therefore, the collection time b3, location a3-2-1, and feature information a3-2-2 corresponding to target object a3-2 are ignored, and the processing flow of target object a3-2 ends.

[0118] If the identity information of target object a3-3 is determined to be "Zhuge X" based on feature information a3-3-2, as shown in Table 2, this identity information corresponds to attendance time. Therefore, the collection time b3, location a3-3-1, and feature information a3-3-2 corresponding to target object a3-3 are ignored, and the processing flow of target object a3-3 ends.

[0119] If the identity information of target object a3-4 cannot be determined based on feature information a3-4-2, then it is determined whether target object a3-4 is the same object as an already identified object. If target object a3-4 is the same object as an already identified object, then the acquisition time b3, location a3-4-1, and feature information a3-4-2 corresponding to target object a3-4 are ignored, and the processing flow for target object a3-4 ends. If target object a3-4 is not the same object as any of the already identified objects, then it is determined whether target object a3-4 is the same object as an unidentified object.

[0120] Referring to Table 2, if the target object a3-4 and the unidentified object a2-4 are the same object, that is, the target object a3-4 was already an unidentified object between the acquisition time b3 and the acquisition time b3, the acquisition time b3, the location a3-4-1, and the feature information a3-4-2 corresponding to the target object a3-4 are ignored, and the processing flow of the target object a3-4 ends.

[0121] If the target object a3-4 is not the same as any of the unidentified objects, then the target object a3-4 is treated as an unidentified object, and the acquisition time b3 and position a3-4-1 corresponding to this unidentified object are recorded.

[0122] Step S17: Repeat the above process, that is, use the earliest collection time as the attendance time of the target object.

[0123] See Figure 3 The diagram shown illustrates a method for determining attendance time, which may include:

[0124] Step 301: Acquire PTZ images of the target scene captured by the PTZ camera, and determine the third acquisition time and third feature information corresponding to the candidate objects in the PTZ images. For ease of distinction, the acquisition time corresponding to the candidate object is recorded as the third acquisition time, and the feature information corresponding to the candidate object is recorded as the third feature information.

[0125] For example, a PTZ camera can periodically acquire PTZ images of the target scene. Since the processing method for each PTZ image is the same, the processing procedure for one PTZ image will be used as an example. If the PTZ image does not contain any object, it is discarded. If the PTZ image contains at least one object, for ease of distinction, each object in the PTZ image is called a candidate object. Since the processing method for each candidate object is the same, the processing procedure for one candidate object will be used as an example. For each PTZ image, the third acquisition time and third feature information corresponding to the candidate object in that PTZ image can be determined.

[0126] In one possible implementation, for a candidate object in a PTZ image, a third acquisition time and third feature information corresponding to the candidate object can be determined. The third acquisition time corresponding to the candidate object can be the acquisition time of the PTZ image; for example, if the PTZ image is acquired at time A, the third acquisition time corresponding to the candidate object is time A. The third feature information corresponding to the candidate object can be the facial features of the candidate object or the human body features of the candidate object; there is no limitation in this regard.

[0127] In one possible implementation, after obtaining the third acquisition time and third feature information corresponding to the candidate object, it can be determined whether the acquisition time and feature information corresponding to the candidate object have already been stored. If yes, the third acquisition time and third feature information corresponding to the candidate object can be ignored, that is, the third acquisition time and third feature information corresponding to the candidate object will no longer be stored. If no, the third acquisition time and third feature information corresponding to the candidate object can be stored, as shown in Table 3.

[0128] Table 3

[0129] Candidate object m1 Data collection time m1-1 Feature information m1-2 Candidate object m2 Data collection time m2-1 Feature information m2-2 Candidate object m3 Data collection time m3-1 Feature information m3-2 … … …

[0130] For example, determining whether the acquisition time and feature information corresponding to the candidate object (taking candidate object m3 as an example) have been stored may include, but is not limited to: sequentially traversing each candidate object in Table 3 (taking candidate object m1 as an example), determining the similarity between the feature information m3-2 of candidate object m3 and the feature information m1-2 of candidate object m1. If both feature information m3-2 and feature information m1-2 are facial features, then the similarity is the similarity between facial features. If the similarity is greater than a preset similarity threshold (configured based on experience, such as 90% or 95%), then candidate object m3 and candidate object m1 are determined to be the same object, that is, the acquisition time and feature information corresponding to candidate object m3 have been stored, and the third acquisition time m3-1 and the third feature information m3-2 corresponding to candidate object m3 are no longer stored. If the similarity is not greater than the preset similarity threshold, then candidate object m3 and candidate object m1 are determined to be different objects. If candidate object m3 is not the same as any of the candidate objects in Table 3, that is, the acquisition time and feature information corresponding to candidate object m3 are not stored, then the third acquisition time and third feature information corresponding to candidate object m3 are stored, as shown in Table 3.

[0131] Step 302: Obtain the first panoramic image of the target scene captured by the panoramic camera, and determine the first acquisition time, first position and first feature information of the target object in the first panoramic image.

[0132] Step 303: If the identity information of the target object is determined based on the first feature information, and this identity information does not correspond to an attendance time, then determine whether the target object and the unidentified object are the same object based on the first location and the second location corresponding to the unidentified object. If not, proceed to step 304; if yes, proceed to step 305.

[0133] Step 304: If, based on the first and third feature information, it is determined that the target object and the candidate object are the same object, then the attendance time of this identity information is determined based on the first and third collection times. If, based on the first and third feature information, it is determined that the target object and all candidate objects are not the same object, then the attendance time of this identity information is determined based on the first collection time.

[0134] Step 305: If, based on the first and third feature information, it is determined that the target object and the candidate object are the same object, then the attendance time of this identity information is determined based on the second and third collection times corresponding to the unidentified object. If, based on the first and third feature information, it is determined that the target object and all candidate objects are not the same object, then the attendance time of this identity information is determined based on the second collection time.

[0135] In one possible implementation, after determining the first acquisition time, first position, and first feature information corresponding to the target object in the first panoramic image, the processing can be divided into the following cases:

[0136] Case 1: The identity information of the target object is determined based on the first feature information, and the identity information already corresponds to the attendance time, that is, the attendance time of the identity information has been obtained. Therefore, the first panoramic image is ignored, and the first acquisition time and the first feature information corresponding to the target object are ignored.

[0137] Scenario 2: The identity information of the target object is determined based on the first feature information, and this identity information does not correspond to an attendance time. Furthermore, based on the first location and the second location corresponding to the unidentified object, it is determined that the target object and the unidentified object are the same object. Therefore, the attendance time corresponding to this identity information can be determined based on the second collection time corresponding to the unidentified object. For example, when determining the attendance time corresponding to this identity information based on the second collection time, each candidate object in Table 3 (taking candidate object m1 as an example) is sequentially traversed. The similarity between the first feature information of the target object and the third feature information of candidate object m1 is determined. If the first feature information includes the facial features corresponding to the target object, and the third feature information includes the facial features corresponding to candidate object m1, then the similarity is the similarity between the facial features corresponding to the target object and the facial features corresponding to candidate object m1. If the similarity is greater than a preset similarity threshold (which can be configured empirically, such as 90% or 95%), then the target object and candidate object m1 are determined to be the same object. If the similarity is not greater than the preset similarity threshold, then the target object and candidate object m1 are determined to be different objects.

[0138] If the target object and a candidate object (e.g., candidate object m2) are the same object, the attendance time for that identity information is determined based on the second collection time and the third collection time corresponding to candidate object m2. For example, if the second collection time is earlier than or equal to the third collection time, the second collection time can be determined as the attendance time for that identity information. If the second collection time is later than the third collection time, the third collection time can be determined as the attendance time for that identity information.

[0139] If the target object is not the same as any of the candidate objects, the attendance time of the identity information is determined based on the second collection time, such as determining the second collection time as the attendance time of the identity information.

[0140] Scenario 3: The identity information of the target object is determined based on the first feature information, and this identity information does not correspond to an attendance time. Furthermore, based on the first location and the second location corresponding to the unidentified object, it is determined that the target object and the unidentified object are not the same object. Therefore, the attendance time corresponding to this identity information can be determined based on the first collection time corresponding to the target object. For example, when determining the attendance time corresponding to this identity information based on the first collection time, each candidate object in Table 3 (taking candidate object m1 as an example) is traversed sequentially to determine the similarity between the first feature information of the target object and the third feature information of candidate object m1. If the similarity is greater than a preset similarity threshold, the target object and candidate object m1 are determined to be the same object; if the similarity is not greater than the preset similarity threshold, the target object and candidate object m1 are determined to be different objects.

[0141] If the target object and a candidate object are the same object, the attendance time for that identity information is determined based on the first collection time and the third collection time of the candidate object. For example, if the first collection time is earlier than or equal to the third collection time, the first collection time is determined as the attendance time for that identity information. If the first collection time is later than the third collection time, the third collection time is determined as the attendance time for that identity information. If the target object and all candidate objects are not the same object, the attendance time for that identity information is determined based on the first collection time, such as using the first collection time as the attendance time for that identity information.

[0142] Case 4: If the identity information of the target object cannot be determined based on the first feature information, the target object can be identified as an unidentified object, and the first acquisition time and first location corresponding to the target object can be recorded.

[0143] For example, cases 1-4 can be referred to in the above embodiments, and will not be repeated here.

[0144] As can be seen from the above technical solutions, in this embodiment of the application, the attendance time of identity information can be determined based on the PTZ image captured by the PTZ camera and the panoramic image captured by the panoramic camera, thereby obtaining an accurate and reliable attendance time that can accurately reflect the actual attendance time of the target object.

[0145] In one possible implementation, see Figure 4 As shown, attendance time can be determined using the following steps:

[0146] Step 401A: First recognition by the panoramic camera.

[0147] Step 401B: When the panoramic camera first identifies the person, obtain the person's coordinate information and the capture time; where the person's coordinate information is the first position of the target object, and the capture time is the first acquisition moment.

[0148] Step 401C: Determine whether facial feature information has been obtained. If facial feature information has been obtained, i.e., the identity information of the target object exists, then proceed to step 401D; otherwise, proceed to step 401E.

[0149] Step 401D: Store the record containing personnel information, coordinates, and time.

[0150] Step 401E: Store records that do not contain personnel information but contain coordinates and time.

[0151] Step 402A: Second recognition by panoramic camera.

[0152] Step 402B: During the second recognition by the panoramic camera, obtain the personnel coordinates and capture time.

[0153] Step 402C: Determine whether facial feature information has been obtained. If facial feature information has been obtained, i.e., the identity information of the target object exists, then execute steps 402D and 401D; otherwise, execute step 401E.

[0154] Step 402D: Calculate the overlap of human body coordinates between the first and second snapshots.

[0155] Step 402E: Determine whether the overlap meets the threshold.

[0156] If yes, proceed to step 401D; otherwise, proceed to step 401E.

[0157] The third recognition process of the panoramic camera can be found in the second recognition process of the panoramic camera, and will not be repeated here.

[0158] Step 403A: The PTZ camera captures facial feature information.

[0159] Step 403B: Store face coordinate information.

[0160] Step 403C: Determine if a match is found with the same person. If so, proceed to step 403D.

[0161] Step 403D: Determine whether the PTZ data time is earlier than the panoramic data time.

[0162] Step 403E: If yes, update the time corresponding to the personnel information.

[0163] Based on the same concept as the above method, this application proposes an attendance time determination device, see [link to relevant documentation]. Figure 5 The diagram shown is a structural schematic of the device, which may include:

[0164] The acquisition module 51 is used to acquire a first candidate image of a target scene captured by a first type of camera, determine a first acquisition time and a first position corresponding to a target object in the first candidate image; and acquire an identity recognition result corresponding to the target object. If the identity recognition result indicates that there is identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then the module determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object; wherein, the second acquisition time of the second candidate image corresponding to the unidentified object is earlier than the first acquisition time, and the unidentified object does not correspond to identity information.

[0165] The determination module 52 is used to determine the attendance time of the identity information based on the second collection time if the target object and the unidentified object are the same object.

[0166] The determining module 52 is further configured to determine the attendance time of the identity information based on the first collection time if the target object and the unidentified object are different objects, and update the target object to an identified object, wherein the identified object is an object with identity information.

[0167] For example, the determining module 52 is further configured to ignore the first collection time and the first location corresponding to the target object if the identity recognition result indicates that there is identity information corresponding to the target object and the identity information corresponds to an attendance time.

[0168] For example, when the acquisition module 51 determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, it is specifically used to: determine the first region of the target object in the first candidate image based on the first position; determine the second region of the unidentified object in the second candidate image based on the second position; and determine whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region.

[0169] For example, when the acquisition module 51 determines whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region, it is specifically used to determine the overlapping area between the first region and the second region; if the area of ​​the overlapping area is greater than a preset area threshold, then the target object and the unidentified object are determined to be the same object; otherwise, the target object and the unidentified object are determined to be different objects.

[0170] For example, the determining module 52 is further configured to: if the identity recognition result indicates that there is no identity information corresponding to the target object, and the target object and the unidentified object are determined to be the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or the target object and the identified object are determined to be the same object based on the first position corresponding to the target object and the position corresponding to the identified object, then ignore the first collection time and the first position corresponding to the target object; if the target object and the unidentified object are different objects, and the target object and the identified object are different objects, then determine the target object as an unidentified object, and record the first collection time and the first position corresponding to the target object; the determining module is further configured to: after determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, if so, update the unidentified object to an identified object, and ignore the first collection time and the first position corresponding to the target object.

[0171] For example, the first type of camera includes a panoramic camera. The acquisition module 51 is further configured to acquire a PTZ image of the target scene captured by a panning, tilting, zooming PTZ camera, and determine a third acquisition time corresponding to a candidate object in the PTZ image. When the determining module 52 determines the attendance time of the identity information based on the first acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, if the first acquisition time is earlier than the third acquisition time, determine the first acquisition time as the attendance time; if the first acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time. When the determining module 52 determines the attendance time of the identity information based on the second acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, if the second acquisition time is earlier than the third acquisition time, determine the second acquisition time as the attendance time; if the second acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time.

[0172] For example, when the determining module 52 determines that the candidate object and the target object are the same object, it is specifically used to: obtain the similarity between the facial features corresponding to the candidate object and the facial features corresponding to the target object; if the similarity is greater than a preset similarity threshold, then the candidate object and the target object are determined to be the same object; wherein the facial features corresponding to the candidate object are determined based on the PTZ image, and the facial features corresponding to the target object are determined based on the first candidate image.

[0173] Based on the same concept as the above method, this application proposes an electronic device, see [link to previous application]. Figure 6 As shown, the electronic device includes a processor 61 and a machine-readable storage medium 62, the machine-readable storage medium 62 storing machine-executable instructions that can be executed by the processor 61; the processor 61 is used to execute the machine-executable instructions to implement the attendance time determination method disclosed in the above example of this application.

[0174] Based on the same application concept as the above method, this application embodiment also provides a machine-readable storage medium storing a plurality of computer instructions, which, when executed by a processor, can implement the attendance time determination method disclosed in the above examples of this application.

[0175] The aforementioned machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device that can contain or store information, such as executable instructions, data, etc. For example, machine-readable storage media can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or combinations thereof.

[0176] Based on the same concept as the above method, this application also provides a computer program stored in a machine-readable storage medium. When the processor executes the computer program, it causes the processor to implement the attendance time determination method disclosed in the above examples of this application.

[0177] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which can take the form of a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email sending and receiving device, game console, tablet computer, wearable device, or any combination of these devices.

[0178] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0179] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0180] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0181] Furthermore, these computer program instructions can also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in the process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0182] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0183] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for determining attendance time, characterized in that, The method for determining the attendance time of objects within a target scenario includes: Acquire a first candidate image of the target scene captured by a first type of camera, and determine the first acquisition time and first position of the target object in the first candidate image; Obtain the identity recognition result corresponding to the target object. If the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then determine whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object. The target object is located at a fixed position in the target scene. A first region of the target object in a first candidate image is determined based on the first position, and a second region of the unidentified object in a second candidate image is determined based on the second position. Based on the overlap information between the first and second regions, determine whether the target object and the unidentified object are the same object. The second acquisition time of the second candidate image corresponding to the unidentified object is earlier than the first acquisition time, and the unidentified object does not correspond to identity information. If so, then determine the attendance time of the identity information based on the second acquisition time, and update the unidentified object to an identified object.

2. The method according to claim 1, characterized in that, After determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, the method further includes: If not, then the attendance time of the identity information is determined based on the first collection time, and the target object is updated to an identified object, which is an object with identity information; Before determining whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, the method further includes: If the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information corresponds to an attendance time, then the first collection time and the first location corresponding to the target object are ignored.

3. The method according to claim 1, characterized in that, The step of determining whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region includes: Determine the overlapping area between the first region and the second region; If the area of ​​the overlapping region is greater than a preset area threshold, then the target object and the unidentified object are determined to be the same object; otherwise, the target object and the unidentified object are determined to be different objects.

4. The method according to claim 1 or 2, characterized in that, After obtaining the identity recognition result corresponding to the target object, if the identity recognition result indicates that there is no identity information corresponding to the target object, the method further includes: if it is determined that the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or it is determined that the target object and the identified object are the same object based on the first position corresponding to the target object and the position corresponding to the identified object, then the first acquisition time and the first position corresponding to the target object are ignored.

5. The method according to claim 2, characterized in that, The first type of camera includes a panoramic camera, and the method further includes: acquiring a PTZ image of the target scene captured by a translation tilt zoom PTZ camera, and determining a third acquisition time corresponding to a candidate object in the PTZ image; The step of determining the attendance time based on the first collection time includes: if the candidate object and the target object are determined to be the same object, if the first collection time is earlier than the third collection time, the first collection time is determined as the attendance time; if the first collection time is later than the third collection time, the third collection time is determined as the attendance time. The method of determining the attendance time of the identity information based on the second collection time includes: if the candidate object and the target object are determined to be the same object, if the second collection time is earlier than the third collection time, the second collection time is determined as the attendance time; if the second collection time is later than the third collection time, the third collection time is determined as the attendance time.

6. The method according to claim 5, characterized in that, Determining that the candidate object and the target object are the same object includes: The similarity between the facial features corresponding to the candidate object and the facial features corresponding to the target object is obtained. If the similarity is greater than a preset similarity threshold, the candidate object and the target object are determined to be the same object. The facial features corresponding to the candidate object are determined based on the PTZ image, and the facial features corresponding to the target object are determined based on the first candidate image.

7. The method according to claim 1, characterized in that, If the identity recognition result indicates that there is no identity information corresponding to the target object, and the target object is different from the unidentified object and different from the identified object, then the target object is determined as an unidentified object, and the first collection time and the first position corresponding to the target object are recorded.

8. An attendance time determination device, characterized in that, The device for determining the attendance time of objects within a target scene includes: The acquisition module is used to acquire a first candidate image of a target scene captured by a first type of camera, determine a first acquisition time and a first position corresponding to a target object in the first candidate image; and acquire an identity recognition result corresponding to the target object. If the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information does not correspond to an attendance time, then the module determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object. Specifically, when the target object is in a fixed position in the target scene, the acquisition module determines whether the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object: based on the first position, determine a first region of the target object in the first candidate image; based on the second position, determine a second region of the unidentified object in the second candidate image; and based on the overlap information between the first region and the second region, determine whether the target object and the unidentified object are the same object. The second acquisition time of the second candidate image corresponding to the unidentified object is earlier than the first acquisition time, and the unidentified object does not correspond to identity information. The determination module is used to determine the attendance time of the identity information based on the second collection time if the target object and the unidentified object are the same object, and update the unidentified object to an identified object.

9. The apparatus according to claim 8, Its features are, in, The determining module is further configured to, if the target object and the unidentified object are different objects, determine the attendance time of the identity information based on the first collection time, and update the target object to an identified object, wherein the identified object is an object with identity information; the determining module is further configured to, if the identity recognition result indicates the existence of identity information corresponding to the target object, and the identity information corresponds to an attendance time, ignore the first collection time and the first location corresponding to the target object; Specifically, when the acquisition module determines whether the target object and the unidentified object are the same object based on the overlap information between the first region and the second region, it is used to determine the overlapping area between the first region and the second region; if the area of ​​the overlapping area is greater than a preset area threshold, then the target object and the unidentified object are determined to be the same object; otherwise, the target object and the unidentified object are determined to be different objects. The determining module is further configured to ignore the first collection time and first position corresponding to the target object if the identity recognition result indicates that there is no identity information corresponding to the target object, and the target object and the unidentified object are the same object based on the first position corresponding to the target object and the second position corresponding to the unidentified object, or the target object and the identified object are the same object based on the first position corresponding to the target object and the position corresponding to the identified object. The first type of camera includes a panoramic camera. The acquisition module is further configured to acquire a PTZ image of the target scene captured by a panning, tilting, zooming PTZ camera, and determine a third acquisition time corresponding to a candidate object in the PTZ image. When the determining module determines the attendance time of the identity information based on the first acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, and the first acquisition time is earlier than the third acquisition time, determine the first acquisition time as the attendance time; if the first acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time. When the determining module determines the attendance time of the identity information based on the second acquisition time, it is specifically configured to: if the candidate object and the target object are determined to be the same object, and the second acquisition time is earlier than the third acquisition time, determine the second acquisition time as the attendance time; if the second acquisition time is later than the third acquisition time, determine the third acquisition time as the attendance time. Specifically, when the determining module determines that the candidate object and the target object are the same object, it is used to: obtain the similarity between the facial features corresponding to the candidate object and the facial features corresponding to the target object; if the similarity is greater than a preset similarity threshold, then the candidate object and the target object are determined to be the same object; wherein the facial features corresponding to the candidate object are determined based on the PTZ image, and the facial features corresponding to the target object are determined based on the first candidate image; The determining module is further configured to determine the target object as an unidentified object if the identity recognition result indicates that there is no identity information corresponding to the target object, and the target object is different from the unidentified object and the target object is different from the identified object, and the first collection time and the first position corresponding to the target object are recorded.

10. An electronic device, characterized in that, include: A processor and a machine-readable storage medium, the machine-readable storage medium storing machine-executable instructions that can be executed by the processor; The processor is configured to execute machine-executable instructions to implement the steps of the method according to any one of claims 1-7.

Images