Object behavior determination method, device, electronic device and storage medium

Abstract

The present invention discloses a method, device, electronic device, and storage medium for determining object behavior, wherein the object behavior determination method includes: when an elevator car arrives at an elevator lobby and opens the elevator door, controlling a sensor to collect multiple frames of images of the area of ​​the elevator lobby; performing object detection and tracking on the images to obtain at least two pieces of status data for each object in the elevator lobby; calculating a behavior score for the object based on a preset weight coefficient for each piece of status data and the status data; and determining that the object's behavioral intention is to take the elevator when the behavior score is greater than a preset score threshold. This embodiment solves the problem of proneness to misjudgment of the object's behavioral intention when determining the object's behavioral intention based on the speed of movement. The object's behavioral intention is determined to be highly accurate by integrating multiple pieces of status data of the object, thereby improving the accuracy of identifying objects outside the elevator door with the intention of taking the elevator, which is conducive to accurately controlling the closing time of the elevator door and improving the operating efficiency of the elevator.

Description

Technical Field

[0001] The present invention relates to the field of elevator control technology, and in particular to a method, device, electronic device and storage medium for determining object behavior. Background Art

[0002] With the demand for intelligent elevator development, technologies have emerged that use sensors to sense the movement trends of passengers in the elevator lobby and control the closing or opening of the elevator doors based on the passengers' movement trends.

[0003] In the prior art, after a sensor collects multiple frames of images of the elevator lobby, the multiple frames of images are used to analyze the moving speed of people outside the elevator door, and the moving speed is used to predict whether the people have the intention to take the elevator, so that the elevator door is controlled according to the intention of the people outside the elevator door. However, the area outside the elevator door is usually a public area, and the people in the area outside the elevator door may be waiting for the elevator or passing by the elevator door. Presetting the intention of the people based only on the moving speed or position may easily lead to misjudgment of the intention of the people. For example, people who pass by the elevator door without the need to take the elevator are identified as people who have the intention to take the elevator, resulting in inaccurate prediction of the people's intention, which is not conducive to the control of the elevator door. For example, people who pass by the elevator door without the need to take the elevator are identified as people who have the intention to take the elevator, which causes the elevator door to close late, reducing the operating efficiency of the elevator door. Summary of the Invention

[0004] The present invention provides a method, device, electronic device and storage medium for determining object behavior to solve the problem that predicting the intention of people outside the elevator door by moving speed leads to inaccurate prediction of people's intentions and is not conducive to elevator door control.

[0005] In a first aspect, the present invention provides a method for determining object behavior, comprising:

[0006] When the elevator car arrives at the elevator lobby and the elevator door opens, the control sensor collects multiple frames of images of the area in the elevator lobby;

[0007] Performing object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby;

[0008] Calculating a behavior score of the subject based on a preset weight coefficient of each status data and the status data;

[0009] When the behavior score is greater than a preset score threshold, it is determined that the behavior intention of the subject is to take an elevator.

[0010] In a second aspect, the present invention provides an object behavior determination device, comprising:

[0011] An image acquisition module is used to control the sensor to capture multiple frames of images of the elevator lobby area when the elevator car arrives at the elevator lobby and the elevator door opens;

[0012] an object detection and tracking module, configured to perform object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby;

[0013] A behavior score acquisition module, configured to calculate the behavior score of the object based on a preset weight coefficient of each status data and the status data;

[0014] The behavior intention determination module is used to determine that the behavior intention of the object is to take the elevator when the behavior score is greater than a preset score threshold.

[0015] In a third aspect, the present invention provides an electronic device, comprising:

[0016] at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the object behavior determination method described in the first aspect of the present invention.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a processor to implement the object behavior determination method described in the first aspect of the present invention when executed.

[0020] According to an embodiment of the present invention, when the elevator car arrives at the elevator lobby and the elevator door is opened, the control sensor collects multiple frames of images of the area of ​​the elevator lobby, performs object detection and tracking on the images to obtain at least two status data of each object in the elevator lobby, calculates the behavior score of the object based on the preset weight coefficient and status data of each status data, and determines that the object's behavioral intention is to take the elevator when the behavior score is greater than the preset score threshold. This realizes the determination of the object's behavior by obtaining at least two status data of each object through images. For example, the object's position, direction, moving speed and other data can be obtained and combined with the preset weight coefficient to calculate the object's behavior score. When the behavior score is greater than the preset score threshold, it is determined that the object's behavioral intention is to take the elevator, which solves the problem that the object's behavioral intention is prone to misjudgment of behavioral intention when determining the object's behavioral intention by the speed of movement. The accuracy of determining the object's behavioral intention by integrating at least two status data of the object is high, which improves the accuracy of identifying objects outside the elevator door who intend to take the elevator, and is conducive to accurately controlling the closing time of the elevator door and improving the operating efficiency of the elevator.

[0021] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a schematic diagram of the sensor installation and detection area;

[0024] Figure 2 This is a flow chart of a method for determining object behavior provided by the first embodiment of the present invention;

[0025] Figure 3A This is a flow chart of a method for determining object behavior provided by Embodiment 2 of the present invention;

[0026] Figure 3B Schematic diagram of the detection area division;

[0027] Figure 3C A flowchart for controlling the elevator door during the process of the target object entering the elevator car from the elevator lobby;

[0028] Figure 3D A flow chart for adjusting the timer and controlling the elevator door according to the timing of the timer;

[0029] Figure 4 This is a schematic diagram of the structure of an object behavior determination device provided by Embodiment 3 of the present invention;

[0030] Figure 5 This is a structural diagram of an electronic device provided in Example 4 of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0032] Example 1

[0033] Figure 2This is a flow chart of a method for determining object behavior provided in the first embodiment of the present invention. This embodiment is applicable to the case where the behavior intention of an object in the elevator lobby is determined after the elevator car arrives at the elevator lobby and stops and opens the door. The method can be executed by an object behavior determination device, which can be implemented in the form of hardware and / or software. The object behavior determination device can be configured in an electronic device, such as in the controller of an elevator. Figure 2 As shown, the object behavior determination method includes:

[0034] S201: When the elevator car arrives at the elevator lobby and opens the elevator door, control the sensor to collect multiple frames of images of the area of ​​the elevator lobby.

[0035] like Figure 1 As shown, in this embodiment, a sensor 3 is installed on the elevator car. The sensor can be an ordinary sensor, such as a black and white camera, an RGB camera, etc., or an active light sensor, such as a TOF (Time of Flight) depth sensor, a structured light sensor, etc., that is, a sensor that senses by emitting and receiving light to obtain a depth image. Of course, the sensor can also be a radar, etc. This embodiment does not limit the type of sensor.

[0036] In one example, the sensor of this embodiment can be an active light sensor, and the sensor 3 can be installed on the car 1 so that the detection area of ​​the sensor 3 covers the area inside the car 1 and the area of ​​the elevator lobby 2. Optionally, the sensor 3 can be installed on the lintel of the car door of the car 1, for example, in the middle position of the lintel of the car door, so that the detection area of ​​the sensor 3 covers the area inside the car 1 and the area of ​​the elevator lobby 2.

[0037] In one example, the number of sensor 3 can be one, that is, sensor 3 can be an active light sensor with a larger field of view. Furthermore, when the number of sensor 3 is one, the angle of sensor 3 can be adjusted or fixed, or the angle between the light emission axis of sensor 3 and the vertical direction is variable, that is, the angle of sensor 3 can be adjusted so that the detection area 5 of sensor 3 in the elevator lobby 2 can be expanded or reduced.

[0038] like Figure 1As shown, in another example, the number of sensors 3 can be two. Among the two sensors 3, the light emission axis of one sensor 3 is directed toward the interior of the car 1 to capture images of the area within the car 1, and the light emission axis of the other sensor 3 is directed toward the elevator lobby 2 to capture images of the area of ​​the elevator lobby 2. Optionally, the detection ranges of the two sensors 3 have an overlapping area to ensure that images of all areas from the car 1 to the elevator lobby 2 can be captured, while avoiding the overexposure problem caused by the elevator door closing when the sensor is an active light sensor with a vertical light emission axis and simultaneously captures images of all areas from the car 1 to the elevator lobby 2. For example, the angles of the light emission axes of the two sensors 3 with the vertical direction can be adjusted to adjust the detection areas of the two sensors 3. Of course, the angles of the light emission axes of the two sensors 3 with the vertical direction can also be fixed.

[0039] In actual applications, the elevator car stops running when it reaches the elevator lobby of the target floor during the upward or downward movement, and the elevator door is controlled to open so that after the passengers in the car get out of the elevator, objects in the elevator lobby who need to take the elevator can enter the car. The objects can be people, robots, pets, and other objects that need to take the elevator.

[0040] When the elevator door is opened, or after the elevator door is fully opened, the sensor 3 can collect the depth image according to the preset frame rate and send the depth image to the controller of the elevator. Figure 1 As shown, when the number of sensors 3 is one, the sensor 3 can be controlled to face the area of ​​the elevator lobby 2 when the elevator door starts to open, so as to collect multiple frames of images. When the number of sensors 3 is two, the sensor facing the elevator lobby 2 can be controlled to collect multiple frames of images of the area of ​​the elevator lobby 2. It should be noted that the image collection of sensor 3 can start from the opening of the elevator door and end when the elevator door is closed.

[0041] S202: Perform object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby.

[0042] In this embodiment, the object can be a person, a pet, a robot, etc., and an object detection and tracking model can be pre-trained. After the image sequence is input, the object detection and tracking model can identify the object in each frame of the image sequence and predict the position, movement speed, movement direction, contour volume size, movement trajectory and other status data of each object. The training method of the object detection and tracking model can refer to the training method of the existing target detection and tracking model, which will not be described in detail here.

[0043] After acquiring the image, this embodiment can generate an image sequence according to the image acquisition time, and input the image sequence into a pre-trained object detection and tracking model to identify each object in the elevator lobby through the object detection and tracking model, as well as each object's position, movement speed, contour volume, orientation and other status data in the elevator lobby.

[0044] Among them, the position can be the distance from the object to the elevator door in the elevator lobby with the elevator door as the origin, the moving speed can be the moving speed of the object in the elevator lobby, for example, taking the object as a person as an example, after the elevator door is opened, the speed at which the person walks from the elevator lobby to the elevator door, the contour volume can be the outer contour volume of the object recognized in each frame image, for example, when a person is still and has no intention of taking the elevator, his feet will not move and his hands will not swing, and the outer contour volume is small. When a person walks towards the elevator door with the intention of taking the elevator, his feet will move and his hands will swing, and the outer contour volume is large. The orientation can be the orientation of the recognized object, and the orientation usually indicates the moving direction of the object. Taking the object as a person as an example, the orientation can be the orientation of the person's face. In this embodiment, the orientation is a positive angle towards the elevator door, and the angle with the plane where the elevator door is located is the size of the orientation.

[0045] S203: Calculate the behavior score of the object according to the preset weight coefficient of each status data and the status data.

[0046] In this embodiment, a weight coefficient table can be pre-set for each status data. In the weight coefficient table, the corresponding weight coefficient can be found according to the value of each status data. Of course, the weight coefficient of each status data can also be calculated according to a predetermined rule. For example, a weight coefficient calculation formula for each status data can be set, and the value of each status data is input into the formula to obtain the weight coefficient of each status data. Of course, a fixed weight coefficient can also be set for each status data.

[0047] After obtaining the weight coefficient of each status data, the weighted sum can be calculated using multiple status data and corresponding weight coefficients, that is, the behavior score of the object can be obtained. The behavior score indicates the possibility of the object taking the elevator. The larger the behavior score, the greater the possibility of the object taking the elevator.

[0048] S204: When the behavior score is greater than a preset score threshold, determine that the subject's behavior intention is to take an elevator.

[0049] In this embodiment, a score threshold can be set. When the behavior score of the object is greater than the preset score threshold, it is determined that the object is likely to take the elevator. The object's behavioral intention can be determined to be to take the elevator, so that the elevator door can be controlled based on whether the behavioral intention of the objects in the elevator lobby is to take the elevator. For example, after determining the behavioral intention of all objects in the elevator lobby, the objects with the behavioral intention of taking the elevator can be tracked using images collected by the sensor. After the objects with the behavioral intention of taking the elevator have all been tracked to have entered the elevator car, if the elevator door closing time has not yet arrived, the elevator door can be controlled to close in advance, eliminating the need for manual door closing in the car, or closing the elevator door without waiting for the door closing time to arrive, thereby improving the operating efficiency of the elevator and the passenger's elevator experience. Alternatively, when the elevator door closing time arrives, if the objects with the behavioral intention of taking the elevator have not yet been tracked to have entered the car, the elevator door closing is delayed, and the elevator door is closed only after all objects with the intention of taking the elevator have entered the car. This avoids the situation where the elevator door closing prevents objects with the intention of taking the elevator from entering the car in time, or even causes the elevator door to pinch objects entering the car during the closing process.

[0050] According to an embodiment of the present invention, when the elevator car arrives at the elevator lobby and the elevator door is opened, the control sensor collects multiple frames of images of the area of ​​the elevator lobby, performs object detection and tracking on the images to obtain at least two status data of each object in the elevator lobby, calculates the behavior score of the object based on the preset weight coefficient and status data of each status data, and determines that the object's behavioral intention is to take the elevator when the behavior score is greater than the preset score threshold. This realizes the determination of the object's behavior by obtaining at least two status data of each object through the image. For example, the object's position, direction, moving speed and other data can be obtained and combined with the preset weight coefficient to calculate the object's behavior score. When the behavior score is greater than the preset score threshold, it is determined that the object's behavioral intention is to take the elevator, which solves the problem that the object's behavioral intention is prone to misjudgment of behavioral intention when determining the object's behavioral intention by the speed of movement. The accuracy of determining the object's behavioral intention by integrating multiple status data of the object is high, which improves the accuracy of identifying objects outside the elevator door who intend to take the elevator, which is conducive to accurately controlling the closing time of the elevator door and improving the operating efficiency of the elevator.

[0051] Example 2

[0052] Figure 3A This is a flow chart of a method for determining object behavior provided in the second embodiment of the present invention. This embodiment of the present invention is optimized based on the above-mentioned first embodiment. Figure 3A As shown, the object behavior determination method includes:

[0053] S301. When the elevator car arrives at the elevator lobby, the door opening distance of the elevator door is detected.

[0054] The elevator car stops running when it reaches the elevator lobby of the target floor during the upward or downward movement, and the elevator door is controlled to open so that after the passengers in the car get out of the elevator, the people in the elevator lobby who need to take the elevator can enter the car. The door opening distance can be detected. The door opening distance refers to the opening degree of the door when it is open. For example, when the elevator door is a center-opening door, the door opening distance is the distance between the edges of the two doors close to each other. When the elevator door is a side-opening door, the door opening distance is the distance between the edge of the door and the door frame.

[0055] In one example, the door opening distance can be obtained from the door controller. The door edge can also be identified through the image collected by the sensor, and the door opening distance can be calculated based on the pixels at the door edge combined with the sensor's installation parameters, imaging principles, etc. Figure 1 As shown, a sensor installed on the lintel of the elevator door of the elevator car can collect images during the process of the elevator door opening. After the door edge is identified in each frame of the image, the door opening distance is calculated by combining the pixels where the door edge is located, the installation position of the sensor on the lintel, the parameters of the sensor image collection, etc. with the imaging principle.

[0056] Of course, the door opening distance may also be detected by a distance sensor installed on the elevator door. This embodiment does not limit the method of detecting the door opening distance.

[0057] S302: When the door opening distance is greater than a preset distance threshold, control the sensor with a shooting direction toward the elevator lobby to capture multiple frames of images at a preset frame rate to obtain an image sequence.

[0058] The preset distance threshold can be determined based on the sensor's field of view, the width of the elevator door, and the sensor's installation position. This embodiment does not impose any restrictions on the preset distance threshold, so that the proportion of the elevator lobby in the image captured by the sensor through the gap in the open elevator door is greater than the preset proportion.

[0059] When the door opening distance is greater than a preset distance threshold, if there is only one sensor, the sensor is controlled to face the elevator lobby to collect images. If there are two sensors, the sensor facing the elevator lobby is controlled to collect images. The frame rate of the image collected by the sensor can be a fixed frame rate or a dynamically adjustable frame rate.

[0060] After the sensor captures the image, an image sequence can be generated according to the order in which the images are captured. It should be noted that when there are two sensors, the images captured by the two sensors can be spliced ​​together to obtain a spliced ​​image, and the image sequence can be generated using the spliced ​​image.

[0061] In this embodiment, when the door opening distance of the elevator door is greater than a preset distance threshold, the sensor is controlled to capture images at a preset frame rate to form an image sequence, so as to avoid the sensor continuously capturing images and capturing invalid images when the door opening distance is small. On the one hand, it can reduce the number of images captured by the sensor and extend the service life of the sensor. On the other hand, it can reduce the amount of image data processing.

[0062] S303: Input the image sequence into a pre-trained object detection and tracking model to obtain at least two items of status data for each object. Each item of object status data includes the orientation, position, volume, and movement speed of the object.

[0063] In this embodiment, the images captured by the sensor can be binarized to generate an image sequence, which is then input into the object detection and tracking model to obtain the orientation, position, volume, and movement speed of each object. Of course, other status data such as the movement trajectory of each object can also be included.

[0064] like Figure 1 As shown, taking a person as an example, the orientation data is the angle θ between the person's face direction and the plane where the elevator door 4 is located, the position refers to the vertical distance of the person in the elevator lobby relative to the elevator door 4, and the volume can be the volume of the detected outer contour of the person. The volume of a person when he is still is smaller than the volume when he is walking.

[0065] It should be noted that the object detection and tracking model can be a neural network such as RNN, CNN, DNN, etc. When training the object detection and tracking model, images with the orientation, position, volume, and movement speed of the objects in the image marked can be used as training data. The marked image sequence is input into the object detection and tracking model to predict the orientation, position, volume, and movement speed of each object in the image sequence. The loss rate is calculated by the predicted data and the marked data, and the model parameters are adjusted according to the loss rate until the loss rate is less than the preset value to obtain a trained object detection and tracking model. Among them, this embodiment does not limit the model structure or the training method of the model.

[0066] S304: Searching for a direction weight coefficient that matches the orientation of the object in a preset orientation-direction weight coefficient table.

[0067] like Figure 1 As shown, different orientations of the object can indicate the object's intention to take the elevator. Figure 1 In the figure, objects A and D are facing the elevator door 4, while object B is facing parallel to the elevator door. It can be determined that objects A and D are more likely to take the elevator than object B. Object C is facing an angle θ less than 90° with the elevator door 4. Therefore, the probability of object C taking the elevator is greater than that of object B but less than that of objects A and D.

[0068] Specifically in the embodiment of the present invention, weights corresponding to different orientations may be pre-set to generate an orientation-direction weight coefficient table. For example, the orientation-direction weight coefficient table is shown in Table 1 below:

[0069] Table 1:

[0070] Orientation range θ Directional weight coefficient θ∈(0, 30°] 0.2 θ∈(30°,65°] 0.7 θ∈(65°,90°] 0.9

[0071] The above Table 1 is only an example. Those skilled in the art may also set different orientation ranges and corresponding direction weight coefficients, which is not limited in the embodiment of the present invention.

[0072] In another optional embodiment, the corresponding direction weight coefficient may be calculated according to the orientation. For example, the direction weight coefficient is positively correlated with the orientation, so that the corresponding direction weight coefficient may be dynamically calculated according to different orientations.

[0073] S305: Determine a position weight coefficient of the object according to the position of the object.

[0074] like Figure 1 As shown, the distance between the object and the elevator door is different, and the probability of the object taking the elevator is also different. Generally, the closer the object is to the elevator door 4, the greater the probability of the object taking the elevator. Figure 1 In the figure, objects A and C are closer to the elevator door 4, while object B is farther away from the elevator door 4. It can be determined that objects A and C are more likely to take the elevator than object B. Therefore, a position-position weight coefficient table can be set in advance, and the position weight coefficient corresponding to the position of the object can be found in the position-position weight coefficient table.

[0075] In another optional embodiment, the position weight coefficient may also be calculated according to the position. The closer the position is to the elevator door, the larger the position weight coefficient is, that is, the position weight coefficient is positively correlated with the position.

[0076] S306: Calculate the swing amplitude of the object based on the volume.

[0077] The volume referred to in this embodiment is the volume of the outer contour of the detected object. When the object is stationary, the feet and hands of the object are almost motionless. Taking a person as an example, when the elevator door opens, a passenger intending to take the elevator walks from the elevator lobby to the car. His hands will swing and his feet will take steps alternately, causing the passenger's outer contour to change and the volume of his outer contour to also change. The maximum volume and the minimum volume can be found from the volume of the passenger determined by multiple frames of images, and the swing amplitude can be calculated based on the maximum volume and the minimum volume. For example, when the elevator door opening sensor captures the first frame of image, the passenger taking the elevator needs to wait for the passengers in the car to walk out of the elevator before he starts to walk into the elevator. At this time, the passenger is stationary and his outer contour volume may be the smallest. In the process of the passenger walking into the car, his feet move and the outer contour volume becomes larger. The difference between the maximum volume and the minimum volume can be calculated, and the ratio of the difference to the minimum volume can be calculated as the swing amplitude.

[0078] In another optional embodiment, the swing amplitude of the object may also be measured according to the number of pixels occupied by the object in each frame of the image sequence. This embodiment does not limit the method of calculating the swing amplitude of the object.

[0079] S307 : Searching for a swing amplitude weight coefficient that matches the swing amplitude in a preset swing amplitude-swing amplitude weight coefficient table.

[0080] The larger the swing amplitude of the object, the greater the probability that the object in the elevator lobby moves after the elevator door opens, and the greater the probability that the object takes the elevator. The larger the swing amplitude weight coefficient corresponding to the swing amplitude is. The amplitude weight coefficient that matches the swing amplitude can be found in the pre-set swing amplitude-swing amplitude weight coefficient table.

[0081] S308: Searching for a speed weight coefficient that matches the moving speed of the object in a preset moving speed-speed weight coefficient table.

[0082] When a subject takes an elevator with the elevator door open, the door's opening time is limited, so passengers usually enter the elevator quickly. That is, the faster the subject walks toward the elevator door, the more obvious the subject's intention to take the elevator is, and the larger the speed weight coefficient corresponding to the subject's speed. The speed weight coefficient that matches the subject's moving speed can be found from the moving speed-speed weight coefficient table.

[0083] S309: Calculate a weighted sum using the orientation, direction weight coefficient, position, position weight coefficient, swing amplitude, swing amplitude weight coefficient, moving speed, and speed weight coefficient to serve as the behavior score of the object.

[0084] In an optional embodiment, the behavior score can be calculated by the following formula:

[0085] S=r1×a+r2×b+r3×c+r4×d

[0086] In the above formula, r1, r2, r3, and r4 are the direction weight coefficient, position weight coefficient, swing amplitude weight coefficient, and speed weight coefficient respectively, and a, b, c, and d are the direction, position, swing amplitude, and moving speed respectively.

[0087] In another embodiment, the sum of the direction weight coefficient, the position weight coefficient, the swing amplitude weight coefficient, and the speed weight coefficient may be directly calculated as the behavior score of the object.

[0088] S310: When the behavior score is greater than a preset score threshold, determine that the object's behavior intention is to take an elevator.

[0089] In this embodiment, a score threshold can be set. When the behavior score of the object is greater than the preset score threshold, it is determined that the object is likely to take the elevator. The object's behavioral intention can be determined to be taking the elevator, so that tracking detection is performed on the object with the behavioral intention of taking the elevator. When the behavior score is less than the preset score threshold, it is determined that the object has no need to take the elevator.

[0090] S311. Determine, based on the current image, whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator.

[0091] Optionally, target detection and tracking can be performed on the target objects whose behavioral intention is to take the elevator to obtain the position of each target object in the current image. When the positions of the target objects are all inside the car, it is determined that the objects in the elevator lobby whose behavioral intention is to take the elevator have entered the car, and S313 is executed, otherwise S312 is executed.

[0092] S312: Control the elevator door according to the position of the target object.

[0093] In an optional embodiment, it is possible to determine whether a target object located in a preset door-interval area is detected in the current image. If so, the elevator door is controlled to remain open, the countdown timer is started, and the next frame image is obtained as the current image. It is further determined whether a target object located in the door-interval area of ​​the elevator door is detected in the current image. If not, it is determined whether a target object located in a preset near-door area on the car side is detected in the current image. When a target object located in the near-door area is detected, the distance from the target object to the door-interval area is obtained, the countdown timer is adjusted according to the distance, and the elevator door is controlled according to the timing. When the target object located in the near-door area is not detected, the next frame image is obtained as the current image, and the step of determining whether there is a target object with the behavioral intention of taking the elevator in the elevator lobby based on the current image is returned.

[0094] Among them, when adjusting the countdown timer according to the distance, it can be determined whether the distance is less than a preset distance threshold. When the distance is less than or equal to the preset distance threshold, the countdown timer is subtracted from a first value (for example, subtracted from 1 second). When the distance is greater than the preset distance threshold, the countdown timer is subtracted from a second value (for example, subtracted from 2 seconds).

[0095] When controlling the elevator door according to the timing of the timer, it can be determined whether the countdown timer is greater than 0. If so, the elevator door is controlled to remain open, the next frame image is obtained as the current image, and the process returns to the step of determining whether a target object located in a preset near-door area on the car side is detected in the current image. If not, the process returns to the step of determining whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator based on the current image.

[0096] In order to enable those skilled in the art to more clearly understand the control process of the elevator door when an object in the elevator lobby enters the elevator car, the process of controlling the closing of the elevator door is described below with reference to the accompanying drawings and examples. Figure 3B As shown, the detection area is divided into the door area P1, the door area P2 and the elevator lobby area P3. The door area P1 refers to the area when the elevator door is open or closed, for example, it can be the width area formed by 10 cm on both sides of the track when the elevator door is open or closed. The door area P2 can be the width area formed by the elevator door track as the center, with a single side of 15 cm on the inside and a single side of 30 cm on the outside. The elevator lobby area P3 is the area of ​​the elevator lobby, such as Figure 3C As shown in FIG. 1 , the door control process of a target object A who intends to take the elevator from position A0 in the elevator lobby to position A3 in the elevator car is as follows:

[0097] S0. The elevator door is fully opened.

[0098] S1. Detect whether there is a target object in the inter-door area P1 through the current image. If so, execute S2; otherwise, execute S3.

[0099] In this embodiment, a target object sequence may be established, where the target object sequence includes various data of each target object, such as at least the position of the target object.

[0100] like Figure 3B As shown, if the target object A enters the position A1 of the door area P1 from the position A0 of the elevator lobby, it is determined that the target object A exists in the door area P1 and S2 is executed; otherwise, S3 is executed.

[0101] S2. If there is a target object in the door area P1, the elevator door is controlled to remain open, the timer starts counting down, the next frame of image is acquired, and the process returns to S1.

[0102] In this embodiment, if there is a target object intending to take the elevator in the door area P1, the initialization timer will not start the countdown for you, and because there is a target object in the door area, the elevator door needs to be kept open, that is, no closing signal is generated, and the next frame of image is continuously acquired and returned to S1 to continue to determine whether there is a target object in the door area.

[0103] S3. Detect whether there is a target object in the door-near area P2 through the current image. If so, execute S4; if not, execute S5.

[0104] Specifically, if Figure 3B As shown, if the target object A is not detected in the door area P1, it means that the target object A is still in the elevator lobby area P3, then continue to acquire the next frame image and return to S1, or the target object A has entered the position A3 in the car 1, then end the tracking of the target object A, if the target object is detected in the door-near area P2 on the car 1 side, it means that the target object enters the door-near area on the car 1 side from the door area P1, for example, the target object A moves from position A1 to position A2, then execute S4, otherwise execute S7.

[0105] S4: If there is a target object in the near-door area P2, obtain the distance between the target object in the near-door area P2 and the inter-door area P1;

[0106] S5. Adjust the countdown timer according to the distance.

[0107] Specifically, if Figure 3D As shown, S5 may include the following sub-steps:

[0108] S51, determining whether the distance is less than a preset distance threshold;

[0109] Exemplarily, the distance threshold is 10 cm. If the distance is less than or equal to 10 cm, execute S52; otherwise, execute S53.

[0110] S52: Subtract the first value from the countdown timer.

[0111] For example, when each frame image detects that the distance between the target object located in the near-door area P2 and the door area P1 is less than or equal to 10 cm, the countdown timer is subtracted by 1 second, that is, the target object has just left the door area and the elevator door is about to be closed, but there is a risk in closing the door immediately. At the same time, in order to avoid waiting too long for the timer to count down to 0, the timer can be subtracted by 1 second.

[0112] S53: Subtract the second value from the countdown timer.

[0113] For example, if a target object is detected in the door-near area P2 at a distance greater than 10 cm from the door-interval area P1 in each image frame, the countdown timer is decremented by 2 seconds. Specifically, if the target object is farther away from the door-interval area after passing the door-interval area and entering the door-near area on the elevator car side, the risk of the target object being pinched by the door closing is reduced. Whenever a target object is detected in the door-near area and at a distance greater than 10 cm from the door-interval area in a frame, the elevator door is immediately closed. However, immediate door closing carries risks. To avoid the excessive waiting time required for the timer to count down to 0, the timer is decremented by 2 seconds to shorten the waiting time for the door to close. Of course, if another target object is detected in the door-interval area at the same time, the countdown timer is reset and the countdown restarts.

[0114] S6. Control the elevator door according to the timing.

[0115] Specifically, such as Figure 3D As shown, S6 may include the following sub-steps:

[0116] S61, determine whether the timing counter is greater than 0, if so, execute S62, if not, execute S63;

[0117] S62, keep the elevator door open, obtain the next frame image as the current image, and return to S1;

[0118] S63: When the current image detects that there is no target object in the elevator lobby area, the elevator door is controlled to close.

[0119] S7. When the target object is not detected in the door-near area, the next frame image is acquired as the current image, and the process returns to S1.

[0120] It should be noted that each time the current image captured by the sensor is received, S311 is first executed. If it is determined to execute S312 based on S311, S1-S7 are executed. That is, in this embodiment, each time a frame of image is received, it is first determined whether there is still a target object taking the elevator in the elevator lobby. If there is no target object taking the elevator, the elevator door is closed. If there is still a target object taking the elevator, S1-S7 are executed to reset the timer countdown when the target object is detected in the door area in each frame of the image, and the distance is calculated according to the position of the target object in the near-door area on the car side to adjust the timer timing. When the timing is greater than 0, the door is not closed and the next frame of image is obtained. When the timer is less than or equal to 0, if there is no target object with the intention of taking the elevator in the elevator lobby, the elevator door is controlled to close, so that the time of closing the elevator door can be dynamically adjusted according to the position of the target object taking the elevator in the process of entering the car.

[0121] S313, control the elevator door to close.

[0122] When the elevator car stops at the elevator lobby on the target floor and opens the elevator door, the passengers in the car get out first, and then the passengers in the elevator lobby who need to take the elevator enter the car. If it is determined that all the passengers in the elevator lobby who need to take the elevator have entered the car, the elevator door can be controlled to close without waiting for the preset door closing time, thereby improving the operating efficiency of the elevator.

[0123] In this embodiment, the image sequence collected by the sensor is input into a pre-trained object detection and tracking model to obtain the orientation, position, volume, and moving speed of each object. After obtaining the weight coefficients of the orientation, position, volume, and moving speed, the weighted sum is calculated as the behavior score of the object. When the behavior score is greater than the preset score threshold, the object's behavioral intention is determined to be taking the elevator. This solves the problem of proneness to misjudgment of the behavior intention when determining the object's behavioral intention based on the speed of movement. The accuracy of determining the object's behavioral intention by comprehensively considering the object's orientation, position, volume, and moving speed is high, thereby improving the accuracy of identifying objects outside the elevator door with the intention of taking the elevator, which is conducive to accurately controlling the closing time of the elevator door and improving the operating efficiency of the elevator.

[0124] Furthermore, based on the current image, it is determined that there is no target object in the elevator lobby with the behavioral intention of taking the elevator, and the elevator door is controlled to close without waiting for the preset door closing time, thereby improving the operating efficiency of the elevator.

[0125] Furthermore, when it is determined based on the current image that there is a target object in the elevator lobby with the behavioral intention of taking the elevator, the timer countdown is started when the target object is in the door area, and the timer is shortened when the target object enters from the door area to the near-door area on the car side. That is, the timer timing is dynamically adjusted according to the various positions of the target object entering the car from the elevator lobby, and finally the elevator door is controlled to close when the timer is equal to 0 and there is no target object with the intention of taking the elevator in the elevator lobby, so that the closing of the elevator door can be flexibly controlled.

[0126] Example 3

[0127] Figure 4 This is a schematic diagram of the structure of an object behavior determination device provided by the third embodiment of the present invention. Figure 4 As shown, the object behavior determination device includes:

[0128] Image acquisition module 401, for controlling the sensor to capture multiple frames of images of the elevator lobby area when the elevator car arrives at the elevator lobby and the elevator door opens;

[0129] An object detection and tracking module 402 is configured to perform object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby;

[0130] A behavior score acquisition module 403 is configured to calculate the behavior score of the object based on a preset weight coefficient of each status data and the status data;

[0131] The behavior intention determination module 404 is configured to determine that the behavior intention of the subject is to take an elevator when the behavior score is greater than a preset score threshold.

[0132] Optionally, the image acquisition module 401 includes:

[0133] The door opening distance detection unit is used to detect the door opening distance of the elevator door when the elevator car arrives at the elevator lobby;

[0134] The image acquisition unit is used to control the sensor with a shooting direction toward the elevator lobby to acquire multiple frames of images at a preset frame rate to obtain an image sequence when the door opening distance is greater than a preset distance threshold.

[0135] Optionally, the object detection and tracking module 402 includes:

[0136] An image sequence generating unit, configured to generate an image sequence from the acquired multiple frames of images according to the acquisition time of the images;

[0137] The model input unit is used to input the image sequence into a pre-trained object detection and tracking model to obtain at least two status data of each object.

[0138] Optionally, the state data includes at least two of the orientation, position, volume, and movement speed of each object.

[0139] Optionally, the behavior score acquisition module 403 includes:

[0140] a direction weight coefficient obtaining unit, configured to search a preset direction weight coefficient table for a direction weight coefficient that matches the direction of the object;

[0141] a position weight coefficient obtaining unit, configured to determine a position weight coefficient of the object according to the position of the object;

[0142] a swing amplitude calculation unit, configured to calculate the swing amplitude of the object based on the volume;

[0143] a swing amplitude weight coefficient obtaining unit, configured to search a preset swing amplitude-swing amplitude weight coefficient table for a swing amplitude weight coefficient that matches the swing amplitude;

[0144] a speed weight coefficient obtaining unit, configured to search a preset moving speed-speed weight coefficient table for a speed weight coefficient that matches the moving speed of the object;

[0145] The behavior score calculation unit is used to calculate a weighted sum using the orientation, direction weight coefficient, position, position weight coefficient, swing amplitude, swing amplitude weight coefficient, moving speed, and speed weight coefficient to serve as the behavior score of the object.

[0146] Optionally, it also includes:

[0147] A target object judgment module is used to judge whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator based on the current image;

[0148] a first elevator door control module, configured to control the elevator door according to the position of the target object;

[0149] The second elevator door control module is used to control the elevator door to close.

[0150] Optionally, the target object determination module includes:

[0151] A position tracking unit, configured to perform target detection and tracking on the target object to obtain the position of the target object in the current image;

[0152] a first determining unit configured to determine that no target object with a behavioral intention of taking the elevator exists in the elevator lobby when all the target objects are located in the elevator car;

[0153] The second determining unit is configured to determine that there is a target object with a behavioral intention of taking the elevator in the elevator lobby when the target object is located outside the elevator door.

[0154] Optionally, the first elevator door control module includes:

[0155] an inter-door area determination subunit, configured to determine whether a target object located in a preset inter-door area is detected in the current image;

[0156] The elevator door is kept open control subunit, which is used to control the elevator door to be kept open, start the countdown timer, and obtain the next frame image as the current image, and return to the door area judgment subunit;

[0157] a door-near-area determination subunit, configured to determine whether a target object located within a door-near area preset on the car side is detected in the current image;

[0158] A distance acquisition subunit is configured to acquire the distance between the target object and the inter-door area when a target object is detected in the near-door area;

[0159] a timer adjustment subunit, configured to adjust the timing of the countdown timer according to the distance, and control the elevator door according to the timing;

[0160] The image acquisition subunit is used to acquire the next frame of image as the current image when the target object located in the near-door area is not detected, and return it to the target object judgment module.

[0161] The timer adjustment subunit includes:

[0162] A distance determination component, configured to determine whether the distance is less than a preset distance threshold;

[0163] a first timer adjustment component, configured to subtract a first value from the countdown timer when the distance is less than or equal to the preset distance threshold;

[0164] The second timer adjustment component is used to subtract a second value from the countdown timer when the distance is greater than the preset distance threshold.

[0165] Optionally, the timer adjustment subunit includes:

[0166] A timing judgment component, used to judge whether the countdown timer is greater than 0;

[0167] An elevator door keeping open control component is used to control the elevator door to keep open, obtain the next frame image as the current image, and return to the door area judgment subunit;

[0168] The image acquisition component is used to obtain the next frame of image and return it to the target object judgment module.

[0169] The object behavior determination device provided in the embodiment of the present invention can execute the object behavior determination method provided in the first and second embodiments of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0170] Example 4

[0171] Figure 5 A schematic diagram of the structure of an electronic device 50 that can be used to implement an embodiment of the present invention is shown. The electronic device 50 is intended to represent various forms of digital computers, such as desktop computers, workstations, servers, blade servers, mainframe computers, etc. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0172] like Figure 5As shown, the electronic device 50 includes at least one processor 51 and a memory, such as a read-only memory (ROM) 52, a random access memory (RAM) 53, etc., which is communicatively connected to the at least one processor 51. The memory stores a computer program that can be executed by the at least one processor. The processor 51 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 52 or the computer program loaded from the storage unit 58 into the random access memory (RAM) 53. Various programs and data required for the operation of the electronic device 50 can also be stored in the RAM 53. The processor 51, ROM 52, and RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0173] Multiple components in the electronic device 50 are connected to the I / O interface 55, including an input unit 56, such as a keyboard, a mouse, a sensor, etc.; an output unit 57, such as various types of displays, speakers, etc.; a storage unit 58, such as a magnetic disk, an optical disk, etc.; and a communication unit 59, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 59 allows the electronic device 50 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0174] The processor 51 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 51 executes the various methods and processes described above, such as the object behavior determination method.

[0175] In some embodiments, the object behavior determination method can be implemented as a computer program that is tangibly contained in a computer-readable storage medium, such as storage unit 58. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 50 via ROM 52 and / or communication unit 59. When the computer program is loaded into RAM 53 and executed by processor 51, one or more steps of the object behavior determination method described above can be performed. Alternatively, in other embodiments, processor 51 can be configured to perform the object behavior determination method in any other appropriate manner (e.g., by means of firmware).

[0176] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0177] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0178] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0179] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0180] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0181] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0182] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0183] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for determining object behavior, characterized in that: Used to determine the behavior of objects in the elevator lobby, including: When the elevator car arrives at the elevator lobby and the elevator door opens, the control sensor collects multiple frames of images of the area in the elevator lobby; Performing object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby; Calculating a behavior score of the subject based on a preset weight coefficient of each status data and the status data; When the behavior score is greater than a preset score threshold, determining that the subject's behavior intention is to take the elevator, and controlling the elevator door according to the subject's position; When it is determined based on the current image that there is no target object in the elevator lobby with the behavioral intention of taking the elevator, controlling the elevator door to close; The controlling the elevator door according to the position of the object comprises: Determining whether a target object located in a preset inter-door area is detected in the current image; If so, controlling the elevator door to remain open, starting a countdown timer, acquiring the next frame of image as the current image, and returning to the step of determining whether the target object located in the door area of ​​the elevator door is detected in the current image; If not, determining whether a target object located in a preset door-near area on the car side is detected in the current image; When a target object is detected in the near-door area, obtaining a distance from the target object to the inter-door area; adjusting the countdown timer according to the distance, and controlling the elevator door according to the countdown timer; When no target object is detected in the door-near area, the next frame image is acquired as the current image, and the process returns to the step of determining whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator based on the current image.

2. The method according to claim 1, wherein When the elevator car arrives at the elevator lobby and the elevator door opens, controlling the sensor to collect multiple frames of images of the area of ​​the elevator lobby includes: When the elevator car arrives at the elevator lobby, detect the door opening distance of the elevator; When the door opening distance is greater than a preset distance threshold, the sensor is controlled to capture multiple frames of images at a preset frame rate to obtain an image sequence.

3. The method according to claim 1, wherein The performing object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby includes: Generating an image sequence from the acquired multiple frames of images according to the acquisition time of the images; The image sequence is input into a pre-trained object detection and tracking model to obtain at least two status data of each object.

4. The method according to claim 1, wherein The state data includes at least two items of the orientation, position, volume, and moving speed of each object.

5. The method according to claim 1, wherein The state data includes the direction, position, volume, and movement speed of the object. The calculation of the behavior score of the object based on the preset weight coefficient of each state data and the state data includes: Searching for a direction weight coefficient that matches the orientation of the object in a preset orientation-direction weight coefficient table, wherein the orientation-direction weight coefficient table includes pre-configured direction weight coefficients for different orientation ranges; determining a position weight coefficient of the object according to the position of the object; calculating a swing amplitude of the object based on the volume; Searching for a swing amplitude weight coefficient that matches the swing amplitude in a preset swing amplitude-swing amplitude weight coefficient table, wherein a larger swing amplitude corresponds to a larger swing amplitude weight coefficient in the swing amplitude-swing amplitude weight coefficient table; Searching for a speed weight coefficient that matches the moving speed of the object in a preset moving speed-speed weight coefficient table, wherein a greater moving speed corresponds to a greater speed weight coefficient in the moving speed-speed weight coefficient table; The orientation, direction weight coefficient, position, position weight coefficient, swing amplitude, swing amplitude weight coefficient, moving speed, and speed weight coefficient are used to calculate a weighted sum to serve as the behavior score of the object.

6. The method according to claim 1, wherein The determining, based on the current image, whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator includes: Performing target detection and tracking on the target object to obtain the position of the target object in the current image; When the positions of the target objects are all located in the elevator car, determining that there are no target objects in the elevator lobby with the behavioral intention of taking the elevator; When the target object is located outside the elevator door, it is determined that there is a target object in the elevator lobby with a behavioral intention of taking the elevator.

7. The method according to claim 1, wherein The adjusting the timing of the countdown timer according to the distance includes: Determining whether the distance is less than a preset distance threshold; When the distance is less than or equal to the preset distance threshold, the countdown timer is deducted by a first value; When the distance is greater than the preset distance threshold, the countdown timer is reduced by a second value.

8. The method according to claim 1, wherein controlling the elevator door according to the timing comprises: Determine whether the countdown timer is greater than 0; If so, controlling the elevator door to remain open, acquiring the next frame of image as the current image, and returning to the step of determining whether the target object located in the preset near-door area on the car side is detected in the current image; If not, the next frame of image is acquired, and the process returns to the step of determining whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator based on the current image.

9. An object behavior determination device, characterized in that: The method is applied to determine the behavior of an object in an elevator lobby, specifically to implement the object behavior determination method according to any one of claims 1 to 8, comprising: An image acquisition module is used to control the sensor to capture multiple frames of images of the elevator lobby area when the elevator car arrives at the elevator lobby and the elevator door opens; an object detection and tracking module, configured to perform object detection and tracking on the image to obtain at least two pieces of status data for each object in the elevator lobby; A behavior score acquisition module, configured to calculate the behavior score of the object based on a preset weight coefficient of each status data and the status data; a behavior intention determination module, configured to determine that the behavior intention of the subject is to take an elevator when the behavior score is greater than a preset score threshold; A target object judgment module is used to judge whether there is a target object in the elevator lobby with the behavioral intention of taking the elevator based on the current image; a first elevator door control module, configured to control the elevator door according to the position of the target object; A second elevator door control module, used for controlling the closing of the elevator door; The first elevator door control module includes: an inter-door area determination subunit, configured to determine whether a target object located in a preset inter-door area is detected in the current image; The elevator door is kept open control subunit, which is used to control the elevator door to be kept open, start the countdown timer, and obtain the next frame image as the current image, and return to the door area judgment subunit; a door-near-area determination subunit, configured to determine whether a target object located within a door-near area preset on the car side is detected in the current image; A distance acquisition subunit is configured to acquire the distance between the target object and the inter-door area when a target object is detected in the near-door area; a timer adjustment subunit, configured to adjust the timing of the countdown timer according to the distance, and control the elevator door according to the timing; The image acquisition subunit is used to acquire the next frame of image as the current image when the target object located in the near-door area is not detected, and return it to the target object judgment module.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the object behavior determination method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the object behavior determination method according to any one of claims 1 to 8 when executed.

Images