Internet-based facial recognition police information tracking platform

Abstract

The present invention discloses an Internet-based face recognition police information tracking platform, which includes a face analysis module, an identity formulation module, a police officer center module and a real-time monitoring module; the face analysis module includes a face information extraction unit, which is used to identify facial feature information of a face and extract and analyze the facial feature information; the identity formulation module is used to formulate an individual's identity based on the facial feature information of the face analysis module; the police officer center module includes an alarm information receiving unit and a police officer dispatch plan formulation unit, which is used for the police officer to predict the processing method and the police force to be dispatched after receiving the alarm information; the real-time monitoring module includes a real-time monitoring unit for movement trajectory and a real-time monitoring unit for the scene of the alarm, and the real-time monitoring module is used to monitor the movement trajectory of the person after face recognition and the scene picture when the alarm occurs.

Description

Technical Field

[0001] The present invention relates to the field of police information technology, and in particular to an Internet-based face recognition police information tracking platform. Background Art

[0002] With the continuous development of society, the application of facial recognition has become common in every corner of life. However, due to the impact of the current epidemic, people traveling in public places need to wear masks, which covers most of their facial features and brings new challenges to facial recognition. At the same time, the combination of modern digital information and facial recognition is prone to system fragmentation due to the complexity of the system. The recognition function is not intelligent and the recognition is not humane, resulting in low case efficiency, information islands between departments, and low case information utilization. Summary of the Invention

[0003] The purpose of the present invention is to provide an Internet-based face recognition police information tracking platform to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions: an Internet-based face recognition police information tracking platform, including a face analysis module, an identity formulation module, a police center module and a real-time monitoring module. The face analysis module is used to identify facial feature information of a face and extract and analyze the facial feature information; the identity formulation module is used to formulate an individual's identity based on the facial feature information of the face analysis module; the police center module is used for the police to predict the processing method and the police force that needs to be dispatched after receiving the alarm information; the real-time monitoring module is used to monitor the whereabouts of the person after face recognition and the on-site pictures when the alarm occurs.

[0005] Furthermore, the face analysis module includes a face information extraction unit, which extracts facial features of the identified person, including neck, hairstyle, and eye pattern features. The face information extraction unit then determines the gender and age of the identified person and transmits the facial features and gender information to the identity determination module. While extracting information about the identified person, the face information extraction unit also transmits a signal to the real-time monitoring module. The process of determining the gender of a face and extracting facial features is as follows:

[0006] Step S100: The facial information extraction unit determines the convexity of the neck of the identified person. If the convexity of the neck of the identified person is greater than the system-preset angle threshold for male convexity, the identified person is male, and the gender information, hairstyle characteristics, and eye pattern characteristics of the identified person are recorded. If the neck of the identified person is blocked or the convexity angle is less than the system-preset angle threshold for male convexity, the facial information extraction unit cannot determine the identity and will proceed to the following steps.

[0007] Step S200: The facial information extraction unit identifies hairstyle features, which include hair volume, hair color, and length. The facial information extraction unit determines the gender of the identified person by identifying the hairstyle features, and records the hairstyle features and the gender of the person.

[0008] Step S300: The facial information extraction unit identifies the eye pattern features after determining the gender of the identified person in step S200. The eye pattern features include the depth of the eye pattern. The facial information extraction unit determines the age of the identified person by identifying the eye pattern features, and records the gender, hairstyle features and age of the identified person.

[0009] The facial information extraction unit extracts three types of information: neck, hairstyle, and eye pattern. This is to avoid the problem of other facial information being obscured by masks. This allows for both identification and the capture of key features for future reference. The facial information extraction unit first uses salient features to determine gender, quickly defining the individual and laying the foundation for subsequent analysis, providing police with a clearer picture of the individual.

[0010] Furthermore, the specific process of determining the gender of the person by identifying the hairstyle feature in step S200 is as follows:

[0011] Step S210: When the difference between the adjacent hair volume per square centimeter of the identified person is compared with the system preset hair volume per square centimeter difference, a hair volume characteristic value a is obtained; when the hair volume characteristic value is negative, its absolute value is taken.

[0012] Step S220: Compare the hair color feature of the identified person with the system preset female hair color classification, where the hair color classification includes a first hair color category and a second hair color category. The first hair color category has a first hair color threshold, and the second hair color category has a second hair color threshold. If the hair color feature of the identified person is the system preset female hair color classification, then the hair color feature value of the identified person is b, where b={first hair color threshold, second hair color threshold}.

[0013] Step S230: The length feature value of the identified person is compared with the length feature cutoff value preset by the system to obtain a length feature difference value c. When c is a negative value, its absolute value is taken; when c is a positive value, it is taken as 0;

[0014] Step S240: Calculate the weighted average using a in step S210, b in step S220, and c in step S230, assuming the weight of the hair volume feature is a1, the weight of the hair color feature is b1, and the weight of the length feature is c1; using the formula:

[0015] The eigenvalue for identifying a person as male is E, where the weights are 10>>a1>c1>b1>>1; let the system preset male eigenvalue be E'. When the obtained male eigenvalue is greater than E', the person is identified as male; otherwise, she is female.

[0016] The facial information extraction unit first determines the amount of hair and gives a larger weight to the hair amount feature because the difference in hair amount between contemporary men and women is more obvious, especially for middle-aged men and women. Therefore, the algorithm can more accurately determine the gender of the identified person. The hair color feature is used for judgment because there is no obvious difference in hair color that can directly distinguish between men and women in real life, so the weight is the smallest. The length feature is used for judgment because hair length is a relatively obvious feature between men and women, so the weight is the middle value of the three. The weighted average can be used to obtain the most likely gender of the identified person.

[0017] Furthermore, in step S300, the eye pattern features are identified, and the specific process is as follows:

[0018] Step S310: The facial information extraction unit extracts the eye image to obtain an eye detail image, and uses the eye detail image to determine whether there are eye wrinkles. If not, the person is young; if so, the determination continues.

[0019] Step S320: The face information extraction unit restores the eye detail image into a black and white image, which contains all eye lines e i , i={1,2,3,.....c}, eye pattern e i Arrange them in the order of 1, 2, 3, .....c from top to bottom, and the eye pattern in the black and white image is a black image. Amplify the black and white image by a certain multiple and calculate the area of ​​the black image in the amplified image;

[0020] Step S330: i The longest horizontal line in the rectangle is the length W1 of the rectangle, and the distance from the highest point of eye pattern e1 to the highest point of eye pattern e is c The vertical distance W2 of the lowest point is used as the width of the rectangle to form an eye pattern rectangle. The area of ​​the eye pattern rectangle is W. Then calculate the area of ​​each eye pattern e separately. i The area w i , the eye pattern i After magnification, it becomes a triangle, and each eye pattern is e i The area of ​​the eye pattern gathering point is the vertex, and the vertex is on the edge of the eye pattern matrix. The face information extraction unit extracts the vertex to the eye pattern e i The vertical distance L between the vertex and the base i and the bottom distance d i ;

[0021] Using the formula: W = W1 × W2;

[0022] Among them, k is the coefficient of the depth ratio of eye wrinkles;

[0023] Step S340: The face information extraction unit presets a standard threshold value K for middle-aged eye lines; when the eye line depth ratio coefficient k is greater than the standard threshold value K for middle-aged eye lines, the person is determined to be an elderly person, otherwise, the person is middle-aged.

[0024] The facial information extraction unit recognizes the eye pattern features, which increases the possibility of successful recognition. This is because the eyes cannot be covered under surveillance, and the eye patterns are the most direct evidence for judging a person's age and are evidence with higher reliability. Since the depth of the eye patterns is a significant feature representing age, the eye image is extracted and amplified, and young people are effectively excluded based on the presence or absence of eye patterns, which reduces the system's algorithm, and the image is amplified to effectively convert the eye patterns from a thin line visible to the naked eye into a triangular shape whose area can be calculated. Due to the irregularity of the eye patterns, in order to summarize the calculation of all patterns, the longest line in the eye pattern is taken as the length of the rectangle, and the distance between the top and the bottom is taken as the width. Then, the ratio of the pattern area to the overall rectangular area is calculated, which can effectively analyze the depth of the eye patterns of the identified person. In addition, the judgment of age helps the police to implement the police dispatch plan formulation unit.

[0025] Furthermore, when the eye wrinkle depth ratio coefficient k is close to the standard threshold value K of middle-aged eye wrinkles, it is impossible to accurately determine the age of the identified person, and the following steps will be performed:

[0026] Step S341: The face information extraction unit extracts the eye area image below the eyebrows and above the cheekbones, and performs area processing on the eye area image. The face information extraction unit extracts the eye area of ​​m identified people and calculates the average eye area. As the standard eye area value;

[0027] Step S342: The eye periorbital image is binarized, the grayscale value of the eye wrinkles is set to 255, and the grayscale values ​​of the rest of the image except the eye wrinkles are set to 0; the face information extraction unit extracts the number of eye wrinkles as h, and calculates the density value of the eye wrinkles under the standard eye periorbital area.

[0028] Step S343: Compare the density value in step S342 with the number of eye wrinkles per square centimeter preset by the system. When the density value γ under the standard eye area is greater than the system preset eye wrinkle density value, , it is further determined that the identified person is an elderly person; if the density value γ under the standard eye area is less than the eye pattern density value preset by the system , it is further determined that the identified person is a middle-aged person.

[0029] When the eyes are partially magnified, there may be certain errors. Therefore, the whole area around the eyes is identified and judged. Further refinement of the eye lines can avoid errors caused by unusual phenomena such as premature aging or better maintenance at the junction of middle-aged and elderly people. The eye area image is binarized to highlight the eye lines, so that the facial information extraction unit can more accurately extract the number of eye lines and calculate the density of the eye lines around the entire eye area. Therefore, neither the depth nor the length can affect the change in density. With the increase of age, the number of lines around the eyes will increase to a certain extent. Therefore, the use of density calculation improves the accuracy of identifying age stages based on eye lines.

[0030] Furthermore, the identity determination module performs identity determination while the facial information extraction unit extracts the facial information of the person. The identity determination module records the gender, hair volume, hair color, length, and eye wrinkle depth of the identified person to form the first element of the identity table I.

[0031] At the same time, the facial information extraction unit transmits a signal to the real-time monitoring module, and the real-time monitoring module identifies and monitors the whereabouts of the identified person. The real-time monitoring module transmits the information of the whereabouts of the identified person to the identity establishment unit. The whereabouts of the identified person constitutes the second element of the identity table I, and the identity table I is composed of the first element and the second element; the information of the first identification of the identified person in the identity table I is recorded as the initial identity table I1. The identity table I is used to extract and determine the identity feature information of the on-site personnel when an alarm occurs.

[0032] The first element included in the identity table is formulated on the one hand to record the facial feature information of the identified person for the purpose of identification under monitoring in other scenarios; on the other hand, it is to improve the efficiency of person locking and can simultaneously extract the second element information related to the first element in the identity table; formulating the identity table helps the police center module to extract the identity information of relevant personnel at the police scene, which is conducive to the sharing and utilization of information, saves manpower and material resources, and achieves the beneficial effect of information flow.

[0033] Furthermore, the real-time monitoring module includes a real-time tracking monitoring unit and a real-time monitoring unit for the scene of an emergency. The real-time tracking monitoring unit is used to identify the trajectory of the person after face recognition. The real-time monitoring unit for the scene of an emergency is used to monitor the scene and retrieve on-site information when an emergency occurs.

[0034] The real-time tracking unit tracks and identifies the identified person based on the first element in the identity table I. The real-time tracking unit records the duration of each monitoring of the identified person as h: the monitoring period ends when the time from the start of the facial information extraction unit to the identification person's absence from monitoring exceeds the system-limited time. The real-time tracking unit records the image of the identified person each time it appears under monitoring. The image of the identified person includes location information and time information. The real-time tracking unit arranges all the images of the identified person within the monitoring period h in chronological order, then connects the location information on the image to form a tracking line, which is stored in the identity determination module.

[0035] The face information extraction unit obtains auxiliary identity table I by identifying the same person at different times j , j = {2,3,4......}; auxiliary identity table I j Compare with the initial identity table I1. If they are the same, keep the first element of the initial identity table I1 and ignore the auxiliary identity table I j The first element in, if not the same, the auxiliary identity table I j The first element is updated to the initial identity table I1 to obtain the latest identity table I 1i , i={1,2,3,4......}, and the latest identity table I 1i Each update retains the original information, and the real-time monitoring unit of the whereabouts track is based on the latest identity table I 1i The first element in the identification record of the person's whereabouts trajectory line, and the whereabouts trajectory line is synchronously updated to the latest identity table I 1i middle;

[0036] The real-time monitoring unit at the scene of the police incident locks on the relevant personnel at the scene, extracts the identity table information of the relevant personnel, and transmits the monitoring image to the police center module.

[0037] The real-time monitoring module uses the identity table information and improves the identity information for the identity table, achieving a win-win effect. The information update in the identity table does not overwrite the past information, which provides an effective basis for data processing and provides a feasible travel plan for the police dispatch plan formulation unit.

[0038] Furthermore, the police center module includes a police information receiving unit and a police dispatch plan formulation unit; the police information receiving unit is used to receive and process alarm information, and the police dispatch plan formulation unit is used to formulate a dispatch plan based on the processed alarm information; the specific implementation process of the police information receiving unit is as follows:

[0039] The alarm receiving unit receives the alarm information and breaks down the alarm personnel's audio information into n audio units. An audio unit is the audio of two adjacent words. The alarm receiving unit analyzes the time length t in each audio unit.n , n={1,2,3......}, t n Indicates the time length of the nth audio unit, using the formula:

[0040]

[0041] in Indicates the average duration of each audio unit in the audio information;

[0042] The alarm receiving unit presets the time length of each audio unit as when When the alarm receiving unit determines that the alarm is non-emergency; when When the alarm situation is detected, the alarm situation receiving unit determines that the alarm situation is an emergency situation; the alarm situation receiving unit transmits the alarm situation to the police dispatch plan formulation unit.

[0043] The alarm receiving unit divides the alarm information and determines the time interval between words. The time interval represents the importance of the alarm personnel's prediction of the alarm situation. The system quickly splits and calculates to effectively classify the alarm situation. The alarm receiving unit's prediction of the alarm information can save police force, avoid waste of police force, and effectively provide the fastest option for case handling.

[0044] Furthermore, the specific implementation process of the police dispatch plan formulation unit is as follows:

[0045] The police dispatch plan formulation unit will dispatch non-emergency police incidents in chronological order and dispatch police incidents according to the idle police officer ratio set by the system; the police dispatch plan formulation unit will directly prioritize emergency police incidents for dispatch; at the same time, the police dispatch plan formulation unit will retrieve the monitoring information from the real-time monitoring unit at the scene of the police incident to make a judgment: when the monitoring information indicates a traffic jam, the police dispatch plan formulation unit will give priority to dispatching police officers to respond on police motorcycles; if the traffic is normal, the police dispatch plan formulation unit will dispatch idle police cars for dispatch.

[0046] When the monitoring screen involves a public place with more monitoring screens, the real-time monitoring module retrieves the latest identity table of the identified person. 1i Information, the police dispatch plan formulation unit according to the latest identity table I 1i The first element judgment: When the person is identified as an elderly person, the police dispatch plan formulation unit determines whether to call the 120 emergency number and whether to prepare an elderly first aid kit; when the person is identified as a female, it determines whether she is pregnant and needs to call the 120 emergency number, and whether to dispatch a female police officer for the police response; when the hair color and length characteristics of the identified person change every time the historical records of the latest identity table are updated, the police dispatch plan formulation unit determines whether there are any abnormalities with the person.

[0047] The police dispatch plan formulation unit makes judgments based on the first element of the identity table, which can fully understand the situation at the scene, reduce losses caused by not taking into account the special needs of special groups, make police dispatch more humane, and increase the possibility of police officers solving problems quickly; and the police dispatch plan formulation unit uses the characteristic information in the identity table to take care of special groups at the police scene, such as whether the elderly are likely to have sudden illnesses, whether women are pregnant or have other inconveniences that require female police officers to respond to the police, or predict situations with abnormal appearance characteristics.

[0048] Further, the police dispatch plan formulation unit shall make the following arrangements according to the latest identity table I: 1i The second factor judgment in the police dispatch plan formulation unit extracts the latest identity table I by identifying the first factor 1i The second element is to determine the whereabouts of the identified person;

[0049] The police dispatch plan formulation unit will identify the latest identity list of the person 1i The whereabouts of the track are checked and the latest identity table I 1i The location represented by the same image in the historical record is taken as the starting point g, and the latest identity table I 1i The initial endpoint in is recorded as x1, and the latest identity table I 1i The other endpoints except the initial endpoint are recorded as x i-1 , calculate the straight-line distance between each end point and the starting point: x1-g, x2-g, ...x i-1 -g, and the straight-line distance between each end point and the initial end point: x i-1 -x1...x2-x1; using the formula:

[0050]

[0051] a i-1 =arccosa i-1

[0052]

[0053] where α i-1 is the angle of the i-th movement trajectory, The average angle of the identified person's whereabouts trajectory, the police officer's response plan formulation unit presets the standard whereabouts trajectory angle threshold of the person's activity range

[0054] like The police dispatch plan formulation unit determines that the identified person's activities are abnormal and records the whereabouts of the person; if The police dispatch plan formulation unit determines that the identified person's activities are normal and do not need to be recorded.

[0055] The prediction of the whereabouts of the police officers by the dispatch plan formulation unit can effectively save the workload of the police officers in recording files, and can also understand the historical trajectories of the people at the crime scene, providing a basis for the next step of judging the people on the scene, making the information flow bidirectional and improving the utilization rate of the identity table; because the whereabouts can be seen as a straight line with a low degree of curvature on a map with a certain degree of reduction, the starting point and end point of each whereabouts trajectory are used as the two end points of the straight line segment, and the approximate length can also be effectively calculated. Each whereabouts trajectory except the initial whereabouts trajectory can be triangled with the initial whereabouts trajectory, and the average deviation angle between the two is calculated to estimate whether the identified person has abnormal whereabouts or to extract the whereabouts trajectory information, quickly and effectively providing useful information for the police officers to handle the case.

[0056] Compared with the existing technology, the beneficial effects achieved by the present invention are: the Internet-based face recognition police information tracking platform is based on face recognition, which enhances the accuracy of face recognition when most of the face is blocked, and uses facial feature information in combination with corresponding police equipment to make information utilization higher, and can effectively integrate and analyze information when a case occurs, make intelligent and humane judgments on police information, and improve the utilization rate of police information. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0058] Figure 1 This is a flow chart of the internet-based face recognition police information tracking platform of the present invention;

[0059] Figure 2 The present invention is a method for determining gender and extracting facial features through a face based on an Internet-based face recognition police information tracking platform;

[0060] Figure 3 The present invention is a method for identifying a person's gender by identifying hairstyle features on an Internet-based face recognition police information tracking platform;

[0061] Figure 4 The present invention is a method for identifying eye pattern features based on an Internet-based face recognition police information tracking platform;

[0062] Figure 5 This is a method used in the Internet-based face recognition police information tracking platform of the present invention when the age of the identified person cannot be accurately determined. DETAILED DESCRIPTION

[0063] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0064] See also Figure 1 The present invention provides a technical solution: an Internet-based face recognition police information tracking platform, including a face analysis module, an identity determination module, a police center module and a real-time monitoring module. The face analysis module is used to identify facial feature information of a face and extract and analyze the facial feature information; the identity determination module is used to determine an individual's identity based on the facial feature information of the face analysis module; the police center module is used for police officers to predict the processing method and the police force to be dispatched after receiving the alarm information; the real-time monitoring module is used to monitor the whereabouts of the person after face recognition and the scene when the alarm occurs.

[0065] The face analysis module includes a face information extraction unit, which extracts facial features of the identified person, including neck, hairstyle and eye pattern features. The face information extraction unit then determines the gender and age of the identified person and passes the facial features and gender information to the identity establishment module. While the face information extraction unit is extracting information about the identified person, it will transmit a signal to the real-time monitoring module.

[0066] The facial information extraction unit extracts three types of information: neck, hairstyle features, and eye pattern features in order to avoid the trouble caused by other facial information blocked by the mask. It can not only identify the person, but also capture the typical features of the identified person as reserve data.

[0067] The process of determining the gender of a face and extracting facial features is as follows:

[0068] Step S100: The facial information extraction unit determines the convexity of the neck of the identified person. If the convexity of the neck of the identified person is greater than the system-preset angle threshold for male convexity, the identified person is male, and the gender information, hairstyle characteristics, and eye pattern characteristics of the identified person are recorded. If the neck of the identified person is blocked or the convexity angle is less than the system-preset angle threshold for male convexity, the facial information extraction unit cannot determine the identity and will proceed to the following steps.

[0069] Step S200: The facial information extraction unit identifies hairstyle features, which include hair volume, hair color, and length. The facial information extraction unit determines the gender of the identified person by identifying the hairstyle features, and records the hairstyle features and the gender of the person.

[0070] The specific process of determining the gender of a person by identifying the hairstyle features in step S200 is as follows:

[0071] Step S210: When the difference between the adjacent hair volume per square centimeter of the identified person is compared with the system preset hair volume per square centimeter difference, a hair volume characteristic value a is obtained; when the hair volume characteristic value is negative, its absolute value is taken.

[0072] If the system presets the difference in hair volume per square centimeter to be 25, and the difference in hair volume per square centimeter between adjacent people being identified is 80, then the difference is a = 80 - 25 = 55;

[0073] Step S220: Compare the hair color feature of the identified person with the system preset female hair color classification, where the hair color classification includes a first hair color category and a second hair color category. The first hair color category has a first hair color threshold, and the second hair color category has a second hair color threshold. If the hair color feature of the identified person is the system preset female hair color classification, then the hair color feature value of the identified person is b, where b={first hair color threshold, second hair color threshold}.

[0074] Assume that the first hair color threshold is 55% of the dark color percentage, and the second hair color threshold is 45% of the light color percentage. If the hair color of the person being recognized is dark, then b = 55; if the hair color of the person being recognized is light, then b = 45.

[0075] Step S230: The length feature value of the identified person is compared with the length feature cutoff value preset by the system to obtain a length feature difference value c. When c is a negative value, its absolute value is taken; when c is a positive value, it is taken as 0;

[0076] Assume that the system preset length feature cutoff value is 25, and the hair length of the identified person is 2, that is, c = 25-2> 0, so c = 0;

[0077] Step S240: Calculate the weighted average using a in step S210, b in step S220, and c in step S230, assuming the weight of the hair volume feature is a1, the weight of the hair color feature is b1, and the weight of the length feature is c1; using the formula:

[0078] The eigenvalue for identifying a person as male is E, where the weights are 10>>a1>c1>b1>>1; let the system preset male eigenvalue be E'. When the obtained male eigenvalue is greater than E', the person is identified as male; otherwise, she is female.

[0079] Assume weights a1 = 7, b1 = 2, c1 = 4, calculate Compare this value with the system's preset male characteristic value.

[0080] The facial information extraction unit first determines the amount of hair and gives a larger weight to the hair amount feature because the difference in hair amount between contemporary men and women is more obvious, especially for middle-aged men and women. Therefore, the algorithm can more accurately determine the gender of the identified person. The hair color feature is used for judgment because there is no obvious difference in hair color that can directly distinguish between men and women in real life, so the weight is the smallest. The length feature is used for judgment because hair length is a relatively obvious feature between men and women, so the weight is the middle value of the three. The weighted average can be used to obtain the most likely gender of the identified person.

[0081] Step S300: The facial information extraction unit identifies the eye pattern features after determining the gender of the identified person in step S200. The eye pattern features include the depth of the eye pattern. The facial information extraction unit determines the age of the identified person by identifying the eye pattern features, and records the gender, hairstyle features and age of the identified person.

[0082] The facial information extraction unit extracts three types of information: neck, hairstyle features, and eye pattern features to avoid the trouble caused by other facial information blocked by the mask. It can not only identify the person, but also grasp the typical features of the identified person as reserve data; the facial information extraction unit first uses significant features to judge the gender, which can quickly divide the boundaries of the person, lay the tone for subsequent judgments, and provide the person direction for police information.

[0083] In step S300, the eye pattern features are identified, and the specific process is as follows:

[0084] Step S310: The facial information extraction unit extracts the eye image to obtain an eye detail image, and uses the eye detail image to determine whether there are eye wrinkles. If not, the person is young; if so, the determination continues.

[0085] Step S320: The face information extraction unit restores the eye detail image into a black and white image, which contains all eye lines e i , i={1,2,3,.....c}, eye pattern e i Arrange them in the order of 1, 2, 3, .....c from top to bottom, and the eye pattern in the black and white image is a black image. Amplify the black and white image by a certain multiple and calculate the area of ​​the black image in the amplified image;

[0086] Step S330: iThe longest horizontal line in the rectangle is the length W1 of the rectangle, and the distance from the highest point of eye pattern e1 to the highest point of eye pattern e is c The vertical distance W2 of the lowest point is used as the width of the rectangle to form an eye pattern rectangle. The area of ​​the eye pattern rectangle is W. Then calculate the area of ​​each eye pattern e separately. i The area w i , the eye pattern i After magnification, it becomes a triangle, and each eye pattern is e i The area of ​​the eye pattern gathering point is the vertex, and the vertex is on the edge of the eye pattern matrix. The face information extraction unit extracts the vertex to the eye pattern e i The vertical distance l from the vertex to the base i and the bottom distance d i ;

[0087] Using the formula: W = W1 × W2;

[0088] Among them, k is the coefficient of the depth ratio of eye wrinkles;

[0089] If the horizontal distance of the eye pattern rectangle is W1 = 2.8, and the vertical distance of the eye pattern rectangle is W2 = 1.6, W = 2.8 × 1.6 = 4.48; assuming that the person to be identified has three eye patterns, w1 = 0.09, w2 = 0.12, and w3 = 0.05; the eye pattern depth ratio coefficient is:

[0090]

[0091] Step S340: The face information extraction unit presets a standard threshold value K for middle-aged eye lines; when the eye line depth ratio coefficient k is greater than the standard threshold value K for middle-aged eye lines, the person is determined to be an elderly person, otherwise, the person is middle-aged.

[0092] The facial information extraction unit recognizes the eye pattern features, which increases the possibility of successful recognition. This is because the eyes cannot be covered under surveillance, and the eye patterns are the most direct evidence for judging a person's age and are evidence with higher reliability. Since the depth of the eye patterns is a significant feature representing age, the eye image is extracted and amplified, and young people are effectively excluded based on the presence or absence of eye patterns, which reduces the system's algorithm, and the image is amplified to effectively convert the eye patterns from a thin line visible to the naked eye into a triangular shape whose area can be calculated. Due to the irregularity of the eye patterns, in order to summarize the calculation of all patterns, the longest line in the eye pattern is taken as the length of the rectangle, and the distance between the top and the bottom is taken as the width. Then, the ratio of the pattern area to the overall rectangular area is calculated, which can effectively analyze the depth of the eye patterns of the identified person. In addition, the judgment of age helps the police to implement the police dispatch plan formulation unit.

[0093] When the eye wrinkle depth ratio coefficient k is close to the middle-aged eye wrinkle standard threshold K, the age of the identified person cannot be accurately determined. The following steps will be performed:

[0094] Step S341: The face information extraction unit extracts the eye area image below the eyebrows and above the cheekbones, and performs area processing on the eye area image. The face information extraction unit extracts the eye area of ​​m identified people and calculates the average eye area. As the standard eye area value;

[0095] Step S342: The eye periorbital image is binarized, the grayscale value of the eye wrinkles is set to 255, and the grayscale values ​​of the rest of the image except the eye wrinkles are set to 0; the face information extraction unit extracts the number of eye wrinkles as h, and calculates the density value of the eye wrinkles under the standard eye periorbital area.

[0096] Step S343: Compare the density value in step S342 with the number of eye wrinkles per square centimeter preset by the system. When the density value γ under the standard eye area is greater than the system preset eye wrinkle density value, , it is further determined that the identified person is an elderly person; if the density value γ under the standard eye area is less than the eye pattern density value preset by the system , it is further determined that the identified person is a middle-aged person.

[0097] Assume that the standard eye area value is square centimeters, the number of eye patterns h identified is 12, then the density value γ is 0.3 per square centimeter, and then the system preset eye pattern density value Compare it to;

[0098] When the eyes are partially magnified, there may be certain errors. Therefore, the whole area around the eyes is identified and judged. Further refinement of the eye lines can avoid errors caused by unusual phenomena such as premature aging or better maintenance at the junction of middle-aged and elderly people. The eye area image is binarized to highlight the eye lines, so that the facial information extraction unit can more accurately extract the number of eye lines and calculate the density of the eye lines around the entire eye area. Therefore, neither the depth nor the length can affect the change in density. With the increase of age, the number of lines around the eyes will increase to a certain extent. Therefore, the use of density calculation improves the accuracy of identifying age stages based on eye lines.

[0099] The identity determination module determines the identity of the person while the facial information extraction unit extracts the facial information. The identity determination module records the gender, hair volume, hair color, length, and eye wrinkle depth of the identified person to form the first element of the identity table I.

[0100] At the same time, the facial information extraction unit transmits a signal to the real-time monitoring module, and the real-time monitoring module identifies and monitors the whereabouts of the identified person. The real-time monitoring module transmits the information of the whereabouts of the identified person to the identity establishment unit. The whereabouts of the identified person constitutes the second element of the identity table I, and the identity table I is composed of the first element and the second element; the information of the first identification of the identified person in the identity table I is recorded as the initial identity table I1. The identity table I is used to extract and determine the identity feature information of the on-site personnel when an alarm occurs.

[0101] The first element included in the identity table is formulated on the one hand to record the facial feature information of the identified person for the purpose of identification under monitoring in other scenarios; on the other hand, it is to improve the efficiency of person locking and can simultaneously extract the second element information related to the first element in the identity table; formulating the identity table helps the police center module to extract the identity information of relevant personnel at the police scene, which is conducive to the sharing and utilization of information, saves manpower and material resources, and achieves the beneficial effect of information flow.

[0102] The real-time monitoring module includes a real-time tracking monitoring unit and a real-time monitoring unit for the scene of an emergency. The real-time tracking monitoring unit is used to identify the trajectory of the person after face recognition. The real-time monitoring unit for the scene of an emergency is used to monitor the scene and retrieve on-site information when an emergency occurs.

[0103] The real-time tracking unit tracks and identifies the identified person based on the first element in the identity table I. The real-time tracking unit records the duration of each monitoring of the identified person as h: the monitoring period ends when the time from the start of the facial information extraction unit to the identification person's absence from monitoring exceeds the system-limited time. The real-time tracking unit records the image of the identified person each time it appears under monitoring. The image of the identified person includes location information and time information. The real-time tracking unit arranges all the images of the identified person within the monitoring period h in chronological order, then connects the location information on the image to form a tracking line, which is stored in the identity determination module.

[0104] The face information extraction unit obtains auxiliary identity table I by identifying the same person at different times j , j = {2,3,4......}; auxiliary identity table I j Compare with the initial identity table I1. If they are the same, keep the first element of the initial identity table I1 and ignore the auxiliary identity table I j The first element in, if not the same, the auxiliary identity table I j The first element is updated to the initial identity table I1 to obtain the latest identity table I 1i , i={1,2,3,4......}, and the latest identity table I 1iEach update retains the original information, and the real-time monitoring unit of the whereabouts track is based on the latest identity table I 1i The first element in the identification record of the person's whereabouts trajectory line, and the whereabouts trajectory line is synchronously updated to the latest identity table I 1i middle;

[0105] The real-time monitoring unit at the scene of the police incident locks on the relevant personnel at the scene, extracts the identity table information of the relevant personnel, and transmits the monitoring image to the police center module.

[0106] The real-time monitoring module uses the identity table information and improves the identity information for the identity table, achieving a win-win effect. The information update in the identity table does not overwrite the past information, which provides an effective basis for data processing and provides a feasible travel plan for the police dispatch plan formulation unit.

[0107] The police center module includes a police information receiving unit and a police dispatch plan formulation unit; the police information receiving unit is used to receive and process alarm information, and the police dispatch plan formulation unit is used to formulate a dispatch plan based on the processed alarm information; the specific implementation process of the police information receiving unit is as follows:

[0108] The alarm receiving unit receives the alarm information and decomposes the audio information of the alarm personnel into n audio units. The audio unit is the audio of two adjacent words. The alarm receiving unit analyzes the time length t in each audio unit. n , n={1,2,3......}, t n Indicates the time length of the nth audio unit, using the formula:

[0109]

[0110] in Indicates the average duration of each audio unit in the audio information;

[0111] If the audio information is decomposed into 6 audio units, t1 = 0.15, t2 = 0.14, t3 = 0.21, t4 = 0.25, t5 = 0.32, t6 = 0.35, then Will The preset duration of each audio unit is Make comparisons;

[0112] The alarm receiving unit presets the time length of each audio unit as when When the alarm receiving unit determines that the alarm is non-emergency; when When the alarm situation is detected, the alarm situation receiving unit determines that the alarm situation is an emergency situation; the alarm situation receiving unit transmits the alarm situation to the police dispatch plan formulation unit.

[0113] The alarm receiving unit divides the alarm information and determines the time interval between words. The time interval represents the importance of the alarm personnel's prediction of the alarm situation. The system quickly splits and calculates to effectively classify the alarm situation. The alarm receiving unit's prediction of the alarm information can save police force, avoid waste of police force, and effectively provide the fastest option for case handling.

[0114] The specific implementation process of the police dispatch plan formulation unit is as follows:

[0115] The police dispatch plan formulation unit will dispatch non-emergency police incidents in chronological order and dispatch police incidents according to the idle police officer ratio set by the system; the police dispatch plan formulation unit will directly prioritize emergency police incidents for dispatch; at the same time, the police dispatch plan formulation unit will retrieve the monitoring information from the real-time monitoring unit at the scene of the police incident to make a judgment: when the monitoring information indicates a traffic jam, the police dispatch plan formulation unit will give priority to dispatching police officers to respond on police motorcycles; if the traffic is normal, the police dispatch plan formulation unit will dispatch idle police cars for dispatch.

[0116] When the monitoring screen involves a public place with more monitoring screens, the real-time monitoring module retrieves the latest identity table of the identified person. 1i Information, the police dispatch plan formulation unit according to the latest identity table I 1i The first element judgment: When the person is identified as an elderly person, the police dispatch plan formulation unit determines whether to call the 120 emergency number and whether to prepare an elderly first aid kit; when the person is identified as a female, it determines whether she is pregnant and needs to call the 120 emergency number, and whether to dispatch a female police officer for the police response; when the hair color and length characteristics of the identified person change every time the historical records of the latest identity table are updated, the police dispatch plan formulation unit determines whether there are any abnormalities with the person.

[0117] The police dispatch plan formulation unit makes judgments based on the first element of the identity table, which can fully understand the situation at the scene, reduce losses caused by not taking into account the special needs of special groups, make police dispatch more humane, and increase the possibility of police officers solving problems quickly; and the police dispatch plan formulation unit uses the characteristic information in the identity table to take care of special groups at the police scene, such as whether the elderly are likely to have sudden illnesses, whether women are pregnant or have other inconveniences that require female police officers to respond to the police, or predict situations with abnormal appearance characteristics.

[0118] The police dispatch plan formulation unit shall make the following arrangements according to the latest identity table I 1i The second factor judgment in the police dispatch plan formulation unit extracts the latest identity table I by identifying the first factor 1i The second element is to determine the whereabouts of the identified person;

[0119] The police dispatch plan formulation unit will identify the latest identity list of the person 1i The whereabouts of the track are checked and the latest identity table I 1i The location represented by the same image in the historical record is taken as the starting point g, and the latest identity table I 1i The initial endpoint in is recorded as x1, and the latest identity table I 1i The other endpoints except the initial endpoint are recorded as x i-1 , calculate the straight-line distance between each end point and the starting point: x1-g, x2-g, ...x i-1 -g, and the straight-line distance between each end point and the initial end point: x i-1 -x1...x2-x1; using the formula:

[0120]

[0121] a i-1 =arccosa i-1

[0122]

[0123] If the latest identity table of the identified person records three movement trajectories, the straight-line distance between the initial end point and the starting point is 10.6 km, the straight-line distance between the second end point and the starting point is 8.2 km, and the straight-line distance between the third end point and the starting point is 16.5 km. The straight-line distance between the initial end point and the second end point is 5.1 km, and the straight-line distance between the third end point and the initial end point is 13.7 km. Calculate the angle between the second movement trajectory and the initial trajectory: a1=arc0.89≈28.36°; the angle between the third trajectory and the initial trajectory is: α2=arc0.56≈55.93°; we get Will and Make comparisons;

[0124] where α i-1 is the angle of each movement trajectory, The average angle of the identified person's whereabouts trajectory, the police officer's response plan formulation unit presets the standard whereabouts trajectory angle threshold of the person's activity range

[0125] like The police dispatch plan formulation unit determines that the identified person's activities are abnormal and records the whereabouts of the person; if The police dispatch plan formulation unit determines that the identified person's activities are normal and do not need to be recorded.

[0126] The prediction of the whereabouts of the police officers by the dispatch plan formulation unit can effectively save the workload of the police officers in recording files, and can also understand the historical trajectories of the people at the crime scene, providing a basis for the next step of judging the people on the scene, making the information flow bidirectional and improving the utilization rate of the identity table; because the whereabouts can be seen as a straight line with a low degree of curvature on a map with a certain degree of reduction, the starting point and end point of each whereabouts trajectory are used as the two end points of the straight line segment, and the approximate length can also be effectively calculated. Each whereabouts trajectory except the initial whereabouts trajectory can be triangled with the initial whereabouts trajectory, and the average deviation angle between the two is calculated to estimate whether the identified person has abnormal whereabouts or to extract the whereabouts trajectory information, quickly and effectively providing useful information for the police officers to handle the case.

[0127] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0128] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. Internet-based face recognition police information tracking platform, characterized by: It includes a face analysis module, an identity determination module, a police center module and a real-time monitoring module. The face analysis module is used to identify facial feature information of a face and extract and analyze the facial feature information; the identity determination module is used to determine the identity of an individual based on the facial feature information of the face analysis module; the police center module is used by police officers to predict the handling method and the police force required to be dispatched after receiving alarm information; the real-time monitoring module is used to monitor the whereabouts of the person after face recognition and the scene when the alarm occurs; The face analysis module includes a face information extraction unit, which extracts facial features of the identified person, including neck, hairstyle and eye pattern features, and then determines the gender and age of the identified person, and transmits the facial features and gender information to the identity determination module. While extracting information from the identified person, the face information extraction unit also transmits a signal to the real-time monitoring module. The real-time monitoring module includes a real-time tracking monitoring unit and a real-time monitoring unit for the scene of an emergency. The real-time tracking monitoring unit is used to identify the trajectory of the person after face recognition. The real-time monitoring unit for the scene of an emergency is used to monitor the scene and retrieve on-site information when an emergency occurs. The real-time tracking unit tracks and identifies the identified person based on the first element in the identity table I. The real-time tracking unit records each monitoring time of the identified person as h: the monitoring ends when the time from when the facial information extraction unit starts working to when the identified person does not appear in the monitoring exceeds the system-limited time; the real-time tracking unit records the image of the identified person each time it appears in the monitoring, the image containing location information and time information. The real-time tracking unit arranges all the images of the identified person within the monitoring time h in chronological order, and then connects the location information on the image to form a tracking line and stores it in the identity formulation module; The face information extraction unit obtains auxiliary identity table I by identifying the same identified person at different times j , j = {2, 3, 4...}; the auxiliary identity table I j Compare with the initial identity table I1. If they are the same, keep the first element of the initial identity table I1 and ignore the auxiliary identity table I j The first element in, if not the same, the auxiliary identity table I j The first element is updated to the initial identity table I1 to obtain the latest identity table I 1i , i={1,2,3,4......}, and the latest identity table I 1i Each update retains the original information, and the real-time tracking unit of the whereabouts track is based on the latest identity table I 1i The first element in the identification table records the whereabouts of the person, and the whereabouts of the person is synchronously updated to the latest identity table I 1i middle; The real-time monitoring unit for the scene of the police incident locks on the relevant personnel on the scene, extracts the identity table information of the relevant personnel, and transmits the monitoring picture to the police center module.

2. The Internet-based facial recognition police information tracking platform according to claim 1 is characterized by: The process of determining the gender of a face and extracting facial features is as follows: Step S100: The facial information extraction unit determines the bulge at the neck of the identified person. If the bulge at the neck of the identified person is greater than a system-preset angle threshold for male bulges, the person is male, and the gender information, hairstyle characteristics, and eye pattern characteristics of the identified person are recorded. If the neck of the identified person is blocked or the bulge angle is less than the system-preset angle threshold for male bulges, the facial information extraction unit cannot determine the identity and will proceed to the following steps. Step S200: The facial information extraction unit identifies hairstyle features, which include hair volume features, hair color features, and hair length features; The face information extraction unit determines the gender of the identified person by identifying the hairstyle features, and records the hairstyle features and the gender of the person; Step S300: The facial information extraction unit identifies the eye pattern features after determining the gender of the identified person in step S200. The eye pattern features include the depth of the eye pattern. The facial information extraction unit determines the age of the identified person by identifying the eye pattern features, and records the gender, hairstyle features and age of the identified person.

3. The Internet-based facial recognition police information tracking platform according to claim 2 is characterized by: The specific process of determining the gender of the person by identifying the hairstyle features in step S200 is as follows: Step S210: Compare the difference between the hair volume per square centimeter of adjacent identified individuals with the preset hair volume per square centimeter difference of the system to obtain a hair volume characteristic value a; if the hair volume characteristic value is negative, take its absolute value; Step S220: Compare the hair color feature of the identified person with a system preset female hair color classification, wherein the system preset female hair color classification includes a first hair color category and a second hair color category, wherein the first hair color category is provided with a first hair color threshold, and the second hair color category is provided with a second hair color threshold. If the hair color feature of the identified person is a system preset female hair color classification, then the hair color feature value of the identified person is b, where b={first hair color threshold, second hair color threshold}. Step S230: Compare the length feature value of the identified person's hairstyle with the length feature cutoff value preset by the system to obtain a length feature difference value c. When c is a negative value, its absolute value is taken; when c is a positive value, it is taken as 0; Step S240: Calculate the weighted average using a in step S210, b in step S220, and c in step S230, assuming the weight of the hair volume feature is a1, the weight of the hair color feature is b1, and the weight of the length feature is c1; using the formula: The characteristic value of the identified person as male is obtained as E, where the weight is 10>>a1>c1>b1>>1; let the system preset male characteristic value be E'. When the obtained male characteristic value is greater than E and is greater than the system preset male characteristic value E', the identified person is male; otherwise, it is female.

4. The Internet-based facial recognition police information tracking platform according to claim 3 is characterized by: In step S300, the eye pattern feature is identified, and the specific process is as follows: Step S310: The face information extraction unit extracts the eye image to obtain a detailed eye image, and uses the detailed eye image to determine whether there are eye wrinkles. If not, the person is young; if so, the determination continues. Step S320: The face information extraction unit restores the eye detail image into a black and white image, which contains all eye lines e i , i={1,2,3,.....c}, and eye pattern e i Arrange them in the order of 1, 2, 3, .....c from top to bottom, and the eye pattern in the black and white image is a black image. Amplify the black and white image by a certain multiple, and calculate the area of ​​the black image in the amplified image; Step S330: Eye pattern e i The longest horizontal line in the rectangle is the length W1 of the rectangle, and the distance from the highest point of eye pattern e1 to the highest point of eye pattern e is c The vertical distance W2 of the lowest point is used as the width of the rectangle to form an eye pattern rectangle. The area of ​​the eye pattern rectangle is W. Then calculate the area of ​​each eye pattern e separately. i The area w i , the eye pattern i After magnification, it becomes a triangle, and each eye pattern is e i The area of ​​the eye pattern gathering point is the vertex, and the vertex is on the edge of the eye pattern matrix. The face information extraction unit extracts the vertex to the eye pattern e i The vertical distance l from the vertex to the base i and the bottom distance d i ; Using the formula: W = W1 × W2; Among them, k is the ratio coefficient of eye wrinkle depth; Step S340: The face information extraction unit presets a standard threshold value K for middle-aged eye lines; when the eye line depth ratio coefficient k is greater than the standard threshold value K for middle-aged eye lines, the identified person is determined to be an elderly person, otherwise, the person is middle-aged.

5. The Internet-based facial recognition police information tracking platform according to claim 4 is characterized by: In step S340, when the eye wrinkle depth ratio coefficient k is close to the middle-aged eye wrinkle standard threshold K, the age of the identified person cannot be accurately determined, and the following steps are performed: Step S341: The face information extraction unit extracts the eye area image below the eyebrows and above the cheekbones, and performs area processing on the eye area image. The face information extraction unit extracts the eye area of ​​m identified people and calculates the average eye area. As the standard eye area value; Step S342: The eye periorbital image is binarized, the grayscale value of the eye wrinkles is set to 255, and the grayscale values ​​of the rest of the image except the eye wrinkles are set to 0; the face information extraction unit extracts the number of eye wrinkles as h, and calculates the density value of the eye wrinkles under the standard eye periorbital area. Step S343: Compare the density value in step S342 with the number of eye wrinkles per square centimeter preset by the system. When the density value γ under the standard eye area is greater than the system preset eye wrinkle density value, , it is further determined that the identified person is an elderly person; if the density value γ under the standard eye area is less than the eye pattern density value preset by the system , it is further determined that the identified person is a middle-aged person.

6. The Internet-based facial recognition police information tracking platform according to claim 5 is characterized by: The identity determination module determines the identity of the person while the face information extraction unit extracts the facial information. The identity determination module records the gender, hair volume, hair color, length, and eye wrinkle depth of the person to be identified to form the first element of the identity table I; At the same time, the facial information extraction unit transmits a signal to the real-time monitoring module, and the real-time monitoring module identifies and monitors the whereabouts of the identified person. The real-time monitoring module transmits the information of the whereabouts of the identified person to the identity establishment unit. The whereabouts of the identified person constitutes the second element of the identity table I, and the identity table I is composed of the first element and the second element; the information of the first identification of the identified person by the identity table I is recorded as the initial identity table I1. The identity table I is used to extract and determine the identity feature information of the on-site personnel when an alarm occurs.

7. The Internet-based facial recognition police information tracking platform according to claim 6 is characterized by: The police center module includes a police information receiving unit and a police dispatch plan formulation unit; the police information receiving unit is used to receive and process alarm information, and the police dispatch plan formulation unit is used to formulate a dispatch plan based on the processed alarm information; the specific implementation process of the police information receiving unit is as follows: The alarm receiving unit receives the alarm information, and the alarm receiving unit disassembles the audio information of the alarm person into n audio units, wherein the audio unit is the audio of two adjacent words, and the alarm receiving unit analyzes the time length t in each audio unit. n , n={1,2,3......}, t n Indicates the time length of the nth audio unit, using the formula: in Indicates the average duration of each audio unit in the audio information; The alarm receiving unit presets the time length of each audio unit as when When the alarm receiving unit determines that the alarm is a non-emergency situation; when When the alarm situation is received, the alarm situation receiving unit determines that the alarm situation is an emergency situation; the alarm situation receiving unit transmits the alarm situation to the police dispatch plan formulation unit.

8. The Internet-based facial recognition police information tracking platform according to claim 7 is characterized by: The specific implementation process of the police dispatch plan formulation unit is as follows: The police dispatch plan formulation unit arranges non-emergency situations in chronological order and dispatches them according to the idle police ratio set by the system; the police dispatch plan formulation unit directly prioritizes emergency situations and dispatches them; at the same time, the police dispatch plan formulation unit retrieves the monitoring information from the real-time monitoring unit at the scene of the situation to make a judgment: if the monitoring information indicates a traffic jam, the police dispatch plan formulation unit will give priority to dispatching police officers driving police motorcycles to respond; If the traffic is normal, the police dispatch plan formulation unit dispatches idle police cars for dispatch; When the monitoring screen involves a public place with more monitoring screens, the real-time monitoring module retrieves the latest identity table of the identified person. 1i Information, the police dispatch plan formulation unit according to the latest identity table I 1i The first element judgment: when the identified person is an elderly person, the police dispatch plan formulation unit determines whether to call the 120 emergency number and whether to prepare an elderly first aid kit; when the identified person is a female, it determines whether she is pregnant and needs to call the 120 emergency number, and whether to dispatch a female police officer for the police response; when the hair color and length characteristics of the identified person change every time the historical records of the latest identity table are updated, the police dispatch plan formulation unit determines whether there are any abnormal problems with the person.

9. The Internet-based facial recognition police information tracking platform according to claim 8, characterized in that: The police dispatch plan formulation unit is based on the latest identity table I 1i The second factor judgment in: the police dispatch plan formulation unit extracts the latest identity table I by identifying the first factor 1i The second element is to determine the whereabouts of the identified person; The police dispatch plan formulation unit will identify the latest identity table I of the person 1i The whereabouts of the track are checked and the latest identity table I 1i The location represented by the same image in the historical record is taken as the starting point g, and the latest identity table I 1i The initial endpoint in is recorded as x1, and the latest identity table I 1i The other endpoints except the initial endpoint are recorded as x i , calculate the straight-line distance between each end point and the starting point: x1-g, x2-g, ...x i-1 -g, and the straight-line distance between each end point and the initial end point: x i-1 -x1...x2-x1; using the formula: a i-1 =arccos i-1 where α i-1 is the angle of the i-th movement trajectory, The average angle of the identified person's whereabouts trajectory, the police officer's response plan formulation unit presets the standard whereabouts trajectory angle threshold of the person's activity range like The police dispatch plan formulation unit determines that the identified person is abnormal in activity and records the whereabouts of the person; if The police dispatch plan formulation unit determines that the identified person is engaged in normal activities and does not record it.

Images