CN115050129B - Data processing method and system for intelligent access control - Google Patents
Data processing method and system for intelligent access control Download PDFInfo
- Publication number
- CN115050129B CN115050129B CN202210745554.9A CN202210745554A CN115050129B CN 115050129 B CN115050129 B CN 115050129B CN 202210745554 A CN202210745554 A CN 202210745554A CN 115050129 B CN115050129 B CN 115050129B
- Authority
- CN
- China
- Prior art keywords
- face frame
- frame position
- value
- queue
- historical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 27
- 238000012545 processing Methods 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000005764 inhibitory process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000007873 sieving Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/30—Individual registration on entry or exit not involving the use of a pass
- G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
- G07C9/37—Individual registration on entry or exit not involving the use of a pass in combination with an identity check using biometric data, e.g. fingerprints, iris scans or voice recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The application relates to a data processing method and a system of an intelligent access control, which relate to the field of access control systems and comprise the steps of acquiring the position of a face frame; determining new face frame positions and IOU values of target face frame positions in a historical face frame position queue, wherein the target face frame positions are generated based on the face frame position queue; comparing the IOU value with a preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, moving the new face frame position to a historical face frame position queue; and calculating the target face frame position in the historical face frame position queue, and taking the result as the final face frame position. The face recognition method and device have the advantages that the possibility of jitter of the face frame is reduced, and the accuracy in face recognition is improved.
Description
Technical Field
The application relates to the field of access control systems, in particular to a data processing method and system for an intelligent access control.
Background
Currently, access control refers to the forbidden authority of a door, and is a guard against the door. The entrance and exit entrance guard safety management system is a novel modern safety management system, integrates microcomputer automatic identification technology and modern safety management measures, relates to a plurality of new technologies such as electronics, machinery, optics, computer technology, communication technology, biotechnology and the like, and is an effective measure for realizing safety precaution management of an important department entrance. The access control system is suitable for various confidential departments, such as banks, hotels, machine rooms, ordnance libraries, confidential rooms, offices, intelligent communities, factories and the like.
In addition, common access control systems comprise a password access control system, a non-contact card access control system, a fingerprint iris palm type biological identification access control system, a face recognition access control system and the like, and in the face recognition access control system, face data are generally collected by adopting a face recognition technology based on a deep convolutional neural network.
However, the nature of face recognition is target detection, and the following problems currently exist: when the face recognition engine detects a video stream, the frame position of the same face in the video often shakes, so that the face frame output by face detection is difficult to stabilize and continue, and the accuracy of face recognition is easy to influence, so that improvement is needed.
Disclosure of Invention
In order to solve the problem that face recognition accuracy is low due to the fact that face frames are prone to shaking when face recognition is carried out on an intelligent face guard in the related art, the application provides a data processing method and system of the intelligent face guard.
The data processing method and system for the intelligent access control adopt the following technical scheme:
a data processing method of intelligent gate inhibition is applied to an intelligent gate inhibition system, and comprises the following steps of;
acquiring the current face frame position;
determining IOU values of a target face frame position in the current face frame position and historical face frame position queues, wherein the target face frame position is generated based on the historical face frame position queues;
comparing the IOU value with the preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, moving the face frame position to a historical face frame position queue to form a current face frame position queue;
and calculating the target face frame position in the current face frame position queue, and taking the result as the final face frame position.
By adopting the technical scheme, when the face information in the video stream is detected by adopting the data processing method of the intelligent access control, the queue of the positions of the face frames can be firstly obtained, namely, a plurality of face frame positions meeting the conditions are obtained, wherein the face frame positions are the left margin, the upper margin, the right margin and the lower margin of the face frame in the video and the video boundary; and the method comprises the steps of putting the target face position into a historical face position queue, processing position data of the face in the queue, solving the target face position, wherein reference data of the target face position can be average, median or mode of all face position queues in the position data queue of the face, and then according to IOU values of the new face position and the target face position in the historical face position queue through calculation, when the IOU values are larger than or equal to a threshold value, namely the jitter amplitude of the new face position is smaller than that of the historical face position, at the moment, merging the new face position into the historical face position queue to become the current face position queue and the target face position in the historical face position queue again, and when the face position in the historical face position queue is required to be used, selecting the target face position in use to determine the face position, so that the possibility of the face is reduced, and the accuracy of face recognition is improved.
Preferably, the number of data stored in the historical face frame position queue is a fixed value; the method further comprises the steps of:
and when the historical face frame position queue is fully loaded and the face frame position is moved to the historical face frame position queue, removing the face frame position with the longest time interval with the face frame position from the historical face frame position queue.
By adopting the technical scheme, the number of the face frame positions in the historical face frame position queue is set, so that when the face frame positions are added into the historical face frame position queue, on one hand, the buffer memory in the historical face frame position queue is smaller, the possibility of overhigh load of a processor caused by excessive face frame positions in the historical face frame position queue is reduced, and on the other hand, the influence of the data with the earliest data for a long time from the current time on the target face frame positions in the historical face frame position queue can be reduced, so that the target face frame positions in the historical face frame position queue are more accurate, and the accuracy in face recognition is further improved.
Preferably, after the final face frame position is identified, the current face frame position queue is emptied.
Through adopting above-mentioned technical scheme, after external equipment accomplishes face frame location discernment, empty initial face frame position parameter and historical face frame position queue to make processing unit's buffer memory less, so that follow-up face identification, and empty historical face frame position queue after discernment a man-hour, so that be convenient for reacquire the position of face when carrying out face identification scanning to the individual according to individual's custom, conditions such as height, promoted the accuracy in the face identification.
Preferably, the number of the face frame positions stored in the initial face frame position parameter is greater than or equal to three.
By adopting the technical scheme, when three or more face frame positions are determined on the one hand, the third face frame position can be screened according to two arbitrarily acquired face frame positions, and the accuracy in face recognition can be improved by continuously adjusting the initial face frame positions.
Preferably, in acquiring the initial face frame position parameter, the method further comprises the following steps:
acquiring a first face frame position and a second face frame position;
acquiring a third face frame position;
calculating IOU values of the third face frame position and the first face frame position, and recording the IOU values as first repeated values;
calculating IOU values of the third face frame position and the second face frame position, and recording the IOU values as second repeated values;
comparing the first repeated value with a preset overlapping threshold value and comparing the second repeated value with the preset overlapping threshold value;
and when the first repeated value and the second repeated value are both greater than or equal to the overlapping threshold value, moving the third face frame position to the historical face frame position queue.
Through adopting the technical scheme, the processor firstly acquires the first face frame position and the second face frame position, then compares the third face frame position with the first face frame data and the second face frame data respectively to realize the initial face frame position parameters, screens out the face frame position with larger jitter amplitude in the first face frame position, the second face frame position and the third face frame position, and supplements in the new face frame position, and after repeating the steps for a plurality of times, determines the initial face frame position parameters, and takes the initial face frame position in the initial face frame position parameters as the initial three values of the historical face frame position queue, so as to calculate the target face frame position and acquire the subsequent face frame position, thereby improving the accuracy degree of the target face frame position of the historical face frame position queue.
Preferably, when only one of the first repetition value and the second repetition value is greater than or equal to the overlap threshold, the method includes the following steps;
moving the third face frame position to a historical face frame position queue;
removing the face frame positions with corresponding IOU values smaller than the overlapping threshold values in the historical face frame position queues;
acquiring a new face frame position;
comparing the IOU value of the existing face frame position in the new face frame position and the historical face frame position queue;
and adjusting the face frame positions in the historical face frame position queue according to the comparison result until the number of data stored in the historical face frame position queue meets the preset number.
By adopting the technical scheme, when only one of the first repetition value and the second repetition value is greater than or equal to the overlapping threshold value, the corresponding face frame position smaller than the overlapping threshold value is screened out, a new face frame position is introduced, the IOU value is obtained for the new face frame position and the face frame position in the screened initial face frame position parameter, at the moment, the steps in the claims 5-6 are repeated until the face frame position in the initial face frame position reaches three, and the accuracy in determining the face frame position is improved.
Preferably, when the first repetition value and the second repetition value are both smaller than the overlap threshold, the method includes the following steps;
acquiring an introduction time parameter of the second face frame position;
acquiring an introduction time parameter of a third face frame position;
calculating a difference value between the introduction time parameter of the third face frame position and the introduction time parameter of the second face frame data, and marking the difference value as a first time difference value;
comparing the first time difference value with a preset time limit threshold;
and when the first time difference value is smaller than the time limit threshold value, removing the third face frame position, acquiring a new face frame position and processing the new face frame position in the same way as the third face frame position.
By adopting the technical scheme, when the first repetition value and the second repetition value are smaller than the overlapping threshold value, the time for acquiring the face frame position each time is read, the newly introduced face frame position is compared with the previous face frame position, if the difference value of the newly introduced face frame position and the previous face frame position is smaller than the time limit threshold value, the newly introduced face frame position is screened out, the face frame position is continuously introduced, and the accuracy of the face frame when being used for face recognition scanning is improved.
Preferably, when the first time difference value is greater than the time limit threshold, removing the first face frame position and the second face frame position, and moving the original third face frame position to be used as a new first face frame position in the initial face frame position parameter; and acquiring the position of the second face frame, and processing the initial face frame position parameters again.
By adopting the technical scheme, when the first time difference value is larger than the time limit threshold value, the newly introduced face frame position is reserved, and the original face frame position is removed from the initial face frame position parameter, so that the influence of overlong data of the reserved time on the calculation of the accuracy degree of the face frame position is reduced.
Preferably, an intelligent access control device includes:
the intelligent access control device comprises an acquisition unit, a determination unit, a mobile unit and a calculation unit; wherein,,
the acquisition unit is used for acquiring the current face frame position;
the determining unit is used for determining the current face frame position and the IOU value of a target face frame position in the historical face frame position queue, and the target face frame position is generated based on the historical face frame position queue;
the mobile unit is used for comparing the IOU value with the preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, moving the face frame position into the historical face frame position queue to form a current face frame position queue;
the calculating unit is used for calculating the target face frame position in the current face frame position queue and taking the result as the final face frame position.
By adopting the technical scheme, the acquisition unit acquires the face in the form of the face frame, then the determination unit determines the IOU value of the current face frame position and the target face frame position in the historical face frame position queue, the mobile unit compares the IOU value with the preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, the face frame position is moved to the historical face frame position queue to form the current face frame position queue; and finally, calculating the target face frame position in the current face frame position queue through a calculation unit, and taking the result as the final face frame position, thereby reducing the possibility of jitter of the face frame and improving the accuracy in face recognition.
A third aspect of the present application provides an electronic device comprising a processor, a memory for storing instructions, and a transceiver for communicating with other devices, the processor for executing the instructions stored in the memory, such that an electronic device unit performs a method according to any of the first aspects of the present application.
In summary, the present application includes at least one of the following beneficial technical effects:
1. when the data processing method of the intelligent access control is adopted to detect the face information in the video stream, the queue of the positions of the face frames can be firstly obtained, namely, a plurality of face frame positions meeting the conditions are obtained, wherein the face frame positions are the left margin, the upper margin, the right margin and the lower margin of the face frames in the video and the video boundary; and the method comprises the steps of putting the target face position into a historical face position queue, processing position data of the face in the queue, solving the target face position, wherein reference data of the target face position can be average, median or mode of all face position queues in the position data queue of the face, and then according to IOU values of the new face position and the target face position in the historical face position queue through calculation, when the IOU values are larger than or equal to a threshold value, namely the jitter amplitude of the new face position is smaller than that of the historical face position, at the moment, merging the new face position into the historical face position queue to become the current face position queue and the target face position in the historical face position queue again, and when the face position in the historical face position queue is required to be used, selecting the target face position in use to determine the face position, so that the possibility of the face is reduced, and the accuracy of face recognition is improved.
2. The processor acquires the first face frame position and the second face frame position firstly, then compares the third face frame position with the first face frame data and the second face frame data respectively to realize the calculation of the target face frame position and the acquisition of the subsequent face frame position by sieving out the face frame position with larger jitter amplitude in the first face frame position, the second face frame position and the third face frame position and supplementing the new face frame position, and after repeating the steps for a plurality of times, the initial face frame position parameter is determined, and the initial face frame position in the initial face frame position parameter is used as the initial three values of the historical face frame position queue so as to calculate the target face frame position and acquire the subsequent face frame position, thereby improving the accuracy degree of the target face frame position of the historical face frame position queue.
3. The method comprises the steps that an acquisition unit acquires a human face in a human face frame mode, then a determination unit determines the IOU value of a current human face frame position and a target human face frame position in a historical human face frame position queue, a mobile unit compares the IOU value with the preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, the human face frame position is moved to the historical human face frame position queue to form the current human face frame position queue; and finally, calculating the target face frame position in the current face frame position queue through a calculation unit, and taking the result as the final face frame position, thereby reducing the possibility of jitter of the face frame and improving the accuracy in face recognition.
Drawings
Fig. 1 is a schematic diagram of a logic processing unit in the present application when acquiring an initial face frame position.
FIG. 2 is a schematic diagram of a select data process in the present application.
Fig. 3 is a schematic diagram of a data processing procedure provided in the present application.
Fig. 4 is a flow chart of the data processing provided in the present application.
Fig. 5 is a schematic diagram of the overall structure of an intelligent access control system in the present application.
Reference numerals illustrate: 1. an acquisition unit; 2. a determination unit; 3. a mobile unit; 4. and a calculation unit.
Detailed Description
The present application is described in further detail below in conjunction with figures 1-5. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of embodiments of the present application, words such as "exemplary," "such as" or "for example," are used to indicate by way of example, illustration, or description. Any embodiment or design described herein as "illustrative," "such as" or "for example" is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "illustratively," "such as" or "for example," etc., is intended to present related concepts in a concrete fashion.
In the description of the embodiments of the present application, the term "and/or" is merely an association relationship describing an association object, and indicates that three relationships may exist, for example, a and/or B may indicate: a alone, B alone, and both A and B. In addition, unless otherwise indicated, the term "plurality" means two or more. For example, a plurality of systems means two or more systems, and a plurality of screen terminals means two or more screen terminals. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating an indicated technical feature. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.
The following is an explanation of some nouns referred to in this application:
face frame position: the human face frame is located in the video boundary and is located at the left margin, the right margin, the upper margin and the lower margin of the video boundary;
historical face frame position queues: a queue composed of face frame positions meeting the conditions;
IOU value, namely the coincidence ratio of two face frames;
the overlapping threshold is a preset coincidence rate threshold of two face frames and can be manually adjusted;
the embodiment of the application discloses an intelligent access control system and a processing mode of intelligent access control data, wherein the processing mode comprises the following steps of
In the process of processing the data of the intelligent access control system, firstly, a historical face frame position queue is established, a plurality of first items of the historical face frame position queue are defined as initial face type position parameters, wherein the initial face frame position parameters comprise three or more face frame positions, three face frame positions are selected in the application, but in actual use, four or more face frame positions are allowed and do not influence the access control system in the application, the number of face frame positions stored in the historical face frame position queue is larger than the number of initial face frame positions, and the capacity of the face frame positions in the historical face frame position queue is 180 in the actual application of the access control system in the application;
the initial face frame position parameters and the historical face frame position queues are in one-to-one correspondence, the number of the initial face frame position parameters and the historical face frame position queues are related to the number of faces in the face image information scanned by the logic processing unit, and specifically, if the logic processing unit scans that three faces exist in the face image information, the logic processing unit establishes three historical face frame position queues at the moment, and the initial face frame position parameters and the historical face frame position queues corresponding to each face work independently with the initial face frame position parameters and the historical face frame position queues corresponding to other faces without affecting each other.
After the establishment of the historical face frame position queue is completed, acquiring initial face frame position parameters;
referring to fig. 1, when initial face frame position parameters are acquired, the number of acquired initial face frame positions in use of the intelligent access control system in the application is 3; when the initial face frame position is acquired, the method comprises the steps of S200-S203;
s200, acquiring a first face frame position and a second face frame position;
the first face frame position and the second face frame position are face frame positions in the two frames of face image information which are preferentially received by the logic processing unit.
S201, acquiring a third face frame position;
s202, calculating IOU values of a third face frame position and a first face frame position and IOU values of the third face frame and a second face frame position, and recording the IOU values as a first repeated value and a second repeated value respectively, wherein the IOU values are the ratio of the overlapping part of two face frames to the total occupied area of the two face frames;
s203, comparing the first repeated value and the second repeated value with preset overlapping thresholds respectively, wherein the overlapping thresholds are preset values of staff, the range of the overlapping thresholds is 0.85-0.95 in the application, the larger the number of the overlapping thresholds is, the smaller the jitter amplitude of the face frame is, the smoother the face frame moves, and the difficulty in collecting the position of the face frame is increased; in the test link, the optimal value of the overlapping threshold value is 0.92, and the overlapping threshold value with the highest use frequency is 0.89 after the overlapping threshold value is continuously and finely adjusted in the actual working process.
Specifically, if the first repetition value and the second repetition value are both greater than or equal to the preset overlap threshold, then it is indicated that the variation amplitude of the third face frame position relative to the first face frame position and the second face frame position is smaller at this time, so that the jitter amplitude of the third face frame is smaller, and the third face frame position can be added to the initial face frame position parameter, and then the initial face frame position parameter is { the first face frame position, the second face frame position, the third face frame position }.
In another embodiment, if one of the first repetition value and the second repetition value is greater than or equal to a preset overlap threshold, selecting a case where the first repetition value is greater than or equal to the preset overlap threshold, indicating that the change amplitude of the third face frame position relative to the first face frame position is smaller, so that the change amplitude of the second third face frame relative to the second face frame position is larger, retaining the third face frame position, removing the second face frame position, merging the third face frame position into an initial face frame position parameter, replacing the position of the original second face frame position, introducing a fourth face frame position, repeating the steps, namely comparing the fourth face frame position with the first face frame position and a new second face frame position (namely the original third face frame position), and still referring to the steps, if the fourth face frame position meets the IOU value with the first face frame position and the new second face frame position and is greater than the preset overlap threshold, determining that the fourth face frame position is the first face position and the fourth face position; when the second repetition value is greater than or equal to the preset overlap threshold, the selection method of the face frame data is consistent with the above, so that details are not repeated herein, but in the step of removing the IOU value less than the preset overlap threshold, if the first face frame position is removed, the original second face frame position replaces the position of the first face frame position, and the third face frame position replaces the position of the second face frame position, namely, under the condition that the fourth face frame position meets the condition, the initial face frame position parameter is { the second face frame position, the third face frame position and the fourth face frame position };
in still another embodiment, referring to fig. 2, if the first repetition value and the second repetition value are both smaller than the predetermined overlapping threshold, steps S300-S303 are included;
s300, acquiring an introduction time parameter of the second face frame position, wherein the introduction time parameter is the time point of acquiring the second face frame position, and is accurate to millisecond;
s301, acquiring an introduction time parameter of a third face frame position;
s302, calculating a difference value between an introduction time parameter of the third face frame position and an introduction time parameter of the second face frame data, and marking the difference value as a first time difference value;
s303, comparing the first time difference value with a preset time limit threshold value, and determining initial face frame position parameters according to a comparison result.
Specifically, a worker can adjust a preset time limit threshold through an external information interaction unit, and in the intelligent access control system in the application, the set length of the time limit threshold is higher than the sum of time intervals obtained by two frames of pictures; the time limit threshold in the application is set to be 1s in actual use; and when the first time difference value is smaller than a preset time limit threshold, removing the third face frame position, introducing fourth face frame data, and similarly obtaining an introduction time parameter of the fourth face frame data, repeating the steps until an IOU value between the first face frame position and the second face frame position is larger than or equal to a preset overlapping threshold and is recorded as an N-th face frame position within a time limit range when the time difference value is smaller than the preset time limit range, obtaining an M+1th face frame position, repeating the steps until the M+1th face frame position reaches the initial position (the M+1th face frame position, and repeatedly setting the M+1th face frame position in the S+200-th face frame position) if the time difference value is larger than the preset time limit range, and the M-th face frame position and the IOU value between the first face frame position and the second face frame position are still smaller than the preset overlapping threshold, and the M+1th face frame position is obtained, and the M+1th face frame position is repeatedly set in the initial position and the S+2nd face frame position is the initial position (the M+2-th face position is the first face position and the M+2 face frame position is the initial position and the 2-face position is the first face position and the 2 face position is the initial position and the 2 face position is judged).
The number of data stored in a historical face frame position queue of the access control system is 180, after initial face frame position parameters are obtained, the first 3 face frame position queues in the historical face frame position queue are determined, namely the initial face frame position parameters are { A-th face frame position, B-th face frame position and C-th face frame position }, and the historical face frame position queues are { A-th face frame position, B-th face frame position and C-th face frame position;
after determining the historical face frame position queue, referring to fig. 3 and 4, steps S100-S103 are further included;
s100, determining the position of a target face frame;
the method comprises the steps that a target face frame position is used for representing the most frequently occurring position of a face frame in a historical face frame position queue, the position is selected as an average, namely when only an initial face frame position exists in the historical face frame position queue, the average of three initial face frame positions is calculated, namely the average of the left edge distances of the average of the three initial face frame positions, the average of the right edge distances of the average of the three initial face frame positions, the average of the upper edge distances of the average of the three initial face frame positions and the average of the lower edge distances of the average of the three initial face frame positions, and a virtual positioning face frame is generated in a logic processing unit according to the calculated average of the left edge distance, the right edge distance, the upper edge distance and the lower edge distance in the historical face frame position queue;
in another embodiment, the target face frame position is selected as the mode, because the logic processing unit is relatively accurate when measuring the left edge distance, the right edge distance, the upper edge distance and the lower edge distance, any tiny movement can lead to the change of the values of the left edge distance, the right edge distance, the upper edge distance and the lower edge distance, so that in the 180 face frame positioning data in the application, more identical values are difficult to find to be used as references of the mode, the mode which is the approximate value of the left edge distance, the right edge distance, the upper edge distance and the lower edge distance is selected, the values of the left edge distance, the right edge distance, the upper edge distance and the lower edge distance can be determined after the values of the left edge distance, the right edge distance, the upper edge distance and the lower edge distance are accurate to the mode after the positions are accurate to be located, the number after the positions are accurate to be located, then the original data corresponding to the screened face frame position is accurate to the position after the position is located to be located to the mode, the mode is determined again, the left edge distance, the right edge distance, the upper edge distance and the lower edge distance are determined to be the mode after the position is accurate to be the position to be the mode, the mode is further accurate to the position of the mode, and the face frame is further accurate to be the position to be the mode is the position of the mode.
S101, introducing a new face frame position, and marking the new face frame position as a fourth face frame position;
s102, calculating and comparing the IOU value of the fourth face frame position and the IOU value of the positioning face frame, if the IOU value of the fourth face frame position and the IOU value of the positioning face frame are larger than or equal to a preset overlapping threshold value, merging the fourth face frame position into a historical face frame position queue, recalculating the target face frame position, and repeating the steps S101-S102 to fill the historical face frame position queue; when the historical face frame position queue is filled with the face frame positions, when a new face frame position is continuously introduced, the face frame position with the longest time interval with the current face frame position is removed from the historical face frame position queue, and the face frame position with longer acquisition time in the historical face frame position queue is removed so as to keep the accuracy of the target face frame position.
An intelligent access control system, referring to FIG. 5, comprises
The acquisition unit is used for acquiring the current face frame position; the camera can be selected to scan and position the face in the form of a face frame;
the determining unit is used for determining the IOU value of the current face frame position and the target face frame position in the historical face frame position queue, and the target face frame position is generated based on the historical face frame position queue;
the mobile unit is used for comparing the IOU value with a preset overlapping threshold value, and if the IOU value is greater than or equal to the overlapping threshold value, moving the face frame position to the historical face frame position queue to form a current face frame position queue;
the computing unit is used for computing the target face frame position in the current face frame position queue and taking the result as the final face frame position.
It should be noted that: in the device provided in the above embodiment, when implementing the functions thereof, only the division of the above functional modules is used as an example, in practical application, the above functional allocation may be implemented by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules, so as to implement all or part of the functions described above. In addition, the embodiments of the apparatus and the method provided in the foregoing embodiments belong to the same concept, and specific implementation processes of the embodiments of the method are detailed in the method embodiments, which are not repeated herein.
The foregoing are all preferred embodiments of the present application, and are not intended to limit the scope of the present application in any way, therefore: all equivalent changes in structure, shape and principle of this application should be covered in the protection scope of this application.
Claims (8)
1. The data processing method of the intelligent access control is characterized by being applied to an intelligent access control system and comprising the following steps of;
establishing a historical face frame position queue;
based on the historical face frame position queue, acquiring initial face frame position parameters, including the following steps:
acquiring a first face frame position and a second face frame position;
acquiring a third face frame position;
calculating IOU values of the third face frame position and the first face frame position, and recording the IOU values as first repeated values, wherein the IOU values are the coincidence rate of the two face frames;
calculating IOU values of the third face frame position and the second face frame position, and recording the IOU values as second repeated values;
comparing the first repeated value with a preset overlapping threshold value and comparing the second repeated value with the preset overlapping threshold value;
when the first repeated value and the second repeated value are both larger than or equal to the overlapping threshold value, moving the third face frame position to the initial face frame position parameter to form a new initial face frame position parameter;
when only one of the first repeated value and the second repeated value is larger than or equal to the overlapping threshold value, the method comprises the following steps;
moving the third face frame position to the initial face frame position parameters to form new initial face frame position parameters;
removing the face frame positions with corresponding IOU values smaller than the overlapping threshold values in the historical face frame position queues;
acquiring the position of a face frame;
comparing the IOU values of the existing face frame positions in the face frame position queue;
according to the comparison result, adjusting the face frame positions in the historical face frame position queue until the number of data stored in the historical face frame position queue meets the preset number;
acquiring the current face frame position;
determining IOU values of a target face frame position in the current face frame position and historical face frame position queues, wherein the target face frame position is generated based on the historical face frame position queues;
comparing the IOU value with the preset overlapping threshold value; if the IOU value is greater than or equal to the overlapping threshold value, moving the face frame position to a historical face frame position queue to form a current face frame position queue;
and calculating the target face frame position in the current face frame position queue, and taking the result as the final face frame position.
2. The data processing method of the intelligent access control according to claim 1, wherein the number of data stored in the historical face frame position queue is a fixed value; the method further comprises the steps of:
and when the historical face frame position queue is fully loaded and the face frame position is moved to the historical face frame position queue, removing the face frame position with the longest time interval with the face frame position from the historical face frame position queue.
3. The method for processing data of an intelligent entrance guard according to claim 1, wherein the current face frame position queue is emptied after the final face frame position is identified.
4. The data processing method of the intelligent access control according to claim 1, wherein the number of the face frame positions stored in the initial face frame position parameter is greater than or equal to three.
5. The method for processing the data of the intelligent access control according to claim 1, wherein when the first repetition value and the second repetition value are smaller than the overlapping threshold value, the method comprises the following steps of;
acquiring an introduction time parameter of the second face frame position;
acquiring an introduction time parameter of a third face frame position;
calculating a difference value between the introduction time parameter of the third face frame position and the introduction time parameter of the second face frame data, and marking the difference value as a first time difference value;
comparing the first time difference value with a preset time limit threshold;
and when the first time difference value is smaller than the time limit threshold value, removing the third face frame position, acquiring the face frame position and processing the face frame position in the same way as the third face frame position.
6. The method for processing data of intelligent entrance guard according to claim 5, wherein when the first time difference is greater than the time limit threshold, removing the first face frame position and the second face frame position and moving the original third face frame position to the original face frame position parameter as the first face frame position; and acquiring the position of the second face frame, and processing the initial face frame position parameters again.
7. An intelligent access control device, characterized in that it is applied in any one of the above claims 1-6, comprising:
the intelligent access control device comprises an acquisition unit, a determination unit, a mobile unit and a calculation unit; wherein,,
the determining unit is used for establishing a historical face frame position queue;
the obtaining unit is configured to obtain an initial face frame position parameter based on a historical face frame position queue, and includes the following steps:
the acquisition unit is also used for acquiring the first face frame position and the second face frame position;
the acquisition unit is further used for acquiring a third face frame position;
the computing unit is used for computing IOU values of the third face frame position and the first face frame position, and recording the IOU values as first repeated values, namely the coincidence rate of the two face frames;
the computing unit is further configured to compute an IOU value of the third face frame position and the second face frame position, and record the IOU value as a second repetition value;
the calculating unit is further used for comparing the first repeated value with a preset overlapping threshold value and comparing the second repeated value with the preset overlapping threshold value;
the mobile unit is used for moving the third face frame position to the initial face frame position parameters to form new initial face frame position parameters when the first repeated value and the second repeated value are both larger than or equal to the overlapping threshold value;
when only one of the first repeated value and the second repeated value is larger than or equal to the overlapping threshold value, the method comprises the following steps;
the mobile unit is further used for moving the third face frame position to form a new initial face frame position parameter in the initial face frame position parameters;
the mobile unit is further used for removing the face frame positions, corresponding to IOU values, smaller than the overlapping threshold value in the historical face frame position queue;
the acquisition unit is also used for acquiring the position of the face frame;
the computing unit is also used for comparing the face frame position with the IOU value of the existing face frame position in the historical face frame position queue;
the mobile unit is further used for adjusting the face frame positions in the historical face frame position queue according to the comparison result until the number of data stored in the historical face frame position queue meets the preset number;
the acquisition unit is also used for acquiring the current face frame position;
the determining unit is further configured to determine an IOU value of a target face frame position in the current face frame position and the historical face frame position queue, where the target face frame position is generated based on the historical face frame position queue;
the mobile unit is further configured to compare the IOU value with the preset overlap threshold, and if the IOU value is greater than or equal to the overlap threshold, move the face frame position to the historical face frame position queue to form a current face frame position queue;
the computing unit is further configured to compute a target face frame position in the current face frame position queue, and take a result as a final face frame position.
8. An electronic device comprising a processor, a memory for storing instructions, and a transceiver for communicating with other devices, the processor for executing the instructions stored in the memory to cause the electronic device to perform the method of any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210745554.9A CN115050129B (en) | 2022-06-27 | 2022-06-27 | Data processing method and system for intelligent access control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210745554.9A CN115050129B (en) | 2022-06-27 | 2022-06-27 | Data processing method and system for intelligent access control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115050129A CN115050129A (en) | 2022-09-13 |
| CN115050129B true CN115050129B (en) | 2023-06-13 |
Family
ID=83163688
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210745554.9A Active CN115050129B (en) | 2022-06-27 | 2022-06-27 | Data processing method and system for intelligent access control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115050129B (en) |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107967677B (en) * | 2017-12-15 | 2020-02-04 | Oppo广东移动通信有限公司 | Image processing method, image processing device, computer-readable storage medium and computer equipment |
| CN109815843B (en) * | 2018-12-29 | 2021-09-14 | 深圳云天励飞技术有限公司 | Image processing method and related product |
| CN110688930B (en) * | 2019-09-20 | 2023-07-18 | Oppo广东移动通信有限公司 | Face detection method and device, mobile terminal and storage medium |
| CN110766726B (en) * | 2019-10-17 | 2022-03-01 | 重庆大学 | Visual positioning and dynamic tracking method for moving target of large bell jar container under complex background |
| CN111028191B (en) * | 2019-12-10 | 2023-07-04 | 上海闻泰电子科技有限公司 | Anti-shake method and device for video image, electronic equipment and storage medium |
| CN111079686B (en) * | 2019-12-25 | 2023-05-23 | 开放智能机器（上海）有限公司 | Single-stage face detection and key point positioning method and system |
| CN111582141B (en) * | 2020-04-30 | 2023-05-09 | 京东方科技集团股份有限公司 | Face recognition model training method, face recognition method and device |
-
2022
- 2022-06-27 CN CN202210745554.9A patent/CN115050129B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN115050129A (en) | 2022-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1981278B1 (en) | Automatic tracking device and automatic tracking method | |
| KR100559471B1 (en) | System and method for detecting face using symmetric axis | |
| WO2021249306A1 (en) | Crowd overdensity prediction method and device | |
| KR20210070945A (en) | Server, method and program for measuring waste discharge and managing dispose of waste | |
| US12041341B2 (en) | Information processing apparatus, camera control method, program, camera apparatus, and image processing system | |
| CN106603968A (en) | Information processing apparatus and information processing method | |
| US7613334B2 (en) | Method and arrangement for electronic recording of rolled fingerprints | |
| CN115050129B (en) | Data processing method and system for intelligent access control | |
| CN112561967A (en) | Control method and device for vehicle door lock | |
| CN111353374A (en) | Information processing apparatus, control method thereof, and computer-readable storage medium | |
| JP2599701B2 (en) | Elevator Standby Passenger Number Detection Method | |
| JP2002029421A (en) | Method and device for determining presence of train, and monitoring system | |
| CN103810460B (en) | Object tracking method and object tracking device | |
| CN110713082A (en) | Elevator control method, system, device and storage medium | |
| JP3630703B2 (en) | Object monitoring device | |
| JP7353825B2 (en) | Image processing device and method, image input device, image processing system, program | |
| RU2175148C1 (en) | Method for recognizing person identity | |
| KR20190134916A (en) | Method and apparatus for collecting floating population data on realtime road image | |
| JP2019204434A (en) | Information processing device, system, control method for information processing device, and program | |
| CN112528818A (en) | Data statistical method and device, electronic equipment and machine-readable storage medium | |
| KR102580715B1 (en) | Apparatus and method for counting number of visitors based on deep learning model | |
| CN109740518A (en) | The determination method and device of object in a kind of video | |
| CN115391054B (en) | Resource allocation method of vehicle-mounted machine system and vehicle-mounted machine system | |
| JP5968402B2 (en) | Image processing method and apparatus | |
| KR102589150B1 (en) | Long-distance object detection system using cumulative difference image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |