Intelligent security method, storage medium and device

Abstract

The present application proposes an intelligent security method, storage medium and device. When a large number of people are detected passing through a security entrance, a mode of collaborative operation of multiple security devices is entered. Specifically, the method comprises the following steps: obtaining a comparison result of the computing power of the first security device and the second security device, and based on the comparison result, dividing the image to be identified into a portion to be identified and monitored by the first security device, and another portion to be identified and monitored by the second security device. In addition, based on the comparison result of the computing power, a target memory is allocated to the first security device, and different verification strategies are set. Through the intelligent security solution proposed in the present application, the resources of each security device can be fully mobilized, and collaborative monitoring can be achieved even if the specifications of the security devices are different; even when the security device configuration is low, the recognition efficiency and accuracy can be improved to meet the needs of scenarios with a large number of people passing through; and because ordinary security devices are used, equipment costs can also be saved.

Description

Technical Field

[0001] The present invention relates to the field of security, and in particular to an intelligent security method, storage medium and device. Background Art

[0002] With the development of the times and the advancement of productivity, people's requirements for security are constantly increasing. Whether it's residential area management or screening personnel entering and exiting a production workshop, security equipment is required. To meet security needs, equipment manufacturers are constantly improving the specifications of security equipment. For example, security equipment with extremely high computing power is used to perform video screening of passers-by to improve security efficiency. Alternatively, users are forced to queue through high-resolution security cameras to accurately identify passers-by, sacrificing their travel time.

[0003] However, either approach significantly increases security costs or severely compromises traffic efficiency, making it difficult to meet enterprise needs. Providing a security solution that is both cost-effective and highly efficient is an urgent issue. Summary of the Invention

[0004] This application proposes an intelligent security method, storage medium and device, which can improve recognition efficiency and reduce security costs.

[0005] In one aspect, the present application provides an intelligent security method, comprising: a first security device receiving headcount information sent by an infrared counter, and determining whether the headcount information indicates a number of people exceeding a first threshold;

[0006] If the threshold is exceeded, the first security device obtains the computing power of the first security device and the computing power of the second security device, compares the computing power of the first security device with the computing power of the second security device, and obtains comparison result information; obtains a target area image corresponding to the comparison result information, and extracts the target object in the target area image, where the size of the target area image is smaller than or equal to the image to be identified captured by the security device;

[0007] Determine a target storage resource that matches the comparison result information, establish a mapping relationship with the target storage resource, extract pattern recognition information from the target storage resource, and determine whether the pattern recognition information matches the target object;

[0008] In the case of no match, the target object is verified according to the verification strategy information corresponding to the comparison result information; if the verification is correct, a prompt message for intercepting the target object is output.

[0009] This application first identifies the number of people passing through the security entrance. If the number of people is too large, it enters a mode of coordinated cooperation among multiple security devices, fully mobilizing the resources of each security device. Even when the security device configuration is low, it can improve recognition efficiency and accuracy, meeting the needs of scenarios with a large number of people passing through. Specifically, this application determines the image area that each security device should recognize and the memory resources that can be called based on the computing power of each security device at the same security entrance. Even if the specifications of each security device are inconsistent, the computing power of each security device is fully utilized. At the same time, different verification strategies are set according to the differences in computing power of each security device, which can improve security recognition efficiency while avoiding misjudgments. In addition, this application uses security devices of ordinary specifications, and can achieve efficient recognition effects without the use of expensive equipment, reducing user costs.

[0010] In a possible implementation, the first security device and the second security device are located on the left and right sides of the security entrance, and the cameras of the first security device and the second security device are both aimed at the security entrance; the infrared counter is set in front of the first security device and the second security device, and is used to obtain the number of passers-by information in advance.

[0011] It's important to note that during quiet periods (e.g., late at night), the two security devices can independently capture individual passerby information and, upon identifying an unusual individual, cross-check their results to improve accuracy. This is because, despite fewer people entering and exiting late at night, the need for more accurate identification is paramount to prevent suspicious individuals from entering the security entrance.

[0012] Obtaining the comparison result information includes: obtaining a first ratio between the local computing power and the total computing power, where the total computing power is the sum of the local computing power and the computing power of the second security device;

[0013] The above-mentioned obtaining of the target area image corresponding to the above-mentioned comparison result information includes: if the above-mentioned first ratio is greater than the first threshold and less than the second threshold, then the image corresponding to the first ratio part in the above-mentioned image to be identified is used as the above-mentioned target area image; if the above-mentioned first ratio is greater than the second threshold, then the above-mentioned image to be identified is used as the above-mentioned target area image.

[0014] After taking the image corresponding to the first ratio part in the above-mentioned image to be identified as the above-mentioned target area image, it also includes: dividing a preset amount of computing power from the above-mentioned local computing power as verification computing power, and sending a first indication information to the above-mentioned second security device, and the above-mentioned first indication information is used to instruct the second security device to identify other area images in the above-mentioned image to be identified except the above-mentioned target area image.

[0015] This application divides the area to be identified according to the computing power of the equipment. The equipment with higher computing power (the first security equipment) undertakes the main computing tasks, and other equipment undertakes auxiliary computing tasks. Even if the specifications of the security equipment are inconsistent, the computing power of each security equipment can still be fully utilized.

[0016] The above-mentioned determination of the target storage resource matching the above-mentioned comparison result information includes: obtaining distribution information of the storage resources, and allocating storage resources with a first ratio from the storage resources indicated by the above-mentioned distribution information as the above-mentioned target storage resource.

[0017] It should be noted that the aforementioned storage resources may be SAN or NAS storage systems. Obtaining storage resource distribution information refers to obtaining the storage resource distribution of the SAN or NAS storage system. Based on the computing power of the security device, different amounts of SAN or NAS resources can be allocated to different security devices. The storage resources are provided to the security devices in the form of storage interfaces.

[0018] The above-mentioned target storage resource includes a first memory interface and a second memory interface, the priority of the above-mentioned first memory interface is higher than the priority of the above-mentioned second memory interface, and the above-mentioned extraction of pattern recognition information from the target storage resource and determination of whether the above-mentioned pattern recognition information matches the above-mentioned target object include: reading the first pattern recognition information from the above-mentioned first memory interface, and if the above-mentioned first pattern recognition information does not match the above-mentioned target object, reading the second pattern recognition information from the above-mentioned second memory interface and determining whether it matches the above-mentioned target object.

[0019] In the solution provided in this application, in order to further improve the recognition efficiency, the memory can be graded, where the memory with a higher reading rate (such as a solid-state drive, RAM memory) has a higher priority. The security equipment can store the higher-priority pattern recognition information in the high-level memory in advance during the idle period, so as to quickly read the pattern recognition information during the busy period.

[0020] The verifying the target object according to the verification strategy information corresponding to the comparison result information includes: if the first ratio is greater than a first threshold and less than a second threshold, the first security device verifies the target object according to the verification computing power;

[0021] If the first ratio is greater than a second threshold, the first security device sends third indication information to the second security device, where the third indication information is used to instruct the second security device to verify the target object.

[0022] If the above-mentioned first ratio is greater than the second threshold, the above-mentioned image to be identified is used as the above-mentioned target area image, including: if the above-mentioned first ratio is greater than the second threshold, the original image captured by the above-mentioned first security device is segmented to obtain the first image; the resolution of the above-mentioned first image is reduced to obtain the above-mentioned image to be identified, and the above-mentioned image to be identified is used as the above-mentioned target area image.

[0023] It should be noted that if the computing power difference between the first and second security devices is too large (the first security device is significantly greater than the second), the first security device will perform the identification and monitoring tasks, while the second security device will perform further verification of unrecognizable targets. To mitigate insufficient computing power on the first security device, the image resolution can be further reduced to reduce computing pressure. Furthermore, while lowering the image resolution may reduce the recognition accuracy of the first security device, the second security device performs the verification function, ensuring overall recognition accuracy.

[0024] In a possible implementation, the method further includes: obtaining historical data indicating that a person has passed through the security entrance, the historical data including a mapping relationship between person information and passing time;

[0025] Predict the probability of each person passing through the above security entrance within the current time period based on the above historical data;

[0026] The pattern recognition information corresponding to the personnel whose passing probability is greater than the third threshold is stored in the storage module corresponding to the above-mentioned first memory interface, and the pattern recognition information corresponding to the personnel whose passing probability is greater than the fourth threshold and less than the third threshold is stored in the storage module corresponding to the above-mentioned second memory interface.

[0027] In the solution of the present application, it is predicted in advance which personnel may pass through the security entrance in the current time period, and the pattern recognition information of the relevant personnel is stored in a higher-level memory in advance so that the security equipment can read it quickly, which can greatly improve the recognition efficiency and avoid excessive consumption of computing power.

[0028] The present application also proposes a computer-readable storage medium, which stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, the processor implements the aforementioned intelligent security method.

[0029] The present application also proposes a memory for intelligent security, which includes a first memory, a second memory and a third memory. The first memory and the second memory are used to store pattern recognition information. The performance of the first memory is higher than that of the second memory. The third memory stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, the processor implements the aforementioned intelligent security method.

[0030] This application also proposes an intelligent security device, comprising:

[0031] a counting module, configured to receive the number of people information sent by the infrared counter and determine whether the number of people indicated by the number of people information exceeds a first threshold;

[0032] A comparison module is configured to obtain the computing power of the local device and the computing power of the second security device when the computing power exceeds the first threshold, compare the computing power of the local device with the computing power of the second security device, and obtain comparison result information;

[0033] an extraction module, configured to obtain a target area image corresponding to the comparison result information, and extract a target object from the target area image, wherein the size of the target area image is smaller than or equal to the image to be identified captured by the security device;

[0034] a matching module, configured to determine a target storage resource that matches the comparison result information, establish a mapping relationship with the target storage resource, extract pattern recognition information from the target storage resource, and determine whether the pattern recognition information matches the target object;

[0035] The verification module is used to verify the target object according to the verification strategy information corresponding to the comparison result information when no match is found; if the verification is correct, output a prompt message to intercept the target object.

[0036] The present application also proposes an intelligent security device, comprising a processor and a memory; wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the aforementioned intelligent security method. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0038] Figure 1 It is a schematic diagram of an intelligent security method proposed in this application;

[0039] Figure 2 This is a schematic diagram of another intelligent security method proposed in this application;

[0040] Figure 3 This is a schematic diagram of another intelligent security method proposed in this application;

[0041] Figure 4 This is a schematic diagram of an intelligent security device proposed in this application;

[0042] Figure 5 This is a schematic diagram of another intelligent security device proposed in this application. DETAILED DESCRIPTION

[0043] The terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements, but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, or apparatus.

[0044] Figure 1 This is a schematic diagram of an intelligent security method proposed in this application. The above method is used for security equipment and specifically includes:

[0045] 101. The first security device receives the number of people information sent by the infrared counter and determines whether the number of people indicated by the number of people information exceeds a first threshold;

[0046] Infrared counters, which include infrared sensors and other components, are located near security entrances to detect the number of people entering. Security entrances include residential entrances, laboratory entrances, production workshop entrances, factory entrances, and other areas that require monitoring.

[0047] The infrared counter obtains the number of people and sends it to the first security device so that the first security device can determine whether the current security entrance is busy. The first security device includes necessary components such as a camera, a processor, and a network module.

[0048] 102. If the value exceeds the first threshold, the first security device obtains the computing power of the first security device and the computing power of the second security device, compares the computing power of the first security device with the computing power of the second security device, and obtains comparison result information;

[0049] Computing power represents the performance of a security device's processor; higher computing power indicates higher processor performance. A first security device can monitor the computing power of its local processor while simultaneously communicating with a second security device to obtain the second device's computing power. The comparison result information can be the ratio of the first security device's local computing power to the second device's computing power, preferably the ratio of the first security device's local computing power to the total computing power (the sum of the computing powers of the first and second security devices). This comparison result information can reflect the difference between the computing power of the first and second security devices.

[0050] In this embodiment, the computing power of the first security device is greater than that of the second security device, and the first security device performs the main computing tasks.

[0051] 103. Obtain a target area image corresponding to the comparison result information, and extract a target object from the target area image, wherein a size of the target area image is smaller than or equal to the image to be identified captured by the security device;

[0052] After capturing the complete image, the first security device can crop the edge of the device (such as other background images that do not involve the security entrance) and retain the image to be identified that covers the security entrance, which can reduce the computing power consumption in the pattern recognition process.

[0053] After determining the comparison results between the first and second security devices, the image to be identified can be segmented to determine the area that the first or second security device should identify. After determining the image of the target area, the first security device can further extract facial information from the image as the target object for subsequent pattern matching.

[0054] In addition, the main computing tasks are generally undertaken by security equipment with stronger computing power. If the comparison result information shows that the computing power differences of the security equipment are too large, the first security equipment can undertake the matching computing tasks, and the second security equipment can undertake the verification tasks.

[0055] 104. Determine a target storage resource that matches the comparison result information, establish a mapping relationship with the target storage resource, extract pattern recognition information from the target storage resource, and determine whether the pattern recognition information matches the target object;

[0056] In order to achieve dynamic allocation of storage resources based on the computing power of the security equipment, the storage of this application can adopt storage systems such as SAN and NAS. The first security device can determine its corresponding target storage resource from the storage system based on the aforementioned comparison result information and establish a communication connection with the target storage resource.

[0057] The target storage resource can provide pattern recognition information to the first security device, and the first security device matches the pattern recognition information with the target object (face information) and determines whether the match is successful.

[0058] The pattern recognition information is pre-stored pattern recognition information of persons allowed to pass through the security entrance. In other words, if the target object matches the pattern recognition information, it means that the target object is a person who can be allowed to pass. The pattern recognition information can specifically be pre-acquired facial feature information.

[0059] 105. If a match fails, verify the target object according to the verification strategy information corresponding to the comparison result information; if the verification is correct, output a prompt message to intercept the target object.

[0060] If the target object does not match the pattern recognition information, it means that the target object is not the information of the person who can be released that has been stored in the pattern library, and the target object should be intercepted.

[0061] In order to save costs, the security equipment selected in this application is of ordinary performance. In order to avoid false interception and improve verification accuracy, the target object can be further verified, that is, to further determine whether the target object is not a person who can be released.

[0062] Since the computing power of multiple security devices may vary to varying degrees, different verification strategy information can be used for different computing power disparities. For example, if the computing power disparity is too large, the security device with lower computing power will be responsible for verification, while the security device with higher computing power will be responsible for pattern matching.

[0063] If the verification is correct but the stored pattern recognition information still cannot match the target object, it indicates that the target object should not be allowed to pass. In this case, a message prompting interception should be output. For example, the image information of the target object can be output to the monitoring room and security personnel can be notified to intercept the target object.

[0064] This application first identifies the number of people passing through the security entrance. If the number of people is too large, it enters a mode of coordinated cooperation among multiple security devices, fully mobilizing the resources of each security device. Even when the security device configuration is low, it can improve recognition efficiency and meet the needs of scenarios with a large number of people. This application determines the image area that each security device should recognize and the memory resources that can be called based on the computing power of each security device at the same security entrance. Even if the specifications of each security device are inconsistent, the computing power of each security device is fully utilized. At the same time, different verification strategies are set according to the differences in computing power of each security device, which can improve security recognition efficiency while avoiding misjudgments.

[0065] Figure 2This is another schematic diagram of an intelligent security method proposed in this application. The above method is a method of the first embodiment ( Figure 1 The method further comprises:

[0066] 201. Obtain a first ratio between the local computing power and the total computing power, where the total computing power is the sum of the local computing power and the computing power of the second security device; if the first ratio is greater than a first threshold and less than a second threshold, use the image corresponding to the first ratio portion of the image to be identified as the target area image; if the first ratio is greater than the second threshold, use the image to be identified as the target area image.

[0067] In addition, the first security device and the second security device are located on the left and right sides of the security entrance, and the cameras of the first security device and the second security device are aimed at the security entrance; the infrared counter is set in front of the first security device and the second security device, and is used to obtain the number of passers-by information in advance.

[0068] It's important to note that during quiet periods (e.g., late at night), the two security devices can independently capture individual passerby information and, upon identifying an unusual individual, cross-check their identification information to improve accuracy. This is because, despite fewer people entering and exiting late at night, the need for higher accuracy in identifying individuals is crucial to prevent suspicious individuals from entering the security entrance during these quiet periods.

[0069] It should also be noted that, in this embodiment, the computing power of the first security device is higher than that of the second security device.

[0070] If the first ratio is greater than the first threshold and less than the second threshold, it means that the computing power difference between the two security devices is still within a reasonable range. At this time, the image corresponding to the first ratio part of the above-mentioned image to be identified is used as the above-mentioned target area image. The target area image is handed over to the first security device for identification and screening, and the remaining images except the target area image are handed over to the second security device for identification and screening.

[0071] If the above-mentioned first ratio is greater than the second threshold, it means that the difference in computing power between the two security devices is too large. At this time, the first security device is responsible for the task of identifying and checking images, and the second security device is responsible for the verification task, so as to fully utilize the computing power of each device.

[0072] The first threshold and the second threshold can be set according to actual needs.

[0073] 202. If the first ratio is greater than the first threshold and less than the second threshold, a preset amount of computing power is divided from the local computing power as verification computing power, and a first indication message is sent to the second security device. The first indication message is used to instruct the second security device to identify other area images in the image to be identified except the target area image.

[0074] If the first ratio is greater than the first threshold and less than the second threshold, it means that the computing power difference between the two security devices is not too large, and both security devices can undertake identification and investigation tasks, but the area identified by the first security device is larger than that of the second security device.

[0075] The first indication information may further include boundary information of the image to be identified, where the boundary information is used to distinguish the target area image in the image to be identified from the image to be identified by the second security device.

[0076] After receiving the first indication information, the second security device performs image recognition on the specific area according to the indication information.

[0077] A preset amount of computing power is divided from the above-mentioned local computing power as verification computing power. For example, a verification thread can be opened and a preset amount of CPU and memory resources are allocated to the verification thread to verify the target object.

[0078] 203. If the above-mentioned first ratio is greater than the first threshold and less than the second threshold, the above-mentioned first security device verifies the above-mentioned target object based on the above-mentioned verification computing power; if the above-mentioned first ratio is greater than the second threshold, the above-mentioned first security device sends a third indication information to the above-mentioned second security device, and the above-mentioned third indication information is used to instruct the above-mentioned second security device to verify the above-mentioned target object.

[0079] If the above-mentioned first ratio is greater than the first threshold (for example, 60%) and less than the second threshold (for example, 80%), the second security device needs to undertake a specific computing task. At this time, the first security device with greater computing power divides the preset amount of computing power for verification.

[0080] In addition, after identifying an unmatched target object, the second security device may also send an image of the target object to the first security device for verification.

[0081] If the first ratio is greater than the second threshold, indicating that the computing power of the first security device is significantly greater than that of the second security device, the verification task is assigned to the second security device, while the first security device performs the identification and investigation tasks. The third instruction information may include an image of the target object and / or information instructing the second security device to capture an image at the current time and conduct its own verification.

[0082] Specifically, the verification method of the first security device and / or the second security device can be: recapture the image information of the target object and perform a homography transformation to obtain an adjusted target object; confirm whether the adjusted target object is consistent with the original target object (i.e., the target object that was initially matched); if not, continue to match the adjusted target object with the pattern recognition information; if they are consistent, output a prompt message to intercept the target object.

[0083] Homography can adjust the image's angle, reconstructing the image from a side view to a frontal view. A possible cause of target object matching failure is image distortion caused by the shooting angle. If reshooting and adjusting the image angle reveals no substantial difference between the before and after images of the target object, this indicates that the matching failure was not caused by image distortion, but rather by the target object illegally passing through the security entrance, and interception should be carried out.

[0084] Furthermore, if the first ratio is greater than a second threshold, the original image captured by the first security device is segmented to obtain a first image; the resolution of the first image is reduced to obtain the image to be identified, and the image to be identified is used as the target area image.

[0085] Because the first ratio is greater than the second threshold, the first security device performs all calculations and troubleshooting tasks. To mitigate insufficient computing power on the first security device, the image resolution can be further reduced to alleviate computing pressure. Although lowering the image resolution may reduce the first security device's recognition accuracy, the second security device acts as a verification device, ensuring overall recognition accuracy.

[0086] The method of reducing the resolution may be to sample the first image.

[0087] 204. Obtain distribution information of storage resources, and divide storage resources with a first ratio from the storage resources indicated by the distribution information as the target storage resources; wherein the target storage resources include a first memory interface and a second memory interface, and the priority of the first memory interface is higher than the priority of the second memory interface.

[0088] The first pattern recognition information is read from the first memory interface. If the first pattern recognition information does not match the target object, the second pattern recognition information is read from the second memory interface to determine whether it matches the target object.

[0089] In this embodiment, storage resources are divided according to the differences in computing power of security equipment. Security equipment with greater computing power corresponds to more storage resources, and security equipment with less computing power corresponds to fewer storage resources. This helps to reasonably allocate storage resources and provides the best auxiliary measures for security equipment.

[0090] This embodiment also tiers memory, with memories with higher read rates (e.g., solid-state drives and RAM) receiving higher priority. The security device can pre-store high-priority pattern recognition information in high-level memory during idle periods, allowing for faster access during busy periods. The second-highest-priority pattern recognition information is stored in the next-highest-level memory (i.e., the memory corresponding to the second memory interface). This next-highest-level memory may be less frequently used and have a faster read rate.

[0091] After obtaining the target object, the first security device will prioritize reading pattern recognition information from a higher-priority memory for matching. Once a match is successful, the current match is terminated; if the match fails, the pattern recognition information will continue to be read from memories at other levels. The reason for this is that the computing power of security devices is limited. If pattern recognition information is read directly from all memories in sequence, a large amount of computing power will be wasted on pattern matching, while there are often only a few objects to be intercepted, making this a waste of computing power. Directly reading higher-priority pattern recognition data from higher-priority memories increases the probability of a successful match, greatly reducing the computing power consumption of security devices and having important implications for the coordination of multiple devices.

[0092] Furthermore, the storage resources of this application can be NAS or SAN storage resources, which can dynamically allocate the quota for each security device. Obtaining storage resource distribution information can be used to obtain the distribution of each storage device in the NAS or SAN system. The NAS or SAN system can be composed of devices such as solid-state drives and mechanical hard drives.

[0093] 205. Obtain historical data representing personnel passing through the security entrance, the historical data including a mapping relationship between personnel information and passing time; predict the passing probability of each person passing through the security entrance in the current time period based on the historical data; and store pattern recognition information of personnel with different passing probabilities in memories with different priorities.

[0094] Specifically, the pattern recognition information corresponding to the personnel whose passing probability is greater than the third threshold is stored in the storage module corresponding to the above-mentioned first memory interface, and the pattern recognition information corresponding to the personnel whose passing probability is greater than the fourth threshold and less than the third threshold is stored in the storage module corresponding to the above-mentioned second memory interface.

[0095] For example, a mapping relationship between each resident's information stored in the cloud and the time they passed through the security entrance is obtained. Based on this mapping relationship, the probability of each resident passing through the security entrance between 6:00 PM and 7:00 PM is calculated. If the probability of passing through is greater than a third threshold (e.g., 50%), the person's pattern recognition information is stored in the highest-priority storage device (e.g., a solid-state drive). If the probability of passing through is greater than a fourth threshold (e.g., 30%) but less than the third threshold, the person's pattern recognition information is stored in the next-highest-priority storage device (e.g., the most recently purchased memory device).

[0096] In this way, security equipment can quickly read pattern recognition information that matches the target object from a specific memory in each time period, saving computing power while improving recognition efficiency.

[0097] Figure 3 This is a schematic diagram of another intelligent security method proposed in this application, which includes:

[0098] 301. The first security device receives the number of people pre-detected by the infrared counter and determines that the number of people passing through the security entrance is greater than 5.

[0099] 302. The first security device obtains the computing power of its processor and the computing power of the second security device, and determines that the ratio of the computing power of the first security device to the total computing power (the sum of the computing power of the two security devices) is 70%;

[0100] 303. The first security device captures an original image, crops the original image to obtain a first image, samples the first image to obtain an image to be identified, and uses 70% of the image to be identified as a target area image.

[0101] 304. Establish a communication connection with a target storage, where the target storage capacity accounts for 70% of the total storage capacity; the target storage includes a solid-state drive portion, a newly purchased mechanical hard drive portion, and an aging mechanical hard drive portion.

[0102] 305. The first security device instructs the target storage to download pattern recognition information of people who may pass through during the current time period of 18:00-19:00 from the cloud, and stores the part with a higher probability of passing through in the solid state hard disk.

[0103] 306. Extract facial information from the target area image, extract pattern recognition information from the target memory, and match the two; wherein, pattern recognition information is preferentially extracted from the solid state drive portion.

[0104] 307. If the match fails, it means that the person corresponding to the facial information does not have the authority to pass through the security entrance. In order to avoid interception errors, the facial information is verified through the verification thread of the first security device; after the verification is correct, it is further confirmed that the person corresponding to the facial information does not have the authority to pass through the security entrance. At this time, information prompting interception of the target object is output.

[0105] Figure 4 This is a schematic diagram of another intelligent security device proposed in this application, including:

[0106] The counting module 401 is configured to receive the number of people information sent by the infrared counter and determine whether the number of people indicated by the number of people information exceeds a first threshold;

[0107] Comparison module 402, configured to obtain the computing power of the local device and the computing power of the second security device when the computing power exceeds the first threshold, compare the computing power of the local device with the computing power of the second security device, and obtain comparison result information;

[0108] An extraction module 403 is configured to obtain a target area image corresponding to the comparison result information and extract a target object from the target area image, wherein the size of the target area image is smaller than or equal to the image to be identified captured by the security device;

[0109] Matching module 404 is used to determine the target storage resource that matches the comparison result information, establish a mapping relationship with the target storage resource, extract pattern recognition information from the target storage resource, and determine whether the pattern recognition information matches the target object;

[0110] The verification module 405 is used to verify the target object according to the verification strategy information corresponding to the comparison result information when no match is found; and output prompt information of the target object when the verification is correct.

[0111] Figure 5 Schematic diagram of an intelligent security device provided by an embodiment of the present application. The device includes: at least one processor 501, such as a central processing unit (CPU), at least one memory 502, and at least one bus 503.

[0112] The memory 502 may store program instructions, and the processor 501 may be used to call the program instructions to execute the aforementioned intelligent security method.

[0113] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by a program instructing related hardware. The program may be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), a solid state disk (SSD), or other optical disk storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

Claims

1. An intelligent security method, characterized in that: include: The first security device receives the number of people information sent by the infrared counter, and determines whether the number of people indicated by the number of people information exceeds a first threshold; If the value exceeds the first threshold, the first security device obtains the computing power of the local device and the computing power of the second security device, compares the computing power of the local device with the computing power of the second security device, and obtains comparison result information; Obtaining a target area image corresponding to the comparison result information, and extracting a target object in the target area image, wherein a size of the target area image is smaller than or equal to the image to be identified captured by the security device; Determining a target storage resource that matches the comparison result information, establishing a mapping relationship between a target object in the target area image and the target storage resource, extracting pattern recognition information from the target storage resource, and determining whether the pattern recognition information matches the target object; If no match is found, verifying the target object according to the verification strategy information corresponding to the comparison result information; If the verification is correct, output the prompt information of intercepting the target object; Wherein, obtaining the comparison result information includes: determining that the computing power of the local device is greater than the computing power of the second security device; obtaining a first ratio between the computing power of the local device and the total computing power; Wherein, obtaining the target area image corresponding to the comparison result information includes: If the first ratio is greater than a first threshold value and less than a second threshold value, the image corresponding to the first ratio portion in the image to be identified is used as the target area image; If the first ratio is greater than a second threshold, the image to be identified is used as the target area image.

2. The method according to claim 1, characterized in that The first security device and the second security device are located on the left and right sides of the security entrance, and the cameras of the first security device and the second security device are both aimed at the security entrance; the infrared counter is set in front of the first security device and the second security device, and is used to obtain the number of passers-by information in advance.

3. The method according to claim 2, characterized in that The total computing power is the sum of the computing power of the local device and the computing power of the second security device.

4. The method according to claim 3, characterized in that After taking the image corresponding to the first ratio portion in the image to be identified as the target area image, the method further includes: A preset amount of computing power is divided from the local computing power as verification computing power, and a first indication message is sent to the second security device, where the first indication message is used to instruct the second security device to identify other area images in the image to be identified except the target area image.

5. The method according to claim 4, characterized in that: The determining of the target storage resource matching the comparison result information includes: The distribution information of the storage resources is obtained, and the storage resources with a proportion of a first ratio are divided from the storage resources indicated by the distribution information as the target storage resources.

6. The method according to claim 5, characterized in that The target storage resource includes a first storage interface and a second storage interface, the priority of the first storage interface is higher than the priority of the second storage interface, and extracting pattern recognition information from the target storage resource and determining whether the pattern recognition information matches the target object includes: First pattern recognition information is read from the first memory interface. If the first pattern recognition information does not match the target object, second pattern recognition information is read from the second memory interface to determine whether it matches the target object.

7. The method according to claim 4, characterized in that: Verifying the target object according to verification strategy information corresponding to the comparison result information includes: If the first ratio is greater than a first threshold and less than a second threshold, the first security device verifies the target object according to the verification computing power; If the first ratio is greater than a second threshold, the first security device sends third indication information to the second security device, where the third indication information is used to instruct the second security device to verify the target object.

8. The method according to any one of claims 3 to 7, characterized in that: If the first ratio is greater than a second threshold, taking the image to be identified as the target area image includes: If the first ratio is greater than a second threshold, segmenting the original image captured by the first security device to obtain a first image; The resolution of the first image is reduced to obtain the image to be recognized, and the image to be recognized is used as the target area image.

9. The method according to claim 6, characterized in that: Also includes: Acquire historical data indicating personnel passing through a security entrance, the historical data including a mapping relationship between personnel information and passing time; Predicting the probability of each person passing through the security entrance within the current time period based on the historical data; The pattern recognition information corresponding to the persons whose passing probability is greater than the third threshold is stored in the storage module corresponding to the first memory interface, and the pattern recognition information corresponding to the persons whose passing probability is greater than the fourth threshold and less than the third threshold is stored in the storage module corresponding to the second memory interface.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor implements the method according to any one of claims 1 to 9.

Images