A method for reducing false triggering of multi-channel gates, and gate system

Abstract

The present application discloses a method for reducing false triggering of multi-channel gates, characterized in that the method includes receiving target image frames from image recognition devices of each channel, and when the same target exists in each target image frame, calculating the offset of the target in each target image frame relative to the center of the target image frame and / or the target screen ratio, comparing the offset and / or the target screen ratio of each target image frame, and taking the channel corresponding to the target image frame with the smallest offset and / or the largest target screen ratio as the correct channel for triggering. The method for reducing false triggering of multi-channel gates provided by the present application does not require human-computer interaction, does not require modification of the installation method of the gate, has a wide range of adaptability to application scenarios, improves the safety of the gate without affecting the traffic efficiency, and improves the user experience.

Description

Technical Field

[0001] The present invention relates to the field of access control, and in particular to a method for reducing false triggering of a multi-channel gate. Background Art

[0002] The gate is a channel blocking device (channel management equipment), which is mainly used in urban rail transit, scenic spots, residential areas and other scenes to manage the flow of people and regulate the entry and exit of pedestrians. Its basic core function is to achieve only one person passing through at a time, and it can be used at the entrance and exit of various toll and access control occasions.

[0003] Existing gate machines are usually equipped with an image recognition device for a single channel, which is responsible for the permission authentication function of the channel. When the target enters the identification area, the image recognition device immediately starts working, and the channel is opened after the permission authentication is passed. In common usage scenarios such as residential areas, scenic spots, and parks, multiple channels are often used in combination, and the permissions of each channel are independently determined by the image recognition devices of each channel. Since the lens of the image recognition device has a certain field of view (common range of 50°~90°), when the same target approaches the channel, there is a situation where the same target enters the identification areas of several adjacent gate channels at the same time, resulting in multiple gate channels being successfully authenticated and opened, which poses a potential access control security risk.

[0004] See also Figure 1 As shown, Figure 1 A schematic diagram of the recognition area of ​​the protruding recognition device of the same target being located in two channels at the same time. The same target is located in the recognition area A of the image recognition device A of channel A and the recognition area B of the image recognition device B of channel B at the same time, and there is no obvious difference between the distance A of the target and the distance B of the target from the image recognition device A and the distance B of the target from the image recognition device B. When the target is authenticated, both channel A and channel B will be triggered to open.

[0005] In order to avoid the false triggering of multiple channels of the gate being opened or closed for the same target, the existing methods are:

[0006] (1) When installing the gate, adjust the angle of the image recognition device to avoid the same target being located in the intersection area of ​​the recognition areas of the image recognition devices of each channel, for example, Figure 1 In the embodiment of the present invention, the installation angle of the image recognition device is adjusted. Or similarly, a shielding device is added to the image recognition device.

[0007] The advantage of this method is that it is easy to operate, but its disadvantage is that after adjusting the angle of the image recognition device or increasing the obstruction, the effective recognition range in the channel direction will also become smaller. The difference in recognition angles will also increase the risk of misrecognition, which will reduce the user experience of using the channel and cannot completely eliminate the problem of false triggering.

[0008] (2) Adding authentication steps to solve the problem, for example, through a combination of card swiping and image recognition, or confirming on the image recognition device after image recognition is completed. This method greatly reduces the authentication efficiency of the channel and is also applicable to a limited number of scenarios. Summary of the invention

[0009] The present invention provides a method for reducing false triggering of a multi-channel gate, so as to solve the problem of false triggering operation of multiple channels caused by image recognition of the same target by the multi-channel gate.

[0010] The present invention provides a method for reducing false triggering of a multi-channel gate, the method comprising:

[0011] receiving target image frames from image recognition devices of each channel,

[0012] When the same target exists in each target image frame, the offset of the target in each target image frame relative to the center of the target image frame and / or the target screen ratio are calculated.

[0013] The offsets and / or target screen ratios of the target image frames are compared, and the channel corresponding to the target image frame having the smallest offset and / or the largest target screen ratio is used as the correct channel to be triggered.

[0014] Preferably, the method further comprises,

[0015] According to the source address information of the target image frame with the minimum offset and / or the maximum target screen ratio, the corresponding image recognition device is notified to perform authority authentication, so that the image recognition device performs a channel triggering operation when the authority authentication passes.

[0016] Preferably, the receiving of the target image frame from the image recognition device of each channel further comprises:

[0017] According to the received target image frame, it is determined whether more than two image recognition devices recognize the target. If so, the same target is recognized based on each target image frame.

[0018] Otherwise, the channel corresponding to the target image frame is used as the correct channel to trigger, and the image recognition device is notified to perform authority authentication.

[0019] Preferably, the calculating of the offset of the target in each target image frame relative to the center of the target image frame and / or the target screen ratio includes:

[0020] For each target image frame:

[0021] The target image frame is rasterized.

[0022] Mark the grid position information of the target feature point in the target image frame, and calculate the offset of the target feature point relative to the center of the target image frame based on the grid position information of the target feature point and the grid position information of the center of the target image frame.

[0023] and / or

[0024] Mark the target contour range in the target image frame, and calculate the target screen ratio according to the area within the target contour range and the area of ​​the target image frame.

[0025] Preferably, the calculation of the offset of the target feature point relative to the center of the target image frame according to the grid position information of the target feature point and the grid position information of the center of the target image frame includes:

[0026] Calculate the absolute value of the difference between the horizontal coordinate of the target feature point and the horizontal coordinate of the center of the target image frame to obtain the offset;

[0027] When the target is a face, the target feature point is the center point between the pupils of both eyes.

[0028] Preferably, the target screen ratio is calculated according to the area within the target contour range and the area of ​​the target image frame, including:

[0029] Count the number of grids within the target contour range,

[0030] Count the number of grids in the target image frame,

[0031] The ratio of the number of grids within the target contour to the number of grids in the target image frame is calculated to obtain the target screen ratio.

[0032] Preferably, the comparing the offsets and / or target screen ratios of the target image frames and taking the channel corresponding to the target image frame having the smallest offset and / or the largest target screen ratio as the correct channel to trigger includes:

[0033] Eliminate target image frames whose target screen ratio is less than the target screen ratio threshold, and retain target image frames whose target screen ratio is not less than the target screen ratio threshold;

[0034] From the retained target image frames, a target image frame with a minimum offset and / or a maximum target screen ratio is selected,

[0035] in,

[0036] The target screen ratio threshold is determined according to the ratio of the area within the target contour range obtained at the target maximum recognition distance to the total area of ​​the target image frame.

[0037] The present invention provides a multi-channel gate system, comprising at least two or more image recognition devices, each channel is equipped with an image recognition device, the image recognition device of each channel is respectively connected to a network side device, each image recognition device sends an acquired target image frame to the network side device, so that the network side device executes the steps of any method for reducing false triggering of a multi-channel gate as claimed in any one of claims 1 to 7.

[0038] Preferably, the network side device includes a cloud platform.

[0039] The present invention also provides a network side device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement any step of the method for reducing false triggering of a multi-channel gate.

[0040] The present invention also provides a multi-channel gate device, including a control device and at least two image recognition devices, each channel is equipped with an image recognition device, the control device includes a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to implement the steps of any method for reducing false triggering of a multi-channel gate as claimed in any one of claims 1 to 7.

[0041] The present invention further provides a multi-channel gate device, comprising at least n image recognition devices, each channel is equipped with an image recognition device, n is a natural number greater than or equal to 2, wherein at least one image recognition device is a designated image recognition device, used to implement the steps of any method for reducing false triggering of a multi-channel gate as claimed in claims 1 to 7.

[0042] Preferably, the designated image recognition device is one of the image recognition devices having intersecting recognition areas;

[0043] Preferably, the number of the designated image recognition devices is n-1.

[0044] The present application provides a method for reducing the false triggering of multi-channel gates, which determines the trigger channel for the same target located in the intersection area of ​​the recognition areas of multiple image recognition devices by comparing the target offset and / or target screen ratio of each target image frame, thereby solving the problem of false triggering of multiple channels. The present application provides a method for reducing the false triggering of multi-channel gates, which does not require human-computer interaction or modification of the gate installation method, and has a wide range of adaptability to application scenarios. It improves the safety of the gates and the user experience without affecting the traffic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of the recognition area of ​​a protruding recognition device for a same target being located in two channels simultaneously.

[0046] Figure 2 A flow chart of a method for reducing false triggering of multi-channel gates according to an embodiment of the present application.

[0047] Figure 3 A flowchart of a method for reducing false triggering of multi-channel gates on the network side according to an embodiment of the present application.

[0048] Figure 4 A schematic diagram of rasterizing and labeling facial image frames detected by a channel image recognition device.

[0049] Figure 5 For Figure 1 The example of target image frame A captured by image recognition device A is shown in FIG. 1 , and the example on the right is a schematic diagram of target image frame B captured by image recognition device B.

[0050] Figure 6 A flow chart of a method for reducing false triggering of a multi-channel gate machine on the local control device side of the gate machine is provided.

[0051] Figure 7 A flow chart of a method for reducing false triggering of a multi-channel gate machine implemented on any image recognition device side of the gate machine.

[0052] Figure 8 A schematic diagram of the distribution of specified image recognition devices.

[0053] Fig. 9 A schematic diagram of a multi-channel gate system.

[0054] Fig.10 A schematic diagram of a multi-channel gate device.

[0055] Fig.11 A schematic diagram of a network side device according to an embodiment of the present application.

[0056] Fig.12 A schematic diagram of a network side device, a control device, or an image recognition device according to an embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the objectives, technical means and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings.

[0058] The present application provides a method for reducing false triggering of a multi-channel gate, which triggers the correct channel by utilizing the difference in target image frames of the same target collected by an image recognition device.

[0059] See also Figure 2 As shown, Figure 2A flow chart of a method for reducing false triggering of multi-channel gates according to an embodiment of the present application. The method comprises the following steps:

[0060] Step 201, receiving a target image frame from an image recognition device of each channel,

[0061] Step 202: when the same target exists in each target image frame, calculate the offset of the target in each target image frame relative to the center of the target image frame and / or the target screen ratio.

[0062] Step 203: The channel corresponding to the target image frame having the minimum offset and / or the maximum target screen ratio is used as the correct channel to be triggered.

[0063] The embodiment of the present application solves the problem of false triggering of multi-channel gates from a software level, without the need for user participation in the operation, thereby improving the passage efficiency of the channel.

[0064] See also Figure 3 As shown, Figure 3 A flow chart of a method for reducing false triggering of multi-channel gates on the network side in an embodiment of the present application. The method includes:

[0065] Step 301: The image recognition device of each channel collects the target within its recognition area, and sends the acquired target image frame to the cloud platform through target detection.

[0066] The target image frame carries source address information of the image recognition device that sends the target image frame.

[0067] Step 302: The cloud platform determines whether the target image frames are received from multiple image recognition devices according to the source address information of the received target image frames.

[0068] If yes, then execute step 303,

[0069] Otherwise, execute step 304 to notify the image recognition device to perform authority authentication.

[0070] Step 303: The cloud platform determines whether the received target image frame contains the same target.

[0071] If the same target exists, then execute step 305,

[0072] Otherwise, execute step 304 to notify each image recognition device to perform authority authentication.

[0073] In this step, considering that target image frames of the same target are usually in adjacent channels, in order to improve the efficiency of identifying the same target, the target image frames from the image recognition devices of adjacent channels can be compared according to the source address information of the target image frames. For example, it can be determined whether the same target exists based on the Euclidean distance between frames.

[0074] Step 305: The cloud platform determines the correct channel to trigger based on the position of the same target in each target image frame and / or the target screen ratio.

[0075] In this step, see Figure 4 As shown, Figure 4 A schematic diagram of rasterizing and marking facial image frames detected by a channel image recognition device. For each target image frame, the cloud platform rasterizes the target image frame, marks the target grid range according to the target outline, and / or selects a grid within the target outline as a target feature point, for example, when the target is a face, the target feature point is the center point between the pupils of both eyes; and marks the grid position information of the target feature point. The finer the grid, the more conducive it is to improving the accuracy of the correct channel.

[0076] According to the grid position information of the target feature point, the offset of the target relative to the center of the target image frame is calculated. Considering that the image frames of the same target captured by the image recognition devices of multiple channels are usually reflected in the u direction (horizontal direction in the figure) of the image frame, see Figure 5 As shown, Figure 5 On the left Figure 1 The target image frame A is captured by the image recognition device A in the middle, and the target image frame B is captured by the image recognition device B on the right. Therefore, the offset of the grid position of the target feature point relative to the center of the target image frame in the horizontal direction is calculated, that is, the absolute value of the difference between the horizontal coordinate of the grid of the target feature point and the horizontal coordinate of the horizontal center axis of the target image frame is calculated, which is expressed as follows:

[0077] Offset = |grid horizontal coordinate of target feature point - horizontal coordinate of center axis of target image frame in horizontal direction |

[0078] According to the target grid range, the target screen ratio is calculated, that is, the ratio of the grid area within the target grid range to the total area of ​​the target image frame is calculated, such as Figure 4 As shown, the ratio of the area of ​​the gray grid area to the area of ​​all grids is calculated to obtain the target screen ratio, which can be expressed mathematically as follows:

[0079] FO = area within the target grid / total area of ​​the target image frame

[0080] = Number of grids within the target grid range / total number of grids in the target image frame

[0081] Among them, FO represents the target screen ratio.

[0082] According to the area within the target contour range collected at the maximum recognition distance of the image recognition device and the total area of ​​the target image frame, the target screen ratio threshold is determined as the system default value, that is, the target screen ratio threshold is the ratio of the area within the target contour range collected at the maximum recognition distance to the total area of ​​the target image frame, which is expressed in mathematical formula as follows:

[0083] Def_FO = area within the target contour / total area of ​​the target image frame

[0084] = Number of grids within target contour range / Total number of target image frames

[0085] Among them, Def_FO represents the target screen ratio threshold.

[0086] Given that the target is closer to the center of the image frame when facing the forward channel, and further away from the center of the image frame in the false trigger channel, the target offset (Offset) in the target image frame captured on the correct channel will be smaller;

[0087] At the same time, due to the perspective principle, the target of the false trigger channel will be relatively small, so the target screen ratio (FO) value collected on the correct channel will be larger. At the same time, combined with the target screen ratio threshold (Def_FO), the situation where the target screen ratio is small can be excluded according to the actual configuration.

[0088] Specifically,

[0089] For each target image frame, determine whether its target screen ratio is less than the target screen ratio threshold. If so, remove the target image frame; otherwise, retain the target image frame.

[0090] Among the retained target image frames, a target image frame having the largest target screen ratio and / or the smallest offset is selected.

[0091] The channel corresponding to the image recognition device from which the selected target image frame originates is taken as the correct channel.

[0092] Step 306: According to the source address information of the selected target image frame, the image recognition device corresponding to the source address information is notified to perform authority authentication.

[0093] Step 307, after receiving the permission authentication notification, the image recognition device performs permission authentication according to the permission database stored locally. When the authentication is passed, the channel corresponding to the image recognition device is triggered, otherwise, it is not triggered.

[0094] The embodiment of the present application utilizes the differences between different target image frames containing the same target to determine the correct channel to trigger, thereby avoiding false triggering of multiple channels. This embodiment does not require modification or additional installation of multi-channel gate machines, and does not affect the recognition and passage efficiency, thereby improving the user experience and increasing security and convenience for access control management.

[0095] To facilitate understanding of the present application, the following uses facial features as an example to illustrate the situation where the same target exists within the recognition range of an image recognition device with two adjacent channels. It should be understood that the target is not limited to facial features, and the number of channels is not limited to two channels.

[0096] Combination Figure 1 As shown, when the same target is simultaneously located in recognition area A of image recognition device A of channel A and recognition area B of image recognition device B of channel B, image recognition device A obtains target image frame A and sends target image frame A to the cloud platform. Similarly, image recognition device B obtains target image frame B and sends target image frame B to the cloud platform.

[0097] The cloud platform receives the target image frame A and the target image frame B, and determines that multiple image recognition devices recognize the target.

[0098] The cloud platform performs the same target recognition based on target image frame A and target image frame B to determine whether they belong to the same target.

[0099] When the same target is identified, the target image frames A and B are rasterized respectively, and the facial image proportions of the target image frames A and B are calculated respectively.

[0100] The target image frames A and B are screened according to the default screen ratio threshold, so as to eliminate the target image frames whose face screen ratio is less than the screen ratio threshold.

[0101] If no target image frame is excluded, the offset of the face in the target image frame is calculated respectively. Since the positions of both eyes are usually identified during facial recognition, the center point between the eyes is used as the target feature point, the grid position information of the center point between the eyes is marked, and the absolute value of the difference between the grid horizontal coordinate of the center point and the horizontal coordinate of the central axis of the target image frame is calculated. According to the facial image ratio and the offset, the channel corresponding to the target image frame with the smallest offset and / or the largest facial image ratio is used as the correct channel to trigger, and according to the source address information of the target image frame, the image recognition device is notified to perform authentication.

[0102] If a target image frame is excluded, the channel corresponding to the retained target image frame is used as the correct channel to trigger, and the image recognition device is notified to perform authentication based on the source address information of the retained target image frame.

[0103] This embodiment determines the correct channel for triggering through the cloud platform device on the network side, which facilitates remote management and maintenance of the gate equipment and is suitable for situations with widely distributed gate equipment.

[0104] See also Figure 6 As shown, Figure 6 A flow chart of a method for reducing false triggering of a multi-channel gate at a local control device side of the gate. The method includes:

[0105] Step 601 : The image recognition device of each channel respectively collects targets within its recognition area, and sends the acquired target image frames to the local control device through target detection.

[0106] The target image frame may be sent to the local control device via a wired or local wireless local area network.

[0107] Step 602: The local control device determines whether the target image frame is received from multiple image recognition devices according to the source address information of the received target image frame.

[0108] If yes, then execute step 603,

[0109] Otherwise, execute step 604 to notify the image recognition device to perform authority authentication.

[0110] Step 603: The local control device determines whether the received target image frame contains the same target.

[0111] If the same target exists, then execute step 605,

[0112] Otherwise, execute step 604 to notify each image recognition device to perform authority authentication.

[0113] Step 605: The local control device determines the correct channel to trigger according to the position of the same target in each target image frame and / or the target screen ratio.

[0114] The specific steps are the same as step 305.

[0115] Step 606: The local control device notifies the image recognition device corresponding to the source address information to perform authority authentication according to the source address information of the selected target image frame.

[0116] Step 607, after receiving the permission authentication notification, the image recognition device performs permission authentication according to the permission database stored locally. When the authentication is passed, the channel corresponding to the image recognition device is triggered, otherwise, it is not triggered.

[0117] This embodiment uses a local control device to determine the correct channel to trigger, which is low-cost and easy to be compatible with and upgrade existing gate equipment.

[0118] See also Figure 7 As shown, Figure 7 A flow chart of a method for reducing false triggering of a multi-channel gate machine on any image recognition device side of the gate machine. The method includes:

[0119] Step 701 : The image recognition device of each channel respectively collects targets within its recognition area, and sends the acquired target image frames to the designated image recognition device through target detection.

[0120] The target image frame may be sent to a designated image recognition device via a wired or local wireless LAN, Bluetooth, or the like.

[0121] The designated image recognition device may be determined in advance when the system is configured.

[0122] In addition, the designated image recognition device may be any one of the image recognition devices of all channels, or another local image recognition device, or one of the image recognition devices whose recognition areas have an intersection area, for example, Figure 1 Any one of the image recognition device A and the image recognition device B is used as the designated image recognition device.

[0123] See also Figure 8 As shown in the figure, the recognition area of ​​the image recognition device B intersects with the recognition area of ​​the image recognition device A and the recognition area of ​​the image recognition device C. Preferably, the image recognition device B serves as the designated image recognition device so that the correct channel for determining the trigger can be concentrated on the designated device. Multiple designated image recognition devices can also be designated. For example, any two of the image recognition devices A, B, and C are designated image recognition devices. In the figure, image recognition devices B and C are designated image recognition devices. In this way, n-1 of the n image recognition devices are designated image recognition devices, where n is a natural number greater than 2.

[0124] Step 702: The designated image recognition device determines whether the target image frame is received from multiple image recognition devices according to the source address information of the received target image frame.

[0125] If yes, then execute step 703,

[0126] Otherwise, execute step 704 to notify the image recognition device to perform authority authentication.

[0127] Step 703, the designated image recognition device determines whether the received target image frame contains the same target.

[0128] If the same target exists, then execute step 705,

[0129] Otherwise, execute step 704 to notify each image recognition device to perform authority authentication.

[0130] Step 705, specifying the image recognition device to determine the correct channel to trigger according to the position of the same target in each target image frame and / or the target screen ratio,

[0131] The specific steps are the same as step 305.

[0132] Step 706: the designated image recognition device notifies the image recognition device corresponding to the source address information to perform authority authentication according to the source address information of the selected target image frame.

[0133] Step 707, after receiving the permission authentication notification, the corresponding image recognition device performs permission authentication according to the permission database stored locally. When the authentication is passed, the channel corresponding to the image recognition device is triggered, otherwise, it is not triggered.

[0134] This embodiment uses an image recognition device to determine the correct channel for triggering, which is low-cost, flexible to install, highly adaptable, and easy to maintain. Fig. 9 As shown, Fig. 9 The diagram is a schematic diagram of a multi-channel gate system. The system includes two or more image recognition devices, each channel is equipped with an image recognition device, and each image recognition device is connected to a network side device, such as a server, a cloud platform, etc. Each image recognition device sends the acquired target image frame to the network side device, so that the network side device executes any step of the method for reducing the false triggering of the multi-channel gate.

[0135] See also Fig.10 As shown, Fig.10 A schematic diagram of a multi-channel gate device. The multi-channel gate device includes a control device and at least two image recognition devices, each channel is equipped with an image recognition device, each image recognition device is connected to the control device, and each image recognition device sends the acquired target image frame to the control device, so that the control device performs any step of the method for reducing the false triggering of the multi-channel gate.

[0136] See also Fig.11 As shown, Fig.11 This is a schematic diagram of a device for reducing false triggering of multi-channel gates according to an embodiment of the present application. The device includes:

[0137] The transceiver module is used to receive the target image frame from the image recognition device of each channel, and notify the image recognition device to perform authority authentication according to the correct channel determined.

[0138] The trigger channel determination unit is used to calculate the offset of the target in each target image frame relative to the center of the target image frame and / or the target screen ratio when the same target exists in each target image frame, and take the channel corresponding to the target image frame with the smallest offset and / or the largest target screen ratio as the correct channel for triggering.

[0139] Preferably, the trigger channel determination unit includes:

[0140] The regional target recognition module is used to identify the same target based on each target image frame.

[0141] A marking module is used to rasterize each target image frame, mark the grid position information of the target feature points in the target image frame, and / or the grid within the target contour range,

[0142] A calculation module is used to calculate the offset of the target feature point relative to the center of the target image frame based on the grid position information of the target feature point and the grid position information of the center of the target image frame, and / or calculate the target screen ratio based on the area within the target contour range and the area of ​​the target image frame.

[0143] The selection module is used to eliminate target image frames whose target screen ratio is less than a target screen ratio threshold, and retain target image frames whose target screen ratio is not less than the target screen ratio threshold; from the retained target image frames, select the target image frame with the minimum offset and / or the maximum target screen ratio.

[0144] See also Fig.12 As shown, Fig.12 A schematic diagram of a network side device, a control device, or an image recognition device according to an embodiment of the present application, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement any step of the method for reducing false triggering of a multi-channel gate.

[0145] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0146] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0147] An embodiment of the present invention further provides a computer-readable storage medium, wherein a computer program is stored in the storage medium, and when the computer program is executed by a processor, any step of the method for reducing false triggering of a multi-channel gate is implemented.

[0148] As for the apparatus / network-side device / storage medium embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0149] In this article, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0150] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for reducing false triggering of multi-channel gates, characterized in that: The method includes, receiving target image frames from image recognition devices of each channel, Eliminate target image frames whose target screen ratio is less than the target screen ratio threshold, and retain target image frames whose target screen ratio is not less than the target screen ratio threshold; wherein the target screen ratio threshold is determined according to the ratio of the area within the target contour range obtained at the maximum target recognition distance to the total area of ​​the target image frame; When the same target exists in each target image frame, for each target image frame: the target image frame is rasterized, the grid position information of the target feature point in the target image frame is marked, and the absolute value of the difference between the horizontal coordinate of the target feature point and the horizontal coordinate of the center of the target image frame is calculated to obtain the offset; wherein, when the target is a face, the target feature point is the center point between the pupils of both eyes; The offsets of the target image frames are compared, and the channel corresponding to the target image frame with the smallest offset is used as the correct channel for triggering.

2. The method according to claim 1, characterized in that The method further comprises, When the same target exists in each target image frame, the target screen ratio of each target image frame is calculated. The method of comparing the offsets of each target image frame and taking the channel corresponding to the target image frame with the smallest offset as the correct channel for triggering includes: comparing the offsets and target screen ratio of each target image frame and taking the channel corresponding to the target image frame with the smallest offset and the largest target screen ratio as the correct channel for triggering.

3. The method according to claim 1, characterized in that The method further comprises, According to the source address information of the target image frame with the minimum offset, the corresponding image recognition device is notified to perform authority authentication, so that the image recognition device performs a channel triggering operation when the authority authentication passes.

4. The method according to claim 1, characterized in that The receiving target image frame from the image recognition device of each channel further includes: According to the received target image frame, it is determined whether two or more image recognition devices recognize the target. If so, the same target is recognized based on each target image frame. Otherwise, the channel corresponding to the target image frame is used as the correct channel to trigger, and the image recognition device is notified to perform authority authentication.

5. The method according to claim 2, characterized in that The calculating of the target screen ratio of each target image frame includes: Mark the target contour range in the target image frame, and calculate the target screen ratio according to the area within the target contour range and the area of ​​the target image frame.

6. The method according to claim 5, characterized in that The target screen ratio is calculated according to the area within the target contour range and the area of ​​the target image frame, including: Count the number of grids within the target contour range, Count the number of grids in the target image frame, The ratio of the number of grids within the target contour to the number of grids in the target image frame is calculated to obtain the target screen ratio.

7. A multi-channel gate system, comprising at least two image recognition devices, each channel is equipped with an image recognition device, characterized in that: The image recognition device of each channel is connected to the network side device respectively, and each image recognition device sends the acquired target image frame to the network side device, so that the network side device executes the steps of the method for reducing false triggering of multi-channel gates as claimed in any one of claims 1 to 6.

8. The system according to claim 7, characterized in that The network side device includes a cloud platform.

9. A network side device, characterized in that: The device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to implement the steps of any method of reducing false triggering of a multi-channel gate according to any one of claims 1 to 6.

10. A multi-channel gate system, comprising a control device and at least two image recognition devices, each channel is equipped with an image recognition device, characterized in that: The control device includes a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to implement the steps of any method of reducing false triggering of a multi-channel gate according to claims 1 to 6.

11. A multi-channel gate system, comprising at least n image recognition devices, each channel is equipped with an image recognition device, n is a natural number greater than or equal to 2, characterized in that: At least one image recognition device is a designated image recognition device, used to implement the steps of any method of reducing false triggering of a multi-channel gate as claimed in any one of claims 1 to 6.

12. The multi-channel gate system according to claim 11, characterized in that: The designated image recognition device is one of the image recognition devices having intersecting recognition areas.

13. The multi-channel gate system according to claim 12, characterized in that: The number of the designated image recognition devices is n-1.

Images