CN113076880A - Hazardous article pattern recognition algorithm - Google Patents

Hazardous article pattern recognition algorithm Download PDF

Info

Publication number
CN113076880A
CN113076880A CN202110369262.5A CN202110369262A CN113076880A CN 113076880 A CN113076880 A CN 113076880A CN 202110369262 A CN202110369262 A CN 202110369262A CN 113076880 A CN113076880 A CN 113076880A
Authority
CN
China
Prior art keywords
image
detected
characteristic curve
wavelet
dangerous goods
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110369262.5A
Other languages
Chinese (zh)
Other versions
CN113076880B (en
Inventor
李元伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Industry Technical College
Original Assignee
Guangdong Industry Technical College
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangdong Industry Technical College filed Critical Guangdong Industry Technical College
Priority to CN202110369262.5A priority Critical patent/CN113076880B/en
Publication of CN113076880A publication Critical patent/CN113076880A/en
Application granted granted Critical
Publication of CN113076880B publication Critical patent/CN113076880B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Image Analysis (AREA)
  • Analysing Materials By The Use Of Radiation (AREA)

Abstract

The invention discloses a dangerous goods image recognition algorithm, belonging to the technical field of electronics and information, and the recognition algorithm comprises the following specific steps: (1) detecting X-rays; (2) performing wavelet decomposition on the image; (3) reconstructing image wavelets; (4) filling in morphology; (5) matching images; (6) detecting a characteristic curve; the method adopts the X-ray image to detect and match the shape of an object with the average effective atomic number more than 10, and matches the characteristic curve fitted by the high-low gray value of the object with the characteristic curve of the standard dangerous goods for the object with the average effective atomic number less than 10; the dangerous goods are detected by combining two modes, so that the dangerous goods carried in the luggage can be comprehensively detected, the efficiency of dangerous goods detection is favorably improved, and the labor intensity of security personnel is favorably reduced.

Description

Hazardous article pattern recognition algorithm
Technical Field
The invention relates to the technical field of electronics and information, in particular to a dangerous goods pattern recognition algorithm.
Background
Through retrieval, the Chinese patent number CN105427712B discloses a dangerous goods automatic identification device and method based on three-dimensional X-ray imaging, the detection method of the X-ray detection system is single, and the detection accuracy is low; at present, safety inspection equipment such as magnetic needles, metal weapon detection doors, X-ray detectors and the like are mostly arranged at places needing safety inspection such as airports, stations, government buildings, prisons and the like, and can find dangerous goods such as weapons, common explosives and the like, so that the safety inspection equipment plays an important role in safety inspection work; however, the devices are limited by the original technical conditions, are not satisfactory in the actual use process, and have high false alarm rate and missed detection rate; with the development of scientific technology, criminals and terrorists have gradually begun to use high and new technologies to manufacture new weapons, explosives, and the like; like high-precision bombs, plastic explosives, many drugs and the like manufactured by utilizing an integrated circuit technology, the traditional detection means cannot be used; therefore, it becomes important to invent a dangerous goods pattern recognition algorithm;
the existing dangerous goods identification method mostly adopts the traditional safety inspection equipment to cooperate with safety inspection personnel to identify dangerous goods, although dangerous goods such as weapons, common explosives and the like can be found, the labor intensity of the safety inspection personnel is easily increased, and the traditional safety inspection equipment is limited by the original technical conditions, so that high-precision bombs, plastic explosives, a plurality of drugs and the like manufactured by the integrated circuit technology cannot be quickly identified, and the identification effect is not ideal for low-density goods such as kerosene, ethanol and the like; therefore, a dangerous goods pattern recognition algorithm is proposed.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides a dangerous article image recognition algorithm.
In order to achieve the purpose, the invention adopts the following technical scheme:
a dangerous goods pattern recognition algorithm comprises the following specific steps:
(1) x-ray detection: obtaining an average effective atomic sequence value of an object image to be detected by an X-ray detector, jumping to the step (2) to perform image wavelet decomposition on the object image to be detected if the average effective atomic sequence value is greater than or equal to 10, and jumping to the step (6) to perform characteristic curve detection on the object image to be detected if the average effective atomic sequence value is less than or equal to 10;
(2) image wavelet decomposition: performing one-layer wavelet decomposition on an object image to be detected by adopting a wavelet decomposition method, wherein each layer of image is decomposed into a low-frequency wavelet coefficient matrix and a high-frequency wavelet coefficient matrix, the low-frequency coefficient represents an approximate part of the image, and the high-frequency wavelet coefficient matrix comprises coefficients in the horizontal direction, the vertical direction and the diagonal direction;
(3) and (3) image wavelet reconstruction: carrying out edge detection on an image approximate part, namely a low-frequency part by using a Canny algorithm to obtain a sub-image of the low-frequency part; simultaneously, utilizing Roberts and Sobel algorithms to carry out edge detection on coefficients in the horizontal direction, the vertical direction and the diagonal direction of the image, namely the high-frequency part, so as to obtain a sub-image of the high-frequency part; performing wavelet reconstruction on the sub-image of the low-frequency part and the sub-image of the high-frequency part to obtain an edge image;
(4) morphological filling: performing morphological filling on the edge image by using a hole filling mode to obtain an enhanced edge image to be detected;
(5) image matching: constructing a dangerous article standard image, and matching the enhanced edge image to be detected with the dangerous article standard image by using a curvature contour angular point method to obtain a dangerous article judgment result of a curvature angular point;
(6) and (3) characteristic curve detection: and constructing a characteristic curve of the common dangerous goods, and simultaneously carrying out dangerous goods detection matching on the image of the object to be detected by adopting an inorganic matter characteristic curve method to obtain a curved dangerous goods judgment result.
Preferably, the wavelet decomposition of the image in the step (2) needs a scale function in a two-dimensional case
Figure BDA0003008657670000031
And two-dimensional wavelet psiH(x,y)、ψV(x, y) and ψD(x, y), each of which can be expressed as a one-dimensional scale function
Figure BDA0003008657670000032
And ψ, as follows:
Figure BDA0003008657670000033
its two-dimensional separable direction-sensitive wavelet can be represented as:
Figure BDA0003008657670000034
Figure BDA0003008657670000035
ψD(x,y)＝ψ(x)ψ(y) (4)
in the formula: psiH、ψVAnd psiDRespectively representing the image intensity variation in different directions.
Preferably, given the separable two-dimensional scale and wavelet functions, the scale and wavelet basis functions are defined as follows:
Figure BDA0003008657670000036
Figure BDA0003008657670000037
in the formula: i ═ { H, V, D } represents wavelets in different directions;
then an f (x, y) discrete wavelet transform of size M × N:
Figure BDA0003008657670000041
Figure BDA0003008657670000042
in the formula: j is a function of0In order to be the initial scale, the method comprises the following steps,
Figure BDA0003008657670000043
define a dimension of j0F (x, y) approximation image; wψ(j, m, n) contains detail images in horizontal, vertical, and diagonal directions; let j00, and N is selected to be M2jJ-0, 1,2, …, j-1 and m, n-0, 1,2, … 2j-1。
Preferably, the wavelet reconstruction formula in step (3) is as follows:
Figure BDA0003008657670000044
preferably, the hole filling formula in step (4) is as follows:
Figure BDA0003008657670000045
in the formula: k 1,2,3.
Preferably, the curvature in step (5) is mathematically defined as follows: taking an arc section from a point M on the smooth arc, wherein the length of the arc section is delta s, and the corresponding tangent rotation angle is delta alpha, and defining the arc section;
wherein the average curvature over Δ S is:
Figure BDA0003008657670000046
the curvature at point M is:
Figure BDA0003008657670000047
the formula for calculating the discrete curvature by using a curve parameter equation is as follows:
Figure BDA0003008657670000048
preferably, the curvature in step (5) includes global and local curvatures of the image, and the specific calculation process is as follows:
s1: calculating the average curvature of each to-be-detected enhanced edge image profile;
s2: then, judging the current contour point by taking the average curvature of the contour as a threshold, and if the current curvature is smaller than the threshold, determining the current contour point as a real contour angular point;
s3: after the angular points of all to-be-detected enhanced edge images are extracted, performing angular point matching on the to-be-detected enhanced edge images and the standard images of the dangerous goods;
when the corner points are matchedBuilding a corner feature descriptor, wherein the corner feature descriptor is defined as: suppose P1、P2And P3Three adjacent corner points on the contour curve S with curvatures of K1、K2And K3，P1To P2Is a distance L1，P2To P3Is L2) Then P is2The descriptor B is:
B＝(L1+L2)(K3-K1) (14)。
preferably, the image matching in step (5) specifically comprises the following steps:
SS 1: extracting real corner points of a complete edge image with distinct outline features and a standard image of the dangerous goods;
SS 2: respectively solving descriptors B and B' of each enhanced edge image corner point to be detected and each dangerous article standard image corner point by using a formula (14);
SS 3: performing sequential difference operation on each corner point descriptor in the enhanced edge image to be detected and all corner point descriptors in the standard image of the hazardous article, wherein if the ratio of the first term to the second term of the vector of the difference value is smaller than a threshold value, and the threshold value is 0.5, the diagonal points of the two images are matched, otherwise, the diagonal points are not matched;
SS 4: and counting the matching similarity percentage of the corner points of the same target contour, and if the similarity is greater than a preset value, determining that the article is a dangerous article.
Preferably, the specific process of detecting the characteristic curve in the step (6) is as follows:
SSS 1: acquiring high and low energy gray values of an object image to be detected through an X-ray security inspection machine, and drawing a characteristic curve of the object image to be detected according to the high and low energy gray values;
SSS 2: constructing a characteristic curve of a common dangerous article, and taking the characteristic curve as a standard for detecting the dangerous article;
SSS 3: comparing the characteristic curve of the image of the object to be detected with the characteristic curve of the common dangerous goods; and if the characteristic curve of the image of the object to be detected is close to the characteristic curve of the dangerous goods, judging that the object is the dangerous goods, otherwise, judging that the object is normal.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the method, the average effective atomic number value of an object is obtained by an X-ray detector, the detected object is divided into two types, the shape of the object with the average effective atomic number larger than 10 is detected and matched by adopting an X-ray image, and for the target with the average effective atomic number smaller than 10, a characteristic curve fitted by using the high-low gray value of the target is matched with a characteristic curve of a standard dangerous article; the dangerous goods are detected by combining two modes, so that the dangerous goods carried in the luggage can be comprehensively detected, the efficiency of dangerous goods detection is favorably improved, and the labor intensity of security personnel is favorably reduced.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
Fig. 1 is an overall flowchart of a hazardous material pattern recognition algorithm according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In the description of the present invention, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention.
Example 1
A dangerous goods pattern recognition algorithm comprises the following specific steps:
(1) x-ray detection: obtaining an average effective atomic sequence value of an object image to be detected by an X-ray detector, jumping to the step (2) to perform image wavelet decomposition on the object image to be detected if the average effective atomic sequence value is greater than or equal to 10, and jumping to the step (6) to perform characteristic curve detection on the object image to be detected if the average effective atomic sequence value is less than or equal to 10;
(2) image wavelet decomposition: performing one-layer wavelet decomposition on an object image to be detected by adopting a wavelet decomposition method, wherein each layer of image is decomposed into a low-frequency wavelet coefficient matrix and a high-frequency wavelet coefficient matrix, the low-frequency coefficient represents an approximate part of the image, and the high-frequency wavelet coefficient matrix comprises coefficients in the horizontal direction, the vertical direction and the diagonal direction;
(3) and (3) image wavelet reconstruction: carrying out edge detection on an image approximate part, namely a low-frequency part by using a Canny algorithm to obtain a sub-image of the low-frequency part; simultaneously, utilizing Roberts and Sobel algorithms to carry out edge detection on coefficients in the horizontal direction, the vertical direction and the diagonal direction of the image, namely the high-frequency part, so as to obtain a sub-image of the high-frequency part; performing wavelet reconstruction on the sub-image of the low-frequency part and the sub-image of the high-frequency part to obtain an edge image;
(4) morphological filling: performing morphological filling on the edge image by using a hole filling mode to obtain an enhanced edge image to be detected;
(5) image matching: constructing a dangerous article standard image, and matching the enhanced edge image to be detected with the dangerous article standard image by using a curvature contour angular point method to obtain a dangerous article judgment result of a curvature angular point;
(6) and (3) characteristic curve detection: and constructing a characteristic curve of the common dangerous goods, and simultaneously carrying out dangerous goods detection matching on the image of the object to be detected by adopting an inorganic matter characteristic curve method to obtain a curved dangerous goods judgment result.
In the step (2), the image wavelet decomposition needs a scale function under the condition of two dimensions
Figure BDA0003008657670000081
And two-dimensional wavelet psiH(x,y)、ψV(x, y) and ψD(x, y), each of which can be expressed as a one-dimensional scale function
Figure BDA0003008657670000082
And ψ, as follows:
Figure BDA0003008657670000083
its two-dimensional separable direction-sensitive wavelet can be represented as:
Figure BDA0003008657670000084
Figure BDA0003008657670000085
ψD(x,y)＝ψ(x)ψ(y) (4)
in the formula: psiH、ψVAnd psiDRespectively representing the image intensity variation in different directions.
After the two-dimensional scale and wavelet function are given separately, the scale and wavelet basis functions are defined as follows:
Figure BDA0003008657670000086
Figure BDA0003008657670000087
in the formula: i ═ { H, V, D } represents wavelets in different directions;
then an f (x, y) discrete wavelet transform of size M × N:
Figure BDA0003008657670000091
Figure BDA0003008657670000092
in the formula: j is a function of0In order to be the initial scale, the method comprises the following steps,
Figure BDA0003008657670000093
define a dimension of j0F (x, y) approximation image; wψ(j, m, n) contains detail images in horizontal, vertical, and diagonal directions; let j00, and N is selected to be M2jJ-0, 1,2, …, j-1 and m, n-0, 1,2, … 2j-1。
The wavelet reconstruction formula in the step (3) is as follows:
Figure BDA0003008657670000094
the hole filling formula in the step (4) is as follows:
Figure BDA0003008657670000095
in the formula: k 1,2,3.
The mathematical definition of the curvature of step (5) is as follows: taking an arc section from a point M on the smooth arc, wherein the length of the arc section is delta s, and the corresponding tangent rotation angle is delta alpha, and defining the arc section;
wherein the average curvature over Δ S is:
Figure BDA0003008657670000096
the curvature at point M is:
Figure BDA0003008657670000097
the formula for calculating the discrete curvature by using a curve parameter equation is as follows:
Figure BDA0003008657670000098
the curvature in the step (5) comprises global curvature and local curvature of the image, and the specific calculation process is as follows:
s1: calculating the average curvature of each to-be-detected enhanced edge image profile;
s2: then, judging the current contour point by taking the average curvature of the contour as a threshold, and if the current curvature is smaller than the threshold, determining the current contour point as a real contour angular point;
s3: after the angular points of all to-be-detected enhanced edge images are extracted, performing angular point matching on the to-be-detected enhanced edge images and the standard images of the dangerous goods;
when the corner points are matched, a corner point feature descriptor is constructed, wherein the corner point feature descriptor is defined as: suppose P1、P2And P3Three adjacent corner points on the contour curve S with curvatures of K1、K2And K3，P1To P2Is a distance L1，P2To P3Is L2) Then P is2The descriptor B is:
B＝(L1+L2)(K3-K1) (14)。
the image matching in the step (5) comprises the following specific steps:
SS 1: extracting real corner points of a complete edge image with distinct outline features and a standard image of the dangerous goods;
SS 2: respectively solving descriptors B and B' of each enhanced edge image corner point to be detected and each dangerous article standard image corner point by using a formula (14);
SS 3: performing sequential difference operation on each corner point descriptor in the enhanced edge image to be detected and all corner point descriptors in the standard image of the hazardous article, wherein if the ratio of the first term to the second term of the vector of the difference value is smaller than a threshold value, and the threshold value is 0.5, the diagonal points of the two images are matched, otherwise, the diagonal points are not matched;
SS 4: and counting the matching similarity percentage of the corner points of the same target contour, and if the similarity is greater than a preset value, determining that the article is a dangerous article.
The specific process of detecting the characteristic curve in the step (6) is as follows:
SSS 1: acquiring high and low energy gray values of an object image to be detected through an X-ray security inspection machine, and drawing a characteristic curve of the object image to be detected according to the high and low energy gray values;
SSS 2: constructing a characteristic curve of a common dangerous article, and taking the characteristic curve as a standard for detecting the dangerous article;
SSS 3: comparing the characteristic curve of the image of the object to be detected with the characteristic curve of the common dangerous goods; and if the characteristic curve of the image of the object to be detected is close to the characteristic curve of the dangerous goods, judging that the object is the dangerous goods, otherwise, judging that the object is normal.
In order to verify the image matching effect, an X-ray dangerous goods image with the size of 256 × 256 is selected as a sample (the sample comprises a pistol, handcuffs and a grenade), simulation is carried out by adopting matlab7.1, and the specific test data is as follows:
Figure BDA0003008657670000111
example 2
A dangerous goods pattern recognition algorithm comprises the following specific steps:
(1) x-ray detection: obtaining an average effective atomic sequence value of an object image to be detected by an X-ray detector, jumping to the step (2) to perform image wavelet decomposition on the object image to be detected if the average effective atomic sequence value is greater than or equal to 10, and jumping to the step (6) to perform characteristic curve detection on the object image to be detected if the average effective atomic sequence value is less than or equal to 10;
(2) image wavelet decomposition: performing one-layer wavelet decomposition on an object image to be detected by adopting a wavelet decomposition method, wherein each layer of image is decomposed into a low-frequency wavelet coefficient matrix and a high-frequency wavelet coefficient matrix, the low-frequency coefficient represents an approximate part of the image, and the high-frequency wavelet coefficient matrix comprises coefficients in the horizontal direction, the vertical direction and the diagonal direction;
(3) and (3) image wavelet reconstruction: carrying out edge detection on an image approximate part, namely a low-frequency part by using a Canny algorithm to obtain a sub-image of the low-frequency part; simultaneously, utilizing Roberts and Sobel algorithms to carry out edge detection on coefficients in the horizontal direction, the vertical direction and the diagonal direction of the image, namely the high-frequency part, so as to obtain a sub-image of the high-frequency part; performing wavelet reconstruction on the sub-image of the low-frequency part and the sub-image of the high-frequency part to obtain an edge image;
(4) morphological filling: performing morphological filling on the edge image by using a hole filling mode to obtain an enhanced edge image to be detected;
(5) image matching: constructing a dangerous article standard image, and matching the enhanced edge image to be detected with the dangerous article standard image by using a curvature contour angular point method to obtain a dangerous article judgment result of a curvature angular point;
(6) and (3) characteristic curve detection: constructing a common dangerous article characteristic curve, and simultaneously carrying out dangerous article detection matching on an object image to be detected by adopting an inorganic matter characteristic curve method to obtain a curved dangerous article judgment result;
the rest is the same as example 1.
In order to utilize the verification characteristic curve to detect the effect of inorganic dangerous goods, firecrackers with different thicknesses are selected in the test, the firecrackers are detected through the established firecracker characteristic curve, and the test detection data are as follows:
Figure BDA0003008657670000121
Figure BDA0003008657670000131
example 3
A dangerous goods pattern recognition algorithm comprises the following specific steps:
(1) x-ray detection: obtaining an average effective atomic sequence value of an object image to be detected by an X-ray detector, jumping to the step (2) to perform image wavelet decomposition on the object image to be detected if the average effective atomic sequence value is greater than or equal to 10, and jumping to the step (6) to perform characteristic curve detection on the object image to be detected if the average effective atomic sequence value is less than or equal to 10;
(2) image wavelet decomposition: performing one-layer wavelet decomposition on an object image to be detected by adopting a wavelet decomposition method, wherein each layer of image is decomposed into a low-frequency wavelet coefficient matrix and a high-frequency wavelet coefficient matrix, the low-frequency coefficient represents an approximate part of the image, and the high-frequency wavelet coefficient matrix comprises coefficients in the horizontal direction, the vertical direction and the diagonal direction;
(3) and (3) image wavelet reconstruction: carrying out edge detection on an image approximate part, namely a low-frequency part by using a Canny algorithm to obtain a sub-image of the low-frequency part; simultaneously, utilizing Roberts and Sobel algorithms to carry out edge detection on coefficients in the horizontal direction, the vertical direction and the diagonal direction of the image, namely the high-frequency part, so as to obtain a sub-image of the high-frequency part; performing wavelet reconstruction on the sub-image of the low-frequency part and the sub-image of the high-frequency part to obtain an edge image;
(4) morphological filling: performing morphological filling on the edge image by using a hole filling mode to obtain an enhanced edge image to be detected;
(5) image matching: constructing a dangerous article standard image, and matching the enhanced edge image to be detected with the dangerous article standard image by using a curvature contour angular point method to obtain a dangerous article judgment result of a curvature angular point;
(6) and (3) characteristic curve detection: constructing a common dangerous article characteristic curve, and simultaneously carrying out dangerous article detection matching on an object image to be detected by adopting an inorganic matter characteristic curve method to obtain a curved dangerous article judgment result;
the rest is the same as example 1.
In order to test the effectiveness of the invention, several articles including a cutter, a toy gun, bottled alcohol, gasoline and the like are selected as test samples, the test samples are respectively and independently mixed with other articles, a dangerous article shape image database (28 types) and common dangerous article characteristic curves (15 types) are established as dangerous article judgment standards, a plurality of tests are carried out in different directions and positions, and the detection accuracy rates of different articles are specifically as follows:
article with a cover Number of samples Rate of accuracy False detection rate Rate of missed examination
Cutting tool 80 91.25％ 7.50％ 8.13％
Toy gun 80 87.50％ 5.20％ 7.50％
Gasoline (gasoline) 80 92.00％ 6.25％ 5.00％
Bottled alcohol 80 90.00％ 5.00％ 7.13％
As can be seen from the data tables of the embodiment 1, the embodiment 2 and the embodiment 3, the dangerous goods object detection accuracy of the invention is high, and the false detection rate and the omission factor are both below 10%, so that the invention can comprehensively detect the dangerous goods carried in the luggage, improve the efficiency of dangerous goods detection, and is beneficial to reducing the labor intensity of security personnel.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.

Claims (9)

1. A dangerous goods pattern recognition algorithm is characterized by comprising the following specific steps:
(1) x-ray detection: obtaining an average effective atomic sequence value of an object image to be detected by an X-ray detector, jumping to the step (2) to perform image wavelet decomposition on the object image to be detected if the average effective atomic sequence value is greater than or equal to 10, and jumping to the step (6) to perform characteristic curve detection on the object image to be detected if the average effective atomic sequence value is less than or equal to 10;
(2) image wavelet decomposition: performing one-layer wavelet decomposition on an object image to be detected by adopting a wavelet decomposition method, wherein each layer of image is decomposed into a low-frequency wavelet coefficient matrix and a high-frequency wavelet coefficient matrix, the low-frequency coefficient represents an approximate part of the image, and the high-frequency wavelet coefficient matrix comprises coefficients in the horizontal direction, the vertical direction and the diagonal direction;
(3) and (3) image wavelet reconstruction: carrying out edge detection on an image approximate part, namely a low-frequency part by using a Canny algorithm to obtain a sub-image of the low-frequency part; simultaneously, utilizing Roberts and Sobel algorithms to carry out edge detection on coefficients in the horizontal direction, the vertical direction and the diagonal direction of the image, namely the high-frequency part, so as to obtain a sub-image of the high-frequency part; performing wavelet reconstruction on the sub-image of the low-frequency part and the sub-image of the high-frequency part to obtain an edge image;
(4) morphological filling: performing morphological filling on the edge image by using a hole filling mode to obtain an enhanced edge image to be detected;
(5) image matching: constructing a dangerous article standard image, and matching the enhanced edge image to be detected with the dangerous article standard image by using a curvature contour angular point method to obtain a dangerous article judgment result of a curvature angular point;
(6) and (3) characteristic curve detection: and constructing a characteristic curve of the common dangerous goods, and simultaneously carrying out dangerous goods detection matching on the image of the object to be detected by adopting an inorganic matter characteristic curve method to obtain a curved dangerous goods judgment result.
2. The hazardous material pattern recognition algorithm of claim 1, wherein the wavelet decomposition of the image in step (2) requires a scale function in two dimensions
Figure FDA0003008657660000021
(x, y) and two-dimensional wavelet psiH(x，y)、ψV(x, y) and ψD(x, y), each of which can be expressed as a one-dimensional scale function
Figure FDA0003008657660000022
And ψ, as follows:
Figure FDA0003008657660000023
its two-dimensional separable direction-sensitive wavelet can be represented as:
Figure FDA0003008657660000024
Figure FDA0003008657660000025
ψD(x，y)＝ψ(x)ψ(y) (4)
in the formula: psiH、ψVAnd psiDRespectively representing the image intensity variation in different directions.
3. The hazardous material pattern recognition algorithm of claim 2, wherein the separable two-dimensional scale and wavelet function are given to define a scale and wavelet basis function as follows:
Figure FDA0003008657660000026
Figure FDA0003008657660000027
in the formula: i ═ { H, V, D } represents wavelets in different directions;
then an f (x, y) discrete wavelet transform of size M × N:
Figure FDA0003008657660000028
Figure FDA0003008657660000029
in the formula: j is a function of0In order to be the initial scale, the method comprises the following steps,
Figure FDA00030086576600000210
define a dimension of j0F (x, y) approximation image; wψ(j, m, n) contains detail images in horizontal, vertical, and diagonal directions; let j00, and N is selected to be M2jJ-0, 1,2, …, j-1 and m, n-0, 1,2, … 2j-1。
4. The algorithm for identifying patterns of dangerous goods as claimed in claim 1, wherein the wavelet reconstruction formula of step (3) is as follows:
Figure FDA0003008657660000031
5. the hazardous article pattern recognition algorithm of claim 1, wherein the hole filling formula in step (4) is as follows:
Figure FDA0003008657660000032
in the formula: k 1,2,3.
6. The hazardous article pattern recognition algorithm of claim 1, wherein the curvature of step (5) is mathematically defined as follows: taking an arc section from a point M on the smooth arc, wherein the length of the arc section is delta s, and the corresponding tangent rotation angle is delta alpha, and defining the arc section;
wherein the average curvature over Δ S is:
Figure FDA0003008657660000033
the curvature at point M is:
Figure FDA0003008657660000034
the formula for calculating the discrete curvature by using a curve parameter equation is as follows:
Figure FDA0003008657660000035
7. the hazardous article pattern recognition algorithm according to claim 1, wherein the curvature in step (5) comprises global and local curvatures of the image, and the calculation process is as follows:
s1: calculating the average curvature of each to-be-detected enhanced edge image profile;
s2: then, judging the current contour point by taking the average curvature of the contour as a threshold, and if the current curvature is smaller than the threshold, determining the current contour point as a real contour angular point;
s3: after the angular points of all to-be-detected enhanced edge images are extracted, performing angular point matching on the to-be-detected enhanced edge images and the standard images of the dangerous goods;
and when the corner points are matched, constructing a corner point feature descriptor, wherein the corner point feature descriptor is defined as: suppose P1、P2And P3Three adjacent corner points on the contour curve S with curvatures of K1、K2And K3，P1To P2Is a distance L1，P2To P3Is L2) Then P is2The descriptor B is:
B＝(L1+L2)(K3-K1) (14)。
8. the hazardous article pattern recognition algorithm according to claim 1, wherein the image matching in step (5) specifically comprises the following steps:
SS 1: extracting real corner points of a complete edge image with distinct outline features and a standard image of the dangerous goods;
SS 2: respectively solving descriptors B and B' of each enhanced edge image corner point to be detected and each dangerous article standard image corner point by using a formula (14);
SS 3: performing sequential difference operation on each corner point descriptor in the enhanced edge image to be detected and all corner point descriptors in the standard image of the hazardous article, wherein if the ratio of the first term to the second term of the vector of the difference value is smaller than a threshold value, and the threshold value is 0.5, the diagonal points of the two images are matched, otherwise, the diagonal points are not matched;
SS 4: and counting the matching similarity percentage of the corner points of the same target contour, and if the similarity is greater than a preset value, determining that the article is a dangerous article.
9. The algorithm for identifying dangerous goods patterns according to claim 1, wherein the detection of the characteristic curve in step (6) is implemented by the following specific process:
SSS 1: acquiring high and low energy gray values of an object image to be detected through an X-ray security inspection machine, and drawing a characteristic curve of the object image to be detected according to the high and low energy gray values;
SSS 2: constructing a characteristic curve of a common dangerous article, and taking the characteristic curve as a standard for detecting the dangerous article;
SSS 3: comparing the characteristic curve of the image of the object to be detected with the characteristic curve of the common dangerous goods; and if the characteristic curve of the image of the object to be detected is close to the characteristic curve of the dangerous goods, judging that the object is the dangerous goods, otherwise, judging that the object is normal.
CN202110369262.5A 2021-04-06 2021-04-06 Dangerous article pattern recognition algorithm Active CN113076880B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110369262.5A CN113076880B (en) 2021-04-06 2021-04-06 Dangerous article pattern recognition algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110369262.5A CN113076880B (en) 2021-04-06 2021-04-06 Dangerous article pattern recognition algorithm

Publications (2)

Publication Number Publication Date
CN113076880A true CN113076880A (en) 2021-07-06
CN113076880B CN113076880B (en) 2023-06-06

Family

ID=76615151

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110369262.5A Active CN113076880B (en) 2021-04-06 2021-04-06 Dangerous article pattern recognition algorithm

Country Status (1)

Country Link
CN (1) CN113076880B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114332512A (en) * 2022-03-14 2022-04-12 南京正驰科技发展有限公司 Security check card port detection method and system based on big data

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103198483A (en) * 2013-04-07 2013-07-10 西安电子科技大学 Multiple time phase remote sensing image registration method based on edge and spectral reflectivity curve
CN110516531A (en) * 2019-07-11 2019-11-29 广东工业大学 A kind of recognition methods of the dangerous mark based on template matching
CN111091111A (en) * 2019-12-25 2020-05-01 河北工业大学 Vehicle bottom dangerous target identification method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103198483A (en) * 2013-04-07 2013-07-10 西安电子科技大学 Multiple time phase remote sensing image registration method based on edge and spectral reflectivity curve
CN110516531A (en) * 2019-07-11 2019-11-29 广东工业大学 A kind of recognition methods of the dangerous mark based on template matching
CN111091111A (en) * 2019-12-25 2020-05-01 河北工业大学 Vehicle bottom dangerous target identification method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114332512A (en) * 2022-03-14 2022-04-12 南京正驰科技发展有限公司 Security check card port detection method and system based on big data

Also Published As

Publication number Publication date
CN113076880B (en) 2023-06-06

Similar Documents

Publication Publication Date Title
CN108510467B (en) SAR image target identification method based on depth deformable convolution neural network
Riffo et al. Automated detection of threat objects using adapted implicit shape model
Mery et al. Automatic defect recognition in x-ray testing using computer vision
CN105095856B (en) Face identification method is blocked based on mask
US8063815B2 (en) Method and system for automatic classification of objects
US8831355B2 (en) Scale robust feature-based identifiers for image identification
CN103761748B (en) Anomaly detection method and device
Mahalanobis et al. Evaluation of MACH and DCCF correlation filters for SAR ATR using the MSTAR public database
Megherbi et al. A classifier based approach for the detection of potential threats in CT based baggage screening
CN112712093B (en) Security check image identification method and device, electronic equipment and storage medium
Wu et al. Dual tree complex wavelet transform approach to copy-rotate-move forgery detection
CN104680158A (en) Face recognition method based on multi-scale block partial multi-valued mode
Yuan et al. A deep learning method for detection of dangerous equipment
CN113076880A (en) Hazardous article pattern recognition algorithm
Xu et al. A defect inspection for explosive cartridge using an improved visual attention and image-weighted eigenvalue
Zhang et al. Automatic shoeprint retrieval system for use in forensic investigations
Riffo et al. Handgun detection in single-spectrum multiple X-ray views based on 3d object recognition
EP2289024A1 (en) Method for automatic region segmentation on cartridge case base and selection of the best mark region for cartridge case comparison
Mosorov et al. Image texture defect detection method using fuzzy C-means clustering for visual inspection systems
Rajkumar et al. Digital image forgery detection using SIFT feature
Zhou Using vector quantization of hough transform for circle detection
Isaac-Medina et al. Seeing through the data: A statistical evaluation of prohibited item detection benchmark datasets for X-ray security screening
Yammen et al. Cartridge case image matching using effective correlation area based method
Wenchao et al. A fingerprint identification algorithm based on wavelet transformation characteristic coefficient
Patel et al. Surface defect detection using SVM‐based machine vision system with optimized feature

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant