CN112149660A - Gun recognition system - Google Patents

Abstract

The invention provides a gun identification system, which includes a gun information collection subsystem, a gun information database, and a central processing control subsystem; a gun information collection subsystem is used to collect the data of the gun and send the data to the central processing control subsystem; The firearms information database is used to store the data of firearms and to carry out data interaction with the central processing and control subsystem; the central processing and control subsystem is used for management and firearms identification. This scheme can identify a variety of different firearms, and has a wide range of applications. In terms of specific identification methods, the database is quickly screened by the area, perimeter, and maximum distance between two points of the contour image, and then combined with the contour image. The Hu moment matching algorithm further screens the database, and finally the inspectors conduct manual screening and identification of the returned results. Through the combination from intuitive to complex, and from machine to manual, the efficiency and accuracy of identification are ensured.

Description

A gun identification system

technical field

The invention relates to the technical field of gun identification, in particular to a gun identification system.

Background technique

Guns are powerful tools of violence, and strict control of guns is an important requirement for maintaining national security and social stability. Every year, my country's customs seize a large number of illegally smuggled guns (including mass-destruction air guns) and bullets. The rapid and accurate identification and inspection of guns and bullets is very important for the study, analysis and detection of the case.

However, the production sites of the seized smuggled firearms are unknown, with various types, numerous models and complex structures, which have brought great difficulties to the identification and identification of firearms. Experience and relevant departments lack a complete airsoft database, and new challenges are faced in the identification, identification and control of firearms. The traditional firearm identification and identification relying solely on human eyes and experience can no longer meet the needs of the new situation, and it is urgent to develop a new method for rapid identification and identification of firearms. At present, there are also some technical solutions to try to solve this technical problem:

The invention patent with the application number CN201811120409.1 discloses a firearm identification management device. It includes a casing, a controller and a radio frequency identifier. The bottom end of the casing is movably installed with a charging mechanism. The interior of the casing is fixedly installed with a battery, and the battery is located at the top of the charging mechanism. The controller is fixedly installed, and the controller is located at the top of the battery, the top of the inside of the casing is fixedly installed with a radio frequency identifier, the top of the casing is movably installed with a cleaning mechanism, and the top of the casing surface is fixedly installed with a A display screen, a control button is fixedly installed on the surface of the casing, and the control button is movably connected with the controller, and an auxiliary mechanism is movably installed on the bottom end of the casing surface. In addition, the invention patent with the application number CN200710090704.2 proposes a gun identification management system. The system is to install an information card on the firearm and place it on the information read-write handle for firearm information management. The system includes a firearm registration module, a firearm cancellation module, a firearm information query module, a personnel registration module, a personnel cancellation module, and personnel information query. Module, fingerprint borrowing and returning gun module, identity card borrowing and returning gun module, borrowing and returning gun information query module.

Specifically, the invention patent with the application number CN201811120409.1 obtains the information of the gun by scanning the electronic label affixed on the surface of the gun with the device. As for the invention patent with the application number CN200710090704.2, it is obtained by placing an information card on the gun , The information reading handle reads the information card to obtain the gun information. These methods not only require the electronic label (information card) and the electronic label (information card) reading device, the hardware cost increases, and can only be used for the identification of internal supervision guns, and cannot be used. Applied to the identification of seized illegal firearms, it cannot meet the needs of rapid identification of smuggled firearms.

Therefore, a better solution is currently required to solve this technical problem.

SUMMARY OF THE INVENTION

In view of the defects of the prior art, the present invention proposes a firearm identification system, which can perform graphical identification of various firearms. In the specific identification method, the area, perimeter, and distance between two points of the contour image are first analyzed. The database is quickly screened with data such as the maximum distance, and then combined with the Hu moment matching algorithm of contour images, the database is further screened, and finally the inspection personnel manually screen and identify the returned results. Efficient and accurate identification.

Specifically, the present invention proposes the following specific embodiments:

An embodiment of the present invention provides a gun identification system, including a gun information collection subsystem, a gun information database, and a central processing control subsystem; wherein,

The gun information collection subsystem is used to collect the data of the gun and send the data to the central processing control subsystem;

The firearm information database is used to store firearm data and perform data interaction with the central processing control subsystem;

The central processing control subsystem is used for binarizing the contour image of the gun to be identified at a specified viewing angle to obtain a binarized contour image;

Determine the size coefficient according to the focal length of the camera that captures the contour image and the distance between the camera and the firearm to be identified;

determining the area, perimeter and maximum distance between two points of the area occupied by the gun in the binarized contour image;

Matching in the firearm information database according to the size factor, the area, the perimeter and the maximum distance between the two points;

If the candidate firearm data set is obtained by matching, the Hu moment calculation of the binarized contour image is performed on each firearm in the candidate firearm data set and the firearm to be identified;

Calculate a weighted squared error based on the Hu moments;

Screening firearms matching the Hu moments of the firearms to be identified in the candidate firearms data set based on the weighted squared error;

If the screening firearm set is obtained, the firearms in the screening firearm set are sorted according to the weighted square error from low to high, and the data of the sorted firearms are returned to the inspector, so that the inspector can perform manual identification.

In a specific embodiment, the outline image is an outline grayscale image;

The area with a grayscale value of 0 in the binarized contour image is the gun area, and the area with a grayscale value of 255 is the background area; the area corresponds to all pixels with a grayscale value of 0 in the binarized contour image The size of the area corresponding to the point; the perimeter is the length integral of the edge of the area with a gray value of 0 in the binarized contour image; the maximum distance between the two points is the grayscale in the binarized contour image The maximum distance between pixels with a value of 0;

The Hu moment is obtained by calculating the contour binarized image according to the definition of Hu moment.

In a specific embodiment,

The size factor is calculated based on the following formula:

k=H/f; wherein, k is the size coefficient; the H is the distance between the camera and the gun to be identified in the imaging direction; the f is the focal length of the imaging lens of the camera;

The matching is based on the following formula:

其中，

in,

The area of the binarized contour image of the firearm to be identified is S, the perimeter of the binarized contour image of the firearm to be identified is C, and the maximum distance between two points of the firearm to be identified is L; The size coefficient for identifying the gun is k; the area of the gun in the gun information database is S ₁ , the perimeter of the gun is C ₁ , the maximum distance between two points of the gun is L ₁ , the gun The size coefficient of is k ₁ ; γ _s , γ _C , γ _L are the error control coefficients of area, perimeter, and maximum distance between two points, respectively;

The weighted squared error is calculated by the following formula:

Wherein, the Hu moment of the contour binarized image of the firearm in the candidate firearm data set is M; the Hu moment of the contour binarized image of the firearm to be identified is N; k represents different components of the Hu moment; w is Hu The weights of different components of the moment; WSE is the weighted squared error;

The set of firearms obtained by screening is the set of firearms in the candidate firearm data set that conforms to WSE< _WT , where _WT is a preset threshold.

In a specific embodiment, the gun information collection subsystem includes: a basic attribute collection module, a contour image collection module, a color image collection module, and an information collection computer with information collection software.

In a specific embodiment, the basic attribute collection module is used to collect basic attribute information of firearms;

The basic attribute information includes any combination of one or more of the following: name, model, surface identification, LOGO text description, size, caliber, structure, type, production country information, manufacturer information, and production year.

In a specific embodiment, the profile image acquisition module is used to acquire profile images of the firearm.

In a specific embodiment, the color image acquisition module is used to acquire a color image of the appearance of the gun.

In a specific embodiment, the contour image acquisition module or the color image acquisition module is an image acquisition device;

The image acquisition device includes: a liftable support frame, a three-dimensional adjustable support frame, a frosted glass plate for placing guns, a high-definition color camera, a backlight source, and a white light illumination source; wherein,

the three-dimensional adjustable bracket is arranged on the top of the support frame;

The frosted glass plate is horizontally arranged in the middle of the support frame;

The backlight light source is arranged below the frosted glass plate, and illuminates the frosted glass plate from bottom to top;

The high-definition color camera and the white light illumination light source are both arranged on the three-dimensional adjustable bracket.

In a specific embodiment, the central processing control subsystem includes: a system management module, a firearm information management module, and a firearm identification module; wherein,

The system management module is used for account and authority management, system status monitoring, data query, and report management;

The firearm information management module is used for firearm information collection, data transmission, filing, addition, deletion, modification, query and retrieval, and statistics;

The firearm identification module is used for identifying firearms to be identified.

In a specific embodiment, the gun data stored in the gun information database includes: serial number, binarized contour image, area, perimeter, maximum distance between two points, size coefficient, and color image.

In a specific embodiment, the firearm data stored in the firearm information database further includes: firearm name, model, type, structure, production country, manufacturer, production year, and identification.

Therefore, compared with the prior art, the present invention has the following effects: it can perform graphical identification for various guns, and has a wide range of applications. In terms of specific identification methods, the area, perimeter, The database is quickly screened with data such as the maximum distance between two points, and then combined with the Hu moment matching algorithm of contour images, the database is further screened, and finally the inspection personnel manually screen and identify the returned results. , to ensure the efficient and accurate identification.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore do not It should be regarded as a limitation of the scope, and for those of ordinary skill in the art, other related drawings can also be obtained according to these drawings without any creative effort.

FIG. 1 is a schematic diagram of the basic structure of a gun identification system proposed by an embodiment of the present invention;

2 is a schematic diagram of the basic structure of a gun information collection subsystem in a gun identification system proposed by an embodiment of the present invention;

FIG. 3 is a schematic diagram of a gun outline image in a gun identification system proposed by an embodiment of the present invention;

4 is a schematic structural diagram of an image acquisition device in a gun identification system according to an embodiment of the present invention;

FIG. 5 is a basic composition framework of a central processing control subsystem in a gun identification system proposed by an embodiment of the present invention;

FIG. 6 is a schematic diagram of the format definition of gun information data in a gun identification system according to an embodiment of the present invention.

Detailed ways

Hereinafter, various embodiments of the present disclosure will be described more fully. The present disclosure is capable of various embodiments, and adaptations and changes may be made therein. It should be understood, however, that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be construed to cover various embodiments falling within the spirit and scope of the present disclosure. All adjustments, equivalents and/or alternatives.

The terminology used in the various embodiments of the present disclosure is for the purpose of describing particular embodiments only and is not intended to limit the various embodiments of the present disclosure. As used herein, the singular is intended to include the plural as well, unless the context clearly dictates otherwise. Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which various embodiments of this disclosure belong. The terms (such as those defined in commonly used dictionaries) will be interpreted as having the same meaning as the contextual meaning in the relevant technical field and will not be interpreted as having an idealized or overly formal meaning, unless explicitly defined in various embodiments of the present disclosure.

Example

An embodiment of the present invention discloses a gun identification system, as shown in FIG. 1 , including a gun information collection subsystem, a gun information database, and a central processing control subsystem; wherein,

The gun information collection subsystem is used to collect the data of the gun and send the data to the central processing control subsystem;

The firearm information database is used to store firearm data and perform data interaction with the central processing control subsystem;

Specifically, as shown in Figure 1, the system is mainly composed of a gun information collection subsystem, a central processing control subsystem, and a gun information database. The gun information collection subsystem is responsible for collecting the basic information and two-dimensional images of guns, and transmits these data to the central processing control subsystem. The firearm information database is responsible for storing firearm information and exchanging information with the central processing control subsystem.

The central processing and control sub-system is responsible for information processing and management control, and is the center for the identification and identification of firearms and the coordination and control of the sub-systems. Specifically, the central processing control subsystem is used to perform the following functions:

Function 1. Identification and identification of guns. Specifically, the gun information collection subsystem collects the contour images of the guns to be identified, and transmits the contour images to the central processing control subsystem. The central processing control subsystem analyzes the contour according to the intelligent analysis process and algorithm. The image is processed and feature extraction is performed, and the firearm information database is retrieved and compared. If the comparison is successful, the specific information of the matched firearm will be returned, and then the inspector will observe and analyze the color image of the matched firearm, and visually identify the matched firearm and the to-be-identified firearm. Whether the firearm matches, if it matches, it is judged that the identification is successful, otherwise it is judged that there is no such firearm data in the database.

Function 2. Gun information management. Including gun information collection, database establishment, data improvement, data modification, data deletion, data query, data statistics and other gun-related information management.

Specifically, the basic process of the firearm identification method is as follows: obtaining a contour image of the firearm to be identified at a specified viewing angle, and performing a binarization process on it to obtain a binarized contour image of the firearm. Calculate the size factor k according to the height ratio of the camera at the time of shooting. Calculate the area S, perimeter C, and maximum distance L between two points of the area occupied by the gun in the binarized contour image. The firearms information database is retrieved and compared, and the firearms data matching the area S, the perimeter C, and the maximum distance L between the two points are selected to form the candidate firearms data set A. If A is not empty, calculate the Hu moments of the firearm to be identified and the candidate firearm, and calculate the weighted squared error WSE. Build a set B of guns with WSE < threshold _WT . If B is not empty, arrange the guns according to the WSE from low to high and return them in sequence. The inspectors can visually identify whether the matching guns match the to-be-identified guns by observing and analyzing the color images of the matching guns. If they match, then If the identification is successful, if all the firearms do not match, it is judged that there is no firearm data to be identified in the database.

Specifically, taking the identification of firearms as an example, the central processing control subsystem is used to perform binarization processing on the contour image of the firearm to be identified at a specified viewing angle to obtain a binarized contour image; specifically, the contour image is a contour Grayscale image; the area with the grayscale value of 0 in the binarized contour image is the gun area, and the area with the grayscale value of 255 is the background area; the area corresponds to the grayscale value in the binarized contour image. The size of the region corresponding to all pixels of 0; the perimeter is the length integral of the edge of the region with a gray value of 0 in the binarized contour image; the maximum distance between the two points is the binarized contour The maximum distance between pixels with a gray value of 0 in the image;

Determine the size coefficient according to the focal length of the camera that captures the contour image and the distance between the camera and the firearm to be identified;

Specifically, the size factor is calculated based on the following formula:

k=H/f; wherein, k is the size coefficient; the H is the distance between the camera and the gun to be identified in the imaging direction; the f is the focal length of the imaging lens of the camera;

determining the area, perimeter and maximum distance between two points of the area occupied by the gun in the binarized contour image;

matching in the firearm information database according to the size factor, the perimeter, the area and the maximum distance between the two points;

If the candidate firearm data set is obtained by matching, the Hu moment calculation of the binarized contour image is performed on each firearm in the candidate firearm data set and the firearm to be identified; The transformed image is calculated according to the definition of Hu moment;

Calculate a weighted squared error based on the Hu moments;

Wherein, the matching is based on the following formula:

其中，

in,

The area of the binarized contour image of the firearm to be identified is S, the perimeter of the binarized contour image of the firearm to be identified is C, and the maximum distance between two points of the firearm to be identified is L; The size coefficient for identifying the gun is k; the area of the gun in the gun information database is S ₁ , the perimeter of the gun is C ₁ , the maximum distance between two points of the gun is L ₁ , the gun The size coefficient of is k ₁ ; γ _s , γ _C , γ _L are the error control coefficients of area, perimeter, and maximum distance between two points, respectively;

The weighted squared error is calculated by the following formula:

Wherein, the Hu moment of the contour binarized image of the firearm in the candidate firearm data set is M; the Hu moment of the contour binarized image of the firearm to be identified is N; k represents different components of the Hu moment; w is Hu The weights of different components of the moment; WSE is the weighted squared error;

The set of firearms obtained by screening is the set of firearms in the candidate firearm data set that conforms to WSE< _WT , where _WT is a preset threshold. If the screening firearm set is obtained, the firearms in the screening firearm set are sorted according to the weighted square error from low to high, and the data of the sorted firearms are returned to the inspector, so that the inspector can perform manual identification.

In the actual process, for the firearms to be identified, basic attribute information, contour images and color images are collected on the firearms information collection subsystem. If the contour image was taken with a color camera, first convert the color image to grayscale. For contour grayscale images, an appropriate threshold T is designed, so that the grayscale value of pixels whose grayscale value is greater than or equal to T is reset to 255; as for the pixels whose grayscale value is less than T, the grayscale value is reset to 0; thus Obtain a binarized contour image p. The grayscale value of the gun area in the image is 0, and the rest of the background is 255. The threshold value T is selected according to experience. The focal length of the imaging lens of the contour image acquisition camera is f, the distance between the camera and the gun in the imaging direction is H, and the size coefficient k=H/f is calculated. For the binarized contour image, calculate the area S, perimeter C, and maximum distance L between two points of the gun area. The area S is defined as the sum of all pixels with a gray value of 0 in the image, the perimeter C is defined as the length integral of the edge of the area with a gray value of 0 in the image, and the maximum distance L between two points is defined as the gray level in the image. The maximum distance between pixels with a value of 0.

The firearm data in the firearms information database is retrieved. The area of the firearm to be retrieved is S ₁ , the perimeter of the firearm is C ₁ , the maximum distance between two points is L ₁ , and the size coefficient is k ₁ . Find gun data that satisfy the following relationship: Among them, γ _s , γ _C , γ _L are the error control coefficients of area, perimeter, and the maximum distance between two points, respectively, mainly considering the processing error of gun parts, gun assembly error, wear error, image acquisition and calculation. effects of errors, etc. The thresholds γ _S , γ _C and γ _L are chosen empirically.

Through the above method, find the firearm data set that matches the area S, area C, and maximum distance L between two points in the area of the firearm to be identified, denoted as A. If A is empty, it means that there is no gun in the gun information database that matches the area, perimeter, and the maximum distance between the two points in the area. It is judged that there is no gun that matches the gun to be identified, and the result is returned. If A is not empty, for each gun in the set, read its contour binarized image m to obtain the Hu moment. The Hu moment is calculated from the second-order and third-order central moments of the image, and has image scaling, translation, rotation, and mirror image invariance. There are altogether 7 Hu moments: M={M ₁ , M ₂ , M ₃ , M ₄ , M ₅ , M ₆ , M ₇ }. For the gun to be identified, read its contour binarized image n, and obtain the Hu moment for it: N={N ₁ , N ₂ , N ₃ , N ₄ , N ₅ , N ₆ , N ₇ }. Calculate the weighted squared error of M and N

The Hu moment of the contour binarized image of the guns in the candidate gun data set is M; the Hu moment of the contour binarized image of the gun to be identified is N; k represents different components of Hu moments; w is different Hu moments The weights of the components; WSE is the weighted squared error; the weighted squared error WSE is calculated for all guns in set A. Design an appropriate threshold _WT to establish a gun set B with WSE< _WT . If B is empty, it means that there is no firearm in the firearm information database that matches the Hu moment of the binarized contour image of the firearm to be identified, judges that there is no firearm that matches the firearm to be identified, and returns the result. The threshold _WT is selected according to experience. If B is not empty, sort the guns in B by WSE from low to high, and return the detailed information in order.

Finally, the inspectors conduct manual identification and identification of the returned firearm information. The method of manual identification and identification is to observe the high-resolution color images of the front and back of the returned firearm, and the high-resolution color images of the front and back of the firearm to be identified to determine whether it belongs to The same gun. If it is determined that they belong to the same type of firearms, the identification and comparison are successful. If it is determined that all the returned firearms do not match the firearms to be identified, there is no firearms information to be identified in the database.

The following describes each part of the entire gun identification system. The gun information collection subsystem includes: a basic attribute collection module, a contour image collection module, a color image collection module, and an information collection computer with information collection software. Wherein, the basic attribute collection module is used to collect the basic attribute information of the gun; the basic attribute information includes any combination of one or more of the following: name, model, surface identification, LOGO text description, size, caliber, structure, Type, country of manufacture information, manufacturer information, production year. As for the profile image acquisition module, it is used to acquire profile images of guns. The color image acquisition module is used for acquiring color images of the appearance of the gun.

The gun information collection subsystem is mainly composed of basic attribute collection module, contour image collection module, color image collection module, information collection computer and information collection software. As shown in Figure 2. Among them, the basic attribute information collection module is responsible for collecting the basic attribute information of the gun, such as the name, model, surface identification, LOGO text description, size, caliber, structure, type, production country, manufacturer, production age, etc. of the gun. For the firearms to be collected, some of the above information is known and some is unknown. For example, for the seized illegal firearms, most of the information may be unknown, while for the supervision of firearms, most of the information may be known. . Therefore, the collection of information is selective (collect it if it can be determined, and not if it cannot be determined).

The contour image acquisition module is responsible for acquiring the contour image of the gun. The so-called contour image is to use the back-illuminated illumination imaging method to image the surface of the gun. The area covered by the gun in the image will be almost completely dark, while the area not covered by the gun will be almost completely bright. An example of a gun profile image is shown in Figure 3. The contour image acquisition module mainly includes a backlight source, a frosted glass plate, a three-dimensional adjustable bracket, and a high-definition camera. The guns to be collected are placed on the frosted glass plate, and the backlight light source is installed under the frosted glass plate to illuminate upwards from the bottom of the frosted glass plate. The backlight source can be a single large-area light source or an array of illumination light sources. The high-definition camera images the gun from above the frosted glass plate facing down. The camera is installed on a three-dimensional adjustable bracket, which can be adjusted in two dimensions in the direction parallel to the ground glass plate and one-dimensional adjustment in the direction perpendicular to the ground glass plate. By adjusting the position of the camera in the direction parallel to the frosted glass, the gun is basically located in the center of the imaging field of view of the camera, and by adjusting the position of the camera in the direction perpendicular to the frosted glass, the gun fills most of the imaging field of view of the camera. The high-definition camera consists of a large depth-of-field imaging lens and a large area array high-pixel resolution industrial camera. Industrial cameras can be black and white cameras or color cameras.

The color image acquisition module is responsible for acquiring color images of the appearance of the gun. The color image acquisition module mainly includes a white light illumination source, a base plate, a three-dimensional adjustable bracket, and a high-definition color camera. The guns to be collected are placed on the bottom plate, the white light source illuminates downward from the top of the bottom plate, and the high-definition color camera images the guns from above the bottom plate. The color camera is installed on a three-dimensional adjustable bracket, which can be adjusted in two dimensions in the direction parallel to the bottom plate and one-dimensional adjustment in the direction perpendicular to the bottom plate. By adjusting the position of the camera in the direction parallel to the base plate, the gun is basically located in the center of the imaging field of view of the camera, and by adjusting the position of the camera in the direction of the vertical base plate, the gun fills most of the imaging field of view of the camera. The high-definition color camera consists of a large depth-of-field imaging lens and a large area array high-pixel resolution color industrial camera.

In a specific embodiment, the contour image acquisition module and the color image acquisition module may be combined into one, and the contour image acquisition module or the color image acquisition module is an image acquisition device; the image acquisition device Including: a liftable support frame, a three-dimensional adjustable support frame, a frosted glass plate for placing guns, a high-definition color camera, a backlight source, and a white light illumination source; wherein the three-dimensional adjustable support frame is arranged on the top of the support frame (specifically The support frame itself can be raised and lowered, and the three-dimensional adjustable support can also be raised and lowered); the frosted glass plate is horizontally arranged in the middle of the support frame; the backlight light source is arranged below the frosted glass plate, illuminating from bottom to top The frosted glass plate; the high-definition color camera and the white light illumination source are all arranged on the three-dimensional adjustable bracket.

The image acquisition device proposed by the present invention combines a contour image acquisition module and a color image acquisition module into one. The structure of the image acquisition device is shown in FIG. 4 . The image acquisition device is composed of a support frame, a frosted glass plate, a backlight source, a white light illumination source, a three-dimensional adjustable bracket, and a high-definition color camera. The gun to be collected is placed in the center of the frosted glass plate. The backlight source is located under the frosted glass plate, and illuminates the frosted glass plate from below. The high-definition color camera is installed on a three-dimensional adjustable bracket. The three-dimensional adjustable bracket can be adjusted in two dimensions in the direction parallel to the ground glass plate, so as to adjust the center position of the camera's field of view, and can be adjusted in one dimension in the direction perpendicular to the ground glass plate to adjust the ground glass plate. The size of the imaging area of the camera. The position of the camera is adjusted by the three-dimensional adjustable bracket so that the gun is substantially centered on the camera's imaging and fills most of the field of view. Two white light illumination sources are also installed on the three-dimensional adjustable bracket, located on both sides of the high-definition color camera, and illuminate the frosted glass plate from above. The adjustment position of the three-dimensional adjustable bracket can be obtained by the scale reading.

The process of image acquisition by the image acquisition device is as follows: place the gun to be collected in the center of the frosted glass plate, turn on the backlight source and the high-definition color camera, and adjust the position of the camera through the three-dimensional adjustable bracket, so that the gun is basically located in the center of the camera's imaging field of view and fills most of it. field of view. Take an outline image of the gun and save it. Turn off the backlight, turn on the white light source, take a color image of the gun and save, turn the gun upside down, take a color image of the gun and save. Read the position scale of the three-dimensional adjustable bracket in the vertical frosted glass direction, calculate the distance H between the camera and the gun in the height direction and save it. The firearm information collection subsystem may also include an information collection computer and collection software. The acquisition software is installed on the acquisition computer.

In a specific embodiment, the central processing control subsystem includes: a system management module, a gun information management module, and a gun identification module; wherein, the system management module is used for account and authority management, system status monitoring, data Inquiry and report management;

The firearm information management module is used for firearm information collection, data transmission, filing, addition, deletion, modification, query and retrieval, and statistics;

The firearm identification module is used for identifying firearms to be identified.

Specifically, the central processing control subsystem, as shown in FIG. 5 , is mainly composed of a central processing server and control processing software. The central processing server has powerful computing power. According to application requirements and layout conditions, it can be a local physical server or a cloud server. The control processing software is responsible for realizing the core control, management and information processing of the system. According to different realization functions, the control processing software is mainly divided into the following modules: (1) System management module. Including account and authority management, system status monitoring, data query, report management, etc. (2) Gun information management module. Mainly is the management of firearms information database, including firearms information collection, data transmission, filing, addition, deletion, modification, query retrieval, statistics, etc. (3) Gun identification module. Responsible for using image processing and intelligent identification algorithms to process and extract features from the collected two-dimensional images of firearms, and to retrieve and compare the data with the firearms information database, and finally realize the rapid identification of firearms.

In a specific embodiment, the gun data stored in the gun information database includes: serial number, binarized contour image, area, perimeter, maximum distance between two points, size coefficient, and color image. In addition, the firearm data stored in the firearm information database also includes: firearm name, model, type, structure, production country, manufacturer, production year, and identification.

Specifically, the firearms information database is mainly composed of a database server and database system management software. The database server has powerful storage capacity and fast data access speed. According to application requirements and layout conditions, it can be a local physical server or a cloud server. Database management software such as Oracle, SQL Server, MySQL, etc. are installed on the server. The format of gun information data is defined as follows: For the whole gun, the data should at least include items such as serial number, binarized contour image, area, perimeter, maximum distance between two points, size factor, color image, etc. (required), and can include guns Items such as name, model, type, structure, country of manufacture, manufacturer, year of manufacture, and logo (optional). The format definition of firearm information data can be as shown in Figure 6.

The system functions and working modes are described in detail below.

For the function of gun identification, for the guns to be identified, basic attributes, contour images, and color images are collected on the gun information collection subsystem, and the above collected data is packaged and sent to the central processing control subsystem through the information collection software. The sub-system searches and compares the firearm information database to see if it matches the above-mentioned collected data; if matching data is found, the specific information of the matched firearm will be returned, and then the inspectors will observe and analyze the color images of the matched firearms to visually identify and identify the matched firearms and the pending firearms. Identify whether the firearms match, and if they match, determine that the identification is successful, otherwise, determine that there is no firearm data in the database.

For the functions of gun information management, 1. For newly added guns, basic attribute information, contour images, and color images are collected on the gun information collection subsystem, and the above collected data is packaged and sent to the central processing control subsystem through the information collection software. , the central processing and control sub-system searches and compares the firearm information database to see if it matches the above-mentioned collected data; if no matching data is found, it means that the database has not created a file for the type of firearm, and a new firearm data file is created in the database. 2. According to the development of system application, gradually update the firearm data in the database, including data supplement, modification, deletion, etc. 3. Query, count, supervise, and analyze information on firearms, such as type statistics, structure statistics, and source statistics, and can correlate gun-related cases.

Compared with the prior art, the advantages of this solution include: (1) The image recognition method is adopted, which can realize the recognition of various firearms. (2) Based on basic attributes, contour image information, color images and other information, the multi-pronged approach plays a role together to improve the accuracy and reliability of recognition. (3) In the identification method, the database is quickly screened by the area, perimeter, and the maximum distance between two points of the contour image, and then the Hu moment feature matching algorithm of the contour image is used to further screen the database. The results are returned for manual screening and identification. Through the combination from intuitive to complex, and from machine to manual, the efficient and accurate identification is guaranteed.

Those skilled in the art can understand that the accompanying drawing is only a schematic diagram of a preferred implementation scenario, and the modules or processes in the accompanying drawing are not necessarily necessary to implement the present invention.

Those skilled in the art can understand that the modules in the device in the implementation scenario may be distributed in the device in the implementation scenario according to the description of the implementation scenario, or may be located in one or more devices different from the implementation scenario with corresponding changes. The modules of the above implementation scenarios may be combined into one module, or may be further split into multiple sub-modules.

The above serial numbers of the present invention are only for description, and do not represent the pros and cons of the implementation scenarios.

The above disclosures are only a few specific implementation scenarios of the present invention, however, the present invention is not limited thereto, and any changes that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims (11)

1. A gun identification system is characterized by comprising a gun information acquisition subsystem, a gun information database and a central processing control subsystem; wherein,

the gun information acquisition subsystem is used for acquiring gun data and sending the data to the central processing control subsystem;

the gun information database is used for storing gun data and performing data interaction with the central processing control subsystem;

the central processing control subsystem is used for carrying out binarization processing on the contour image of the gun to be identified at a specified visual angle to obtain a binarized contour image;

determining a size coefficient according to a focal length of a camera for shooting the contour image and a distance between the camera and the gun to be identified;

determining the area and the perimeter of a region occupied by the gun in the binary contour image and the maximum distance between two points;

matching in the firearm information database according to the size coefficient, the area, the perimeter and the maximum distance between the two points;

if a candidate gun data set is obtained through matching, performing Hu moment calculation on each gun in the candidate gun data set and the gun to be identified;

calculating a weighted square error based on the Hu moment;

screening guns that match the Hu moment of the gun to be identified in the gun candidate data set based on the weighted square error;

and if the screened gun set is obtained through screening, sorting the guns in the screened gun set from low to high according to the weighted square error, and returning the data of the sorted guns to the inspector so as to facilitate the inspector to perform manual identification.

2. A gun recognition system as claimed in claim 1 wherein said outline image is a gray-scale outline image;

the area with the gray value of 0 in the binary contour image is a gun area, and the area with the gray value of 255 is a background area; the area corresponds to the size of an area corresponding to all pixel points with the gray scale value of 0 in the binary contour image; the circumference is the length integral of the edge of the area with the gray value of 0 in the binaryzation contour image; the maximum distance between the two points is the maximum distance between pixel points with the gray value of 0 in the binary contour image;

the Hu moment is obtained by calculating the contour binary image according to the Hu moment definition.

3. A gun recognition system according to claim 1 or 2,

the size factor is calculated based on the following formula:

k is H/f; wherein k is a size coefficient; the H is the distance between the camera and the gun to be identified in the imaging direction; f is the focal length of an imaging lens of the camera;

the matching is based on the following formula:

|k₁C₁/kC-1|<γ_C,|k₁L₁/kL-1|<γ_L(ii) a Wherein,

the area of the binarization contour image of the gun to be identified is S, the perimeter of the binarization contour image of the gun to be identified is C, and the maximum distance between two points of the gun to be identified is L; the size coefficient of the gun to be identified is k; the area of the firearm in the firearm information database is S₁The perimeter of the gun is C₁The maximum distance between two points of the gun is L₁The gun has a size coefficient of k₁；γ_s、γ_C、γ_LError control coefficients of area, perimeter and maximum distance between two points are respectively;

the weighted square error is calculated by the following formula:

wherein the Hu moment of the contour binary image of the gun in the gun candidate data set is M; the Hu moment of the contour binary image of the gun to be identified is N; k represents the different components of the Hu moment; w is the weight of different components of the Hu moment; WSE is the weighted square error;

screening to obtain a screened gun set which is in accordance with WSE in the candidate gun data set<W_TA collection of firearms of wherein W_TIs a preset threshold.

4. A firearm identification system in accordance with claim 1 wherein said firearm information collection subsystem comprises: the system comprises a basic attribute acquisition module, a contour image acquisition module, a color image acquisition module and an information acquisition computer internally loaded with information acquisition software.

5. A firearm identification system in accordance with claim 4 wherein said basic attribute collection module is adapted to collect basic attribute information for a firearm;

the basic attribute information includes any combination of one or more of the following: name, model, surface identification, LOGO description, size, caliber, structure, type, country of production information, manufacturer information, and year of production.

6. A firearm identification system in accordance with claim 4 wherein said outline image capture module is adapted to capture an outline image of the firearm.

7. A firearm identification system in accordance with claim 4 wherein the color image capture module is adapted to capture a color image of the appearance of the firearm.

8. A gun identification system as claimed in claim 4 wherein said outline image capture module or said color image capture module is an image capture device;

the image acquisition apparatus includes: the device comprises a lifting support frame, a three-dimensional adjustable support frame, a ground glass flat plate for placing guns, a high-definition color camera, a backlight source and a white light illumination light source; wherein,

the three-dimensional adjustable bracket is arranged at the top of the support frame;

the ground glass flat plate is horizontally arranged in the middle of the support frame;

the back light source is arranged below the ground glass plate and irradiates the ground glass plate from bottom to top;

the high-definition color camera and the white light illuminating light source are both arranged on the three-dimensional adjustable support.

9. A firearm identification system in accordance with claim 1 wherein said central processing control subsystem comprises: the system comprises a system management module, a gun information management module and a gun identification module; wherein,

the system management module is used for account number and authority management, system state monitoring, data query and report management;

the gun information management module is used for acquiring gun information, transmitting data, filing, adding, deleting, modifying, inquiring, retrieving and counting;

and the gun identification module is used for identifying the gun to be identified.

10. A firearm identification system in accordance with claim 1 wherein said firearm information database stores firearm data comprising: serial number, binary outline image, area, perimeter, maximum distance between two points, size coefficient and color image.

11. A firearm identification system in accordance with claim 10 wherein said firearm information database stores firearm data further comprising: gun name, model, kind, structure, country of production, producer, year of production, sign.

Images