CN-115661734-B - Full-segment fuzzy matching-based packet association method

CN115661734BCN 115661734 BCN115661734 BCN 115661734BCN-115661734-B

Abstract

The invention discloses a full-segment fuzzy matching-based package association method which comprises the following steps of S1, uninterruptedly obtaining a current camera picture, extracting package image data, caching the package image data to obtain a package image data queue M, S2, obtaining a package X-ray picture which is completely imaged in a security inspection machine, extracting the imaging time of the package X-ray picture and recording as a time point t2, S3, carrying out time interval operation according to the time point t2 and the time point t to obtain a (t 2-delta t, t 2) time period, S4, updating package image data of the package image data queue M ', wherein the package image data queue M' comprises package image data of the time period (t 2-delta t, t 2), carrying out fuzzy matching on the package image data in the time period (t 2-delta t, t 2) and the package X-ray picture, and S5, obtaining package image data with the best adaptation degree as final package association image information and package X-ray picture to carry out package association binding according to a fuzzy matching result.

Inventors

BAI ZHENGBIN
ZHANG LI
JIANG XU
PU LEI
ZHAO LING
HUANG RUI
DENG JUN

Assignees

成都智元汇信息技术股份有限公司

Dates

Publication Date: 20260508
Application Date: 20220818

Claims (4)

1. The utility model provides a package association method based on full section fuzzy matching, is applied to the security inspection machine system, is provided with X ray imaging area in the security inspection machine, installs the camera above the conveyer belt before the security inspection machine entry, the camera make a video recording the direction with the conveyer belt is perpendicular and the camera make a video recording the scope of camera cover the conveyer belt entirely, its characterized in that, the method includes following steps: S1, continuously acquiring a current camera picture, extracting package image data, and caching the package image data to obtain a package image data queue M, wherein the M comprises package image data in a (t-Deltat, t) time period, deltat is package image data caching time, deltat is time when a package runs from one end of a conveyor belt to the other end, and t is a time point when the current camera picture is extracted; When executing S1, the method comprises the following steps: S11, obtaining RGB values of current camera picture composition pixels; S12, wrapping image data according to the RGB values, wherein the wrapping image data comprises a wrapping external contour, a wrapping visible light picture, leftmost and rightmost coordinates wrapped in a current camera picture and the length wrapped in the current camera picture; the method also comprises the following steps: S1a, acquiring package image data B at an initial moment; S1B, tracking packages from the camera picture and updating package image data to obtain package image data B' for fuzzy matching; Setting an initial time as T0, wherein the package image data B at the initial time comprises: b= { ID, img0, xL0, xR0, LB, T0}, wherein ID is a package ID, img0 is a package visible light picture extracted from a camera picture at time T0, xL0 is a leftmost coordinate wrapped in the camera picture at time T0, xR0 is a rightmost coordinate wrapped in the camera picture at time T0, LB is a length wrapped in the camera picture at time T0, LB = xR 0-xL 0; Tracking a package from a camera picture, and obtaining package image data B ', B' = { ID, img1, xL1, xR1, LB1 and T1} at a time T1, wherein ID is a package ID, img1 is a package visible light picture extracted from the camera picture at the time T1, xL1 is the leftmost coordinate of the package in the camera picture at the time T1, xR1 is the rightmost coordinate of the package in the camera picture at the time T1, LB1 is the length of the package in the camera picture at the time T1, and Lb1=xR1-xL 1; when LB1> LB, updating the parcel image data, wherein the parcel image data for fuzzy matching is parcel image data B', and caching the parcel image data into a parcel image data queue; S2, acquiring a package X-ray image which is completely imaged in a security inspection machine, extracting imaging time of the package X-ray image, and recording the imaging time as a time point t2; S3, carrying out time interval operation according to a time point t2 and a time point t to obtain a (t 2-delta t, t 2) time period, wherein the time point t is positioned in the (t 2-delta t, t 2) time period; s4, at a time point t2, updating a package image data queue M into a package image data queue M ', wherein M' comprises package image data of a (t 2-delta t, t 2) time period, and performing fuzzy matching with the package X-ray picture by using the package image data of the (t 2-delta t, t 2) time period; And S5, according to the fuzzy matching result, obtaining the package image data with the best adaptation degree as final package association image information, and carrying out package association binding on the final package association image information and the package X-ray picture.
2. The method of claim 1, wherein S4 is performed by intercepting the package image data of M' in the (t 2- Δt, t) time period and blurring the package X-ray picture by using the package image data in the (t 2- Δt, t) time period.
3. An electronic device for packet association based on full segment fuzzy matching, comprising: One or more processors; A storage unit configured to store one or more programs, which when executed by the one or more processors, enable the one or more processors to implement a full-segment fuzzy matching based packet association method according to any one of claims 1 to 2.
4. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, is capable of implementing a packet association method based on full-segment fuzzy matching according to any one of claims 1 to 2.

Description

Full-segment fuzzy matching-based packet association method Technical Field The invention relates to the technical field of picture data processing, in particular to a packet association method, equipment and medium based on full-segment fuzzy matching. Background In the prior art, when a subway, an airport or other security inspection scenes are used for package security inspection, after a passenger puts the package into a conveyor belt to enter the security inspection machine, contraband identification is carried out on the package, and when the contraband is identified in the package, the corresponding package needs to be determined immediately. Therefore, the package picture of the security inspection machine and the RGB picture of the package are required to be associated, so that the security inspection personnel can conveniently confirm the package picture. The existing solution thinking is that the time for finding the package on the conveyor belt is pushed back through the imaging time of the X-ray picture and the time for moving the package from the package inlet to the imaging position of the X-ray picture, so that the corresponding package visible light picture is found. The method has the following problems that 1, because the lengths of the packages are not uniform and the running speed of the conveyor belt is deviated, the time from the package inlet to the X-ray image imaging position of the packages is changed, and if the packages are fixed to a certain value, matching deviation is caused. 2. Meanwhile, the visible light picture of the package is a panoramic picture, possibly including other packages, and the package matching can be interfered. Aiming at the problems, the invention provides a packet association method based on full-segment fuzzy matching. Disclosure of Invention The invention provides a method, equipment and medium for associating package based on full-segment fuzzy matching, which utilizes package X-ray pictures to carry out fuzzy matching with package image data in a cached package image data queue M to obtain package image data with the best adaptation degree, and carries out package association according to the matched package image data and the package X-ray pictures so as to achieve the purpose of immediately determining corresponding packages when contraband is identified in the package X-ray pictures. The technical aim of the invention is realized by the following technical scheme: the method is applied to a security inspection machine system, an X-ray imaging area is arranged in the security inspection machine, a camera is arranged above a conveyor belt in front of an entrance of the security inspection machine, the shooting direction of the camera is vertical to the conveyor belt, and the shooting range of the camera completely covers the conveyor belt, and the method comprises the following steps: S1, continuously acquiring a current camera picture, extracting package image data, and caching the package image data to obtain a package image data queue M, wherein M comprises package image data in a (t-Deltat, t) time period, deltat is package image data caching time, deltat is time for a package to run from one end of a conveyor belt to the other end, and t is a time point for extracting and acquiring the current camera picture; s2, acquiring a package X-ray image which is completely imaged in a security inspection machine, extracting the imaging time of the package X-ray image, and recording the imaging time as a time point t2; S3, carrying out time interval operation according to a time point t2 and a time point t to obtain a (t 2-delta t, t 2) time period, wherein the time point t is positioned in the (t 2-delta t, t 2) time period; s4, at a time point t2, updating the package image data queue M into a package image data queue M ', wherein the M' comprises package image data of a (t 2-delta t, t 2) time period, and performing fuzzy matching with the package X-ray pictures by using the package image data of the (t 2-delta t, t 2) time period; And S5, according to the fuzzy matching result, obtaining the package image data with the best adaptation degree as final package association image information, and carrying out package association binding on the final package association image information and the package X-ray picture. Further, when executing S1, the method includes the following steps: S11, obtaining RGB values of current camera picture composition pixels; And S12, according to the RGB values, wrapping image data, wherein the wrapping image data comprises, but is not limited to, wrapping external outlines, wrapping visible light pictures, leftmost and rightmost coordinates of wrapping in the current camera picture and length of wrapping in the current camera picture. Further, the method also comprises the following steps: S1a, acquiring package image data B at an initial moment; S1B, tracking the package from the camera picture and updating the package im