An electronic password lock system based on motion recognition
By introducing unlocking action analysis, user action recognition and identity verification modules into the action recognition electronic password lock, the problem of imitating lock unlocking and environmental impact is solved, security and recognition accuracy are improved, and the convenience of action correction is provided.
Patent Information
- Application Number
- CN202411311411.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-20
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-20
AI Technical Summary
The electronic password lock for action recognition has safety hazards and errors in recognition when imitating lock unlocking, too fast speed, unstandard action amplitude or external environment.
Adopt the unlocking action analysis module, user action recognition module, secondary recognition analysis module and identity recognition module. By analyzing historical data, correcting action intervals, matching real-time actions, analyzing the causes of abnormalities, and authenticating them through high-definition cameras to generate specific reason information. to improve identification accuracy.
It effectively avoids imitating lock unlocking situations, improves security, helps users correct actions in a timely manner to ensure successful unlocking, and enhances the system's recognition accuracy and user experience.
Smart Images

Figure CN118823909B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of electronic code locks, in particular to an electronic code lock system based on motion recognition. Background Art
[0002] An electronic combination lock is an electronic device that controls the operation of a circuit or chip through password input, thereby controlling the closing of a mechanical switch to complete unlocking and locking tasks. Electronic combination locks include not only digital password types, but also face recognition and motion recognition types.
[0003] According to publication number CN103116918B, an electronic password lock system based on motion recognition is disclosed, which includes a mechanical switch, a single-chip microcomputer for controlling the mechanical switch, a central processor connected to the single-chip microcomputer, a touch screen connected to the central processor, an audio circuit, a camera, and a memory.
[0004] The above invention can decode only by human body making decoding actions, without the need for touch-type human-computer interaction equipment, thus effectively avoiding risks, reducing the possibility of password theft, and improving the security of electronic password locks.
[0005] However, when using motion recognition, on the one hand, compared with traditional password locks, the unlocking action of the motion recognition electronic password lock may be more easily observed and imitated by others. If someone secretly observes the user's action pattern when unlocking the door and imitates it, it is possible to successfully open the door lock, which poses certain safety hazards. On the other hand, if the user moves too fast, with a small amplitude or irregular movements, the system may misjudge and fail to unlock the door normally. At the same time, the external environment has a greater impact on the accuracy of motion recognition, resulting in unrecognizable situations. Summary of the invention
[0006] In view of the shortcomings of the prior art, the present invention provides an electronic password lock system based on motion recognition, which solves the problem that there will be imitation of motion unlocking without identity verification, causing safety hazards, and secondly, the problem that the motion speed is too fast, the motion amplitude is not standard, or the external environment causes the impact on unlocking.
[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an electronic password lock system based on motion recognition, comprising:
[0008] The unlocking action analysis module is used to analyze the stored unlocking information, segment the unlocking actions corresponding to the unlocking information, generate corresponding recognition intervals based on the unlocking actions in the historical data, correct the deviation of the recognition intervals based on the influence of environmental factors, generate correction intervals, and transmit the correction intervals to the user action recognition module;
[0009] A user action recognition module is used to obtain the real-time action of the unlocking user, segment and recognize the real-time action, and match the segmented real-time action with the correction interval to determine the correctness of the unlocking user and generate an unlocking result, wherein the unlocking result includes: a correct matching result and an incorrect matching result, and the incorrect matching result is transmitted to the secondary recognition analysis module, and the correct matching result is transmitted to the recognition information output module;
[0010] The secondary recognition and analysis module is used to analyze the obtained matching error results, identify the abnormal actions, and generate the corresponding abnormal causes from the execution speed and execution range of the abnormal actions, generate specific cause information, and then transmit the specific cause information to the recognition information output module.
[0011] As a further solution of the present invention: it also includes a user identity recognition module and an identification information output module;
[0012] A user identity recognition module is used to obtain the identity information of the unlocking user, authenticate the unlocking user according to the identity information, and generate an identity authentication result, wherein the identity authentication result includes a correct authentication result and an incorrect authentication result, and transmit the correct authentication result to the user action recognition module, and transmit the incorrect authentication result to the identification information output module;
[0013] The identification information output module is used to display the acquired specific reason information to the corresponding unlocking user.
[0014] As a further solution of the present invention: the specific method in which the user identification module authenticates the unlocking user according to the identity information is:
[0015] The facial information of the unlocking user is collected through a high-definition camera, and the collected facial information is verified and matched with the corresponding facial information recorded in the repository. If the collected facial information exists in the repository, a correct verification result is generated. Otherwise, if the collected facial information does not exist in the repository, an incorrect verification result is generated.
[0016] As a further solution of the present invention: the specific manner in which the unlocking action analysis module analyzes the stored unlocking information is:
[0017] Acquire the stored unlocking information, wherein the unlocking information includes: unlocking action, action duration and action amplitude, and acquire one set of unlocking information as the target object, and acquire the unlocking action corresponding to the target object, and segment the unlocking action. The specific segmentation method is: acquire the unlocking action, and acquire the motion trajectory corresponding to the unlocking action, and acquire the motion nodes of the motion trajectory, and then segment the motion trajectory into multiple segments based on the motion nodes as the segmentation standard to obtain multiple groups of motion trajectory segments denoted as i, and i=1, 2, ..., j, wherein j represents the number of motion trajectory segments, and acquire the action amplitude corresponding to the motion trajectory segment i, and then acquire all the unlocking information, and sort out the action amplitude corresponding to the same motion trajectory segment i to obtain the action amplitude interval;
[0018] The acquired motion amplitude intervals are analyzed, and different light intensities are divided into intervals to obtain light division information, and the intensity impact deviations corresponding to different light intensities are obtained. Then, the overall deviation mean is calculated based on the calculated intensity impact deviation, and the deviation mean is substituted into the motion amplitude interval for correction to obtain the corrected interval, and the corrected interval is transmitted to the user motion recognition module.
[0019] As a further solution of the present invention: the specific method in which the user action recognition module acquires and analyzes the real-time action of the unlocking user is:
[0020] The real-time action of the unlocking user is obtained, and the obtained real-time action is segmented to obtain real-time segmented actions, which are labeled as n, and n=1, 2, ..., m, where m represents the number of real-time segmented actions. Then the real-time segmented action n is matched with the correction interval. If the real-time segmented action is the same as the correction interval, the real-time segmented action is recorded as a normal action. If the real-time segmented action is different from the correction interval, the real-time segmented action is recorded as an abnormal action. At the same time, the normal action and abnormal action obtained by classification are verified and analyzed.
[0021] As a further solution of the present invention, the specific method in which the user action recognition module verifies and analyzes normal actions and abnormal actions is as follows:
[0022] If an abnormal action is generated when matching the correction interval, a matching error result is generated. Conversely, if no abnormal action is generated when matching the correction interval, a matching correct result is generated.
[0023] As a further solution of the present invention: the specific method in which the secondary recognition and analysis module analyzes the matching error result is:
[0024] The abnormal action corresponding to the matching error result is obtained, and the action starting point and action end point corresponding to the abnormal action are obtained, and the motion trajectory corresponding to the abnormal action is obtained. Then the motion trajectory is segmented to obtain multiple groups of segments, denoted as a, and a=1, 2, ..., b, where b represents the number of segments. At the same time, the length corresponding to the segment a is obtained, denoted as Da, and the time corresponding to the segment a is obtained, denoted as Ta. Then, according to the length Da and time Ta of the segment a, the trajectory speed of the segment a is calculated, denoted as Va, and the average trajectory speed of all the segment a is calculated, denoted as Vp, and the trajectory speed average Vp is compared with the preset value.
[0025] As a further solution of the present invention, the specific method of comparing the trajectory velocity mean value Vp with the preset value by the secondary recognition and analysis module is:
[0026] If the trajectory speed average Vp is greater than the preset value, it means that the real-time action execution speed of the unlocking user is too high, and speeding information is generated. Conversely, if the speed average Vp is less than the preset value, it means that the real-time action execution speed of the unlocking user is normal, and action amplitude abnormality information is generated.
[0027] The present invention provides an electronic password lock system based on motion recognition. Compared with the prior art, it has the following beneficial effects:
[0028] The present invention collects the identity information of the unlocking user through a camera and verifies the identity, thereby preventing incorrect users from imitating unlocking and improving overall security. Secondly, the unlocking action is analyzed according to past historical data, and a corresponding unlocking action interval is generated according to the historical data and external influences. At the same time, matching analysis is performed based on the real-time actions of the unlocking user. In case of mismatch, the action speed and action amplitude are analyzed to determine the specific cause of the unlocking failure, and the specific cause is displayed to the user, so that the user can correct the action in time. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 This is a block diagram of the system principle of the present invention;
[0030] Figure 2 It is a schematic diagram of distinguishing the action range of the present invention. DETAILED DESCRIPTION
[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0032] For example, see Figure 1 and Figure 2 The present application provides an electronic password lock system based on motion recognition, including: a user identity recognition module, an unlocking motion analysis module, a user motion recognition module, a secondary recognition analysis module and an identification information output module, and the unlocking motion analysis module, the user motion recognition module, the secondary recognition analysis module and the identification information output module are unidirectionally electrically connected.
[0033] The user identity recognition module is used to obtain the identity information of the unlocking user, authenticate the unlocking user according to the identity information, and generate an identity authentication result, wherein the identity authentication result includes a correct authentication result and an incorrect authentication result. At the same time, the correct authentication result is transmitted to the user action recognition module, and the incorrect authentication result is transmitted to the identification information output module. The specific method of authenticating the unlocking user identity is as follows:
[0034] The facial information of the unlocking user is collected through a high-definition camera. During the collection process, the camera will fully consider factors such as different lighting conditions, angle changes, and possible differences in user expressions to ensure that the collected facial information is as complete and accurate as possible. At the same time, the collected facial information is verified and matched with the corresponding recorded facial information in the repository. If the collected facial information exists in the repository, it means that the facial information is verified correctly, and a verification result is generated. On the contrary, if the collected facial information does not exist in the repository, it means that the facial information is verified incorrectly, and a verification error result is generated;
[0035] An identification information output module is used to display the obtained verification error result to the corresponding unlocking user.
[0036] Embodiment 2, as Embodiment 2 of the present invention, is implemented on the basis of Embodiment 1, and differs from Embodiment 1 in that:
[0037] The unlocking action analysis module is used to analyze the stored unlocking information, segment the unlocking actions corresponding to the unlocking information, and generate corresponding recognition intervals in combination with the unlocking actions in the historical data. At the same time, the recognition intervals are corrected for deviations in combination with environmental factors to generate correction intervals, and the correction intervals are transmitted to the user action recognition module. The specific method of generating the correction intervals is as follows:
[0038] The stored unlocking information is obtained, wherein the unlocking information includes: an unlocking action, an action duration, and an action amplitude, and the stored unlocking information here represents all the recorded identifiable unlocking actions, and one group of unlocking information is obtained as a target object, and the unlocking action corresponding to the target object is obtained, and the unlocking action is segmented. The specific segmentation method is: obtain the unlocking action, and obtain the motion trajectory corresponding to the unlocking action, and obtain the motion node of the motion trajectory, and the motion node here represents the node corresponding to the direction change, and then the motion trajectory is segmented into multiple segments based on the motion node as the segmentation standard to obtain multiple groups of motion trajectory segments denoted as i, and i=1, 2, ..., j, where j represents the number of motion trajectory segments, and the action amplitude corresponding to the motion trajectory segment i is obtained, and the action amplitude here can be represented by the coordinates of the motion end point, such as establishing a coordinate system with the center point of the display screen as the origin, as shown in the attached figure. Figure 2 As shown, the starting point of the trajectory can be expressed as (X1, Y1), the end point of the trajectory can be expressed as (X2, Y2), and the corresponding motion amplitude can be expressed by the ordinate height of the end point of the trajectory, thereby obtaining the amplitude size Y2, and then all the unlocking information is obtained, and the motion amplitude corresponding to the same motion trajectory segment i is sorted to obtain the motion amplitude interval, and the same motion trajectory i is represented as the motion trajectory segment corresponding to the same number after multi-segment segmentation, and the motion amplitude interval obtained by sorting is the motion amplitude interval of the motion trajectory segment, which can be specifically represented by Ymin and Ymax;
[0039] The obtained motion amplitude interval is analyzed, and different light intensities are divided into intervals to obtain light division information, and the specific light division information here is expressed as the division intervals under daytime lighting conditions and nighttime lighting conditions, and the intensity impact deviation corresponding to different light intensities is obtained, and then the overall deviation mean is calculated based on the calculated intensity impact deviation, and the deviation mean is substituted into the motion amplitude interval for correction to obtain the correction interval, and the correction interval is transmitted to the user motion recognition module. The specific method of calculating the deviation mean here is to calculate the difference in motion amplitudes recognized by the system under daytime light intensity, and calculate the overall deviation mean, sum the deviation mean and the motion amplitude interval, and so on to calculate the deviation mean under nighttime light intensity and sum it up.
[0040] For example, the standard amplitude interval corresponding to a certain unlocking motion trajectory segment during the day is [10, 20], while the amplitude interval recognized by the system under actual daytime lighting conditions is [12, 22]. At this time, the deviation during the day is (12 - 10) + (22 - 20) = 4. Assuming there are multiple such motion trajectory segments, calculate their average deviation as the mean deviation during the day. Sum the deviation mean and the motion amplitude interval to obtain the corrected motion amplitude interval during the day.
[0041] For example, assuming that under night lighting conditions, the standard amplitude range corresponding to the motion trajectory segment of a certain unlocking action is [8, 18]. After actual testing, the amplitude range recognized by the system at night is [7, 17], and the deviation is (7 - 8) + (17 -18) = -2. If there are multiple similar motion trajectory segments, the average deviation at night is calculated to be -1.5. The average deviation is summed with the standard amplitude range to obtain the corrected night action amplitude range of [6.5, 16.5].
[0042] The user action recognition module is used to obtain the real-time action of the unlocking user, segment and recognize the real-time action, and match the segmented real-time action with the correction interval to determine the correctness of the unlocking user and generate an unlocking result, wherein the unlocking result includes: a correct match result and an incorrect match result. At the same time, the incorrect match result is transmitted to the secondary recognition analysis module, and the correct match result is transmitted to the recognition information output module. The specific method of generating the unlocking result is:
[0043] The real-time action of the unlocking user is obtained, and the obtained real-time action is segmented to obtain real-time segmented actions, which are labeled as n, and n=1, 2, ..., m, where m represents the number of real-time segmented actions. Then, the real-time segmented action n is matched with the correction interval, and the matching here is to match the motion trajectory and action amplitude of the real-time segmented action. If the real-time segmented action is the same as the correction interval, the real-time segmented action is recorded as a normal action. If the real-time segmented action is different from the correction interval, the real-time segmented action is recorded as an abnormal action.
[0044] At the same time, the normal actions and abnormal actions obtained by classification are verified and analyzed. If an abnormal action is generated when matching the correction interval, a matching error result is generated. On the contrary, if an abnormal action is not generated when matching the correction interval, a matching correct result is generated. For example, when analyzing a certain section of the real-time segmentation action, if there is a mismatch with the correction interval, a mismatch result is directly generated. Only when all real-time segmentation actions are successfully matched, a matching correct result is generated.
[0045] The secondary identification and analysis module is used to analyze the obtained matching error results, identify the abnormal action, and generate the corresponding abnormal cause from the execution speed and execution range of the abnormal action, generate specific cause information, and then transmit the specific cause information to the identification information output module. The specific method of generating the specific cause information is:
[0046] The abnormal action corresponding to the matching error result is obtained, and the action start point and action end point corresponding to the abnormal action are obtained, and the motion trajectory corresponding to the abnormal action is obtained. Then the motion trajectory is segmented to obtain multiple groups of segments, denoted as a, and a=1, 2, ..., b, where b represents the number of segments, and the segmentation here is to divide the motion trajectory into small straight line segments, and at the same time, the length corresponding to the segment a is obtained, denoted as Da, and the time corresponding to the segment a is obtained, denoted as Ta, and the time Ta here is the time value of the straight line segment a. At the same time, the time unit needs to be converted for the case where the time unit is different. The final time unit value is s, and then the trajectory speed of the segment a is calculated according to the length Da and time Ta of the segment a, denoted as Va, and the average trajectory speed of all the segment a is calculated, denoted as Vp;
[0047] The calculated trajectory speed mean Vp is compared with the preset value, and the preset value here is obtained by summing up the trajectory speeds corresponding to the historical data and calculating the average value. If the trajectory speed mean Vp is greater than the preset value, it means that the real-time action execution speed of the unlocking user is too high, and speeding information is generated. On the contrary, if the speed mean Vp is less than the preset value, it means that the real-time action execution speed of the unlocking user is normal, and abnormal action amplitude information is generated.
[0048] For example, by counting the trajectory speeds of the past 100 normal actions, the trajectory speeds are v1, v2, ..., v100, and the units are all meters per second (m / s). First, these speed values are summed, that is, ∑vi (i from 1 to 100), and then divided by 100 to obtain the preset value V1. If the trajectory speed mean Vp is greater than the preset value V1, it means that the real-time action execution speed of the unlocking user is too high, and an overspeed information is generated. For example, the preset value V1 is 2m / s. When the calculated trajectory speed mean Vp is 2.5m / s, it means that the user's action speed is too fast. The system generates an overspeed information to remind the user to adjust the action speed to ensure the accuracy and safety of the action. On the contrary, if the speed mean Vp is less than the preset value V1, it means that the real-time action execution speed of the unlocking user is normal, but there may be other problems. At this time, abnormal action amplitude information is generated. For example, when Vp is 1.5m / s, although the speed is within the normal range, the system may generate abnormal motion amplitude information due to reasons such as the motion amplitude not meeting the requirements, prompting the user to check whether the motion is standardized in order to successfully complete the unlocking operation.
[0049] An identification information output module is used to display the acquired specific reason information to the corresponding unlocking user.
[0050] Embodiment 3, as the embodiment 3 of the present invention, focuses on combining the implementation processes of embodiment 1 and embodiment 2 for implementation.
[0051] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.
[0052] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. An electronic password lock system based on motion recognition, characterized in that: include: The unlocking action analysis module is used to analyze the stored unlocking information, segment the unlocking actions corresponding to the unlocking information, and generate corresponding recognition intervals in combination with the unlocking actions in the historical data. At the same time, the recognition intervals are corrected for deviations in combination with environmental factors to generate correction intervals, and the correction intervals are transmitted to the user action recognition module. The specific method for analyzing the stored unlocking information is as follows: Acquire the stored unlocking information, wherein the unlocking information includes: unlocking action, action duration and action amplitude, and acquire one set of unlocking information as the target object, and acquire the unlocking action corresponding to the target object, and segment the unlocking action. The specific segmentation method is: acquire the unlocking action, and acquire the motion trajectory corresponding to the unlocking action, and acquire the motion nodes of the motion trajectory, and then segment the motion trajectory into multiple segments based on the motion nodes as the segmentation standard to obtain multiple groups of motion trajectory segments denoted as i, and i=1, 2, ..., j, wherein j represents the number of motion trajectory segments, and acquire the action amplitude corresponding to the motion trajectory segment i, and then acquire all the unlocking information, and sort out the action amplitude corresponding to the same motion trajectory segment i to obtain the action amplitude interval; The obtained motion amplitude interval is analyzed, and different light intensities are divided into intervals to obtain light division information, and the intensity influence deviation corresponding to different light intensities is obtained. Then, the overall deviation mean is calculated according to the calculated intensity influence deviation, and the deviation mean is substituted into the motion amplitude interval for correction to obtain the correction interval, and the correction interval is transmitted to the user motion recognition module; A user action recognition module is used to obtain the real-time action of the unlocking user, segment and recognize the real-time action, and match the segmented real-time action with the correction interval to determine the correctness of the unlocking user and generate an unlocking result, wherein the unlocking result includes: a correct matching result and an incorrect matching result, and the incorrect matching result is transmitted to the secondary recognition analysis module, and the correct matching result is transmitted to the recognition information output module; The secondary recognition and analysis module is used to analyze the obtained matching error results, identify the abnormal actions, and generate the corresponding abnormal causes from the execution speed and execution range of the abnormal actions, generate specific cause information, and then transmit the specific cause information to the recognition information output module.
2. The electronic password lock system based on motion recognition according to claim 1 is characterized in that: It also includes a user identity recognition module and an identification information output module; A user identity recognition module is used to obtain the identity information of the unlocking user, authenticate the unlocking user according to the identity information, and generate an identity authentication result, wherein the identity authentication result includes a correct authentication result and an incorrect authentication result, and transmit the correct authentication result to the user action recognition module, and transmit the incorrect authentication result to the identification information output module; The identification information output module is used to display the acquired specific reason information to the corresponding unlocking user.
3. The electronic password lock system based on motion recognition according to claim 2 is characterized in that: The specific method in which the user identification module authenticates the unlocking user according to the identity information is as follows: The facial information of the unlocking user is collected through a high-definition camera, and the collected facial information is verified and matched with the corresponding facial information recorded in the repository. If the collected facial information exists in the repository, a correct verification result is generated. Otherwise, if the collected facial information does not exist in the repository, an incorrect verification result is generated.
4. The electronic password lock system based on motion recognition according to claim 1, characterized in that: The specific method for the user action recognition module to acquire and analyze the real-time action of the unlocking user is: The real-time action of the unlocking user is obtained, and the obtained real-time action is segmented to obtain real-time segmented actions, which are labeled as n, and n=1, 2, ..., m, where m represents the number of real-time segmented actions. Then the real-time segmented action n is matched with the correction interval. If the real-time segmented action is the same as the correction interval, the real-time segmented action is recorded as a normal action. If the real-time segmented action is different from the correction interval, the real-time segmented action is recorded as an abnormal action. At the same time, the normal action and abnormal action obtained by classification are verified and analyzed.
5. The electronic password lock system based on motion recognition according to claim 4 is characterized in that: The specific way in which the user action recognition module verifies and analyzes normal actions and abnormal actions is as follows: If an abnormal action is generated when matching the correction interval, a matching error result is generated. Conversely, if no abnormal action is generated when matching the correction interval, a matching correct result is generated.
6. The electronic password lock system based on motion recognition according to claim 1, characterized in that: The specific method in which the secondary recognition and analysis module analyzes the matching error results is as follows: The abnormal action corresponding to the matching error result is obtained, and the action starting point and action end point corresponding to the abnormal action are obtained, and the motion trajectory corresponding to the abnormal action is obtained. Then the motion trajectory is segmented to obtain multiple groups of segments, denoted as a, and a=1, 2, ..., b, where b represents the number of segments. At the same time, the length corresponding to the segment a is obtained, denoted as Da, and the time corresponding to the segment a is obtained, denoted as Ta. Then, according to the length Da and time Ta of the segment a, the trajectory speed of the segment a is calculated, denoted as Va, and the average trajectory speed of all the segment a is calculated, denoted as Vp, and the trajectory speed average Vp is compared with the preset value.
7. The electronic password lock system based on motion recognition according to claim 6, characterized in that: The specific method in which the secondary recognition and analysis module compares the trajectory velocity mean Vp with the preset value is as follows: If the trajectory speed mean Vp is greater than the preset value, it means that the real-time action execution speed of the unlocking user is too high, and a high speed information is generated. Conversely, if the speed mean Vp is less than the preset value, it means that the real-time action execution speed of the unlocking user is normal, and an abnormal action amplitude information is generated.
