CN104166963B - The detection method of luggage castor in a kind of X-ray safety check equipment - Google Patents
The detection method of luggage castor in a kind of X-ray safety check equipment Download PDFInfo
- Publication number
- CN104166963B CN104166963B CN201410378395.9A CN201410378395A CN104166963B CN 104166963 B CN104166963 B CN 104166963B CN 201410378395 A CN201410378395 A CN 201410378395A CN 104166963 B CN104166963 B CN 104166963B
- Authority
- CN
- China
- Prior art keywords
- caster axle
- castor
- luggage
- point
- interest
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Analysing Materials By The Use Of Radiation (AREA)
Abstract
The present invention relates to X-ray transmission imaging safety check technical field, and in particular to be luggage castor in a kind of X-ray safety check equipment detection method.The invention provides luggage castor knows method for distinguishing in a kind of X-ray safety check equipment image, use for reference the technology of image object identification, based on aspect features such as the distinctive material of castor, shape, positions, the present invention proposes a kind of recognition methodss for first looking for caster axle to judge castor again, the method can find the position of castor in luggage image and scope, so as to the content analysis to luggage provide extracted information, and the spurious alarm that castor is mistaken for explosive is reduced, improve the efficiency of safety inspection.
Description
Technical field
The present invention relates to X-ray transmission imaging safety inspection technical field, and in particular to be a kind of X-ray safety inspection
The detection method of luggage castor in equipment.
Background technology
At home and abroad in the disclosed document with regard to safety inspection technology delivered, have no and be directly related to the identification of luggage castor
Document, Canadian Optosecurity companies propose a series of dangerous goods or chaff interference in identification luggage, such as gun,
The patent of antitank grenade and notebook, such as its United States Patent (USP) 20120093367,20130003135, and its Canadian Patent
CA02608124, but without the technology of identification referred directly to for castor, the angle of concern is not invested in luggage
The feature of itself.Kong Weiwu of domestic the First Research Institute of Ministry of Public Security et al. is on December 10th, 2010 in the 15th national image
Graphics academic meeting paper concentrates the paper delivered《Dual-energy x-ray luggage case image characteristic region elimination method》In, propose
Strategy that luggage pull bar is rejected.Huang Jiayi of China Peoples Public Security University et al. exists《Security protection science and technology》2011 the 03rd
The paper delivered on phase《Sigmatron wraps up image edge features recognition methodss》, the edge feature for mainly using luggage searches
The feature of the aspects such as the simple metal parts such as rope handle, lock, document above simultaneously underuse the material properties and shape of castor
Shape feature is directly identified to castor.Current X-ray rays safety detection apparatus are being detected in luggage case during explosive, some rows
Lee's case feature of itself, such as castor, for the precision of detection has very important impact, castor is used as the important of luggage case
Feature, finds to cause a large amount of false alarms, hence it is evident that have impact on equipment performance index in practice.Main cause is castor itself
The difference of gray scale, material behavior and explosive is less.Additionally, identification castor additionally aids the feature of analysis luggage case, for example, speculate
The species of luggage case, position of pull bar etc..
The content of the invention
In order to overcome defect of the prior art, the invention provides a kind of identification side for first looking for caster axle to judge castor again
Method.The technology of image object identification is used for reference, luggage has been recognized based on aspect features such as the distinctive material of castor, shape, positions
Castor.So as to offer convenience to the content analysis of luggage, and can effectively reduce the quantity of explosive detection false alarm.
In a kind of X-ray safety check equipment that the present invention is provided, the detection method of luggage castor, specifically includes following step
Suddenly:
Step one:Set up luggage caster axle point of interest detection model;
Step 2:When luggage is entered in the working region of X-ray safety check equipment, the position according to caster for trunk is special
Property, the body region of Luggage trunk body is determined as overall search scope, the region easily occurred in castor scans for, and obtains institute
State the seed region of caster axle;
Step 3:With seed region to guide, take the method that point of interest is detected to determine the region at caster axle place;
The caster axle point of interest detection model set up using step one, searches caster axle color close near seed region
The point of interest region of collection distribution, locks position and the size of caster axle point of interest;
Step 4:The shape obtained according to point of interest or positional information, judge with the presence or absence of caster axle around point of interest,
Method is as follows:
In the region that each caster axle suspicion position for obtaining is made up of single or 2 point of interest windows, it is right to find
Claim axle, and by the interestingness score upright projection of each pixel in the region on axis of symmetry, it is desirable to the curve that projection accumulation is formed
Both sides are high, the middle low grown form for meeting caster axle, while it is preferably symmetrical to require that point of interest region has along axis of symmetry
Property;
Step 5:Castor is determined whether there is in the surrounding of caster axle, method is as follows:
Step A:It is for the castor shaft position for existing, oval in its surrounding searching 1, endpoint detections are first carried out,
Determine the parameter i.e. X-axis radius and Y-axis radius of elliptic equation again by Hough transformation, and ensure that there is between parameter rational size
Relation, it is determined that there is elliptoid castor;
Step B:Luggage case main shaft is commonly present certain angle with image main shaft, and Hough transformation is needed repeatedly, by oval X
Axle/Y-axis rotates multiple angles in the picture and repeats to process, and finds maximum Hough transformation value, and the maximum Hough transformation value is big
In certain threshold value, otherwise must not believe that periphery is present oval;
Step C:Eliminate the obvious irrational castor position of isolated presence or position.
Further, in one the step of methods described, prepare the caster axle example of a number of manual mark in advance
Image, belongs to the probability density of caster axle color with the window function estimation technique come estimated color point.
Further, the step 2 is comprised the following steps:
Step A:With gray level threshold segmentation method Luggage trunk body region segmentation out, formed 0-1 binary map;
Step B:The 0-1 binary map obtained in the step A, by way of first micro 12 does opening operation again again, disappear
Except details, the subsidiary stand-alone component of Luggage trunk body surrounding is removed, then micro binary map is enlarged into into original size again, obtained
The body region of Luggage trunk body;
Step C:With the front end of luggage case, rear end, left and right two ends as order, searched near corresponding luggage case border
Rope;
D steps:In radioscopic image, the specific gray scale having using the caster axle of luggage case and RGB pseudo-colourss models
Enclose, the seed region of caster axle is split with the mode of Threshold segmentation.
Further, three the step of methods described in point of interest detection method, including:
Step A:It is in seed region, with interest point detecting method proposed by the present invention, close based on caster axle color probability
Degree, finds the point of interest window including caster axle, and the window describes the position of caster axle and size;
Step B:Due to the difference of caster axle imaging angle, caster axle is presented dumbbell shape, calabash shape or sub-circular, wherein
Dumbbell shape, can form 1 point of interest, in each split according to syntagmatic, apart from two similar interest of close, size
Point combines, and forms 1 caster axle;
Step C:Eliminate not prominent enough the point of interest of those point of interest correlation properties.
Further, the feature of the caster axle is gray scale, material behavior and the shape of caster axle.
Further, the caster axle characteristic adopts light alloy material for caster axle.
Compared with prior art, superior effect is:The invention provides row in a kind of X-ray safety check equipment image
Lee's castor knows method for distinguishing, and the method can find the position of castor in luggage image and scope, so as to the content analysis to luggage
Extracted information is provided, and reduces the spurious alarm that castor is mistaken for explosive, improve the efficiency of safety inspection.
Description of the drawings
Fig. 1 is the method flow diagram that the present invention carries out castor detection in radioscopic image;
Fig. 2 is the Luggage trunk body figure that castor wheel shaft is embedded in luggage case base plate;
Outside Fig. 3 is suspended from luggage case for castor and the Luggage trunk body figure that can rotate freely;
Fig. 4 is caster axle shape graph;
Fig. 5 is the testing process exemplary plot of the present invention.
Reference is as follows:
A- dumbbell shape caster axle figures, b- calabash shape caster axle figures, c- sub-circular caster axle figures, d- luggage case artworks, e-
Luggage case segmentation result, the hunting zone of f- caster axles, the point of interest window that g- black box correspondence is detected, h- black box correspondence
The castor for detecting.
Specific embodiment
Below in conjunction with the accompanying drawings the specific embodiment of the invention is described in further detail.
Embodiment 1
With reference to Figure of description 1-5, the present invention is illustrated, the invention provides in a kind of X-ray safety check equipment
The detection method of luggage castor, comprises the following steps:
Step one:Set up luggage caster axle " point of interest " detection model;
Step 2:When luggage is entered in the working region of X-ray safety check equipment, the position according to caster for trunk is special
Property, " body region " of Luggage trunk body is determined as overall search scope, the region easily occurred in castor scans for, and obtains
" seed region " of the caster axle, there is provided preliminary hunting zone;
Step 3:With " seed region " to guide, take the method that " point of interest " is detected to determine the area at caster axle place
Domain;
Caster axle " point of interest " detection model set up using step one, searches caster axle face near " seed region "
" the point of interest region " of color dense distribution, locks position and the size of caster axle point of interest;
Step 4:The shape obtained according to " point of interest " or positional information, judge to whether there is foot around " point of interest "
Wheel shaft, method are as follows:
In the region that each caster axle suspicion position for obtaining is made up of single or 2 point of interest windows, it is right to find
Claim axle, and by the interestingness score upright projection of each pixel in the region on axis of symmetry, it is desirable to the curve that projection accumulation is formed
Both sides are high, the middle low grown form for meeting caster axle, while it is preferably symmetrical to require that point of interest region has along axis of symmetry
Property;
Step 5:Castor is determined whether there is in the surrounding of caster axle, method is as follows:
Step A:It is for the castor shaft position for existing, oval in its surrounding searching 1, endpoint detections are first carried out,
Determine the parameter i.e. X-axis radius and Y-axis radius of elliptic equation again by Hough transformation, and ensure that there is between parameter rational size
Relation, it is determined that there is elliptoid castor;
Step B:Luggage case main shaft is commonly present certain angle with image main shaft, and Hough transformation is needed repeatedly, by oval X
Axle/Y-axis rotates multiple angles in the picture and repeats to process, and finds maximum Hough transformation value, and the maximum Hough transformation value is big
In certain threshold value, otherwise must not believe that periphery is present oval;
Step C:Eliminate the obvious irrational castor position of isolated presence or position.
The step 2 is comprised the following steps:
Step A:With gray level threshold segmentation method Luggage trunk body region segmentation out, formed 0-1 binary map;
Step B:The 0-1 binary map obtained in the step A, by way of first micro 12 does opening operation again again, disappear
Except details, the subsidiary stand-alone component of Luggage trunk body surrounding is removed, then micro binary map is enlarged into into original size again, obtained
The body region of Luggage trunk body;
Step C:With the front end of luggage case, rear end, left and right two ends as order, searched near corresponding luggage case border
Rope;
D steps:In radioscopic image, the specific gray scale having using the caster axle of luggage case and RGB pseudo-colourss models
Enclose, the seed region of caster axle is split with the mode of Threshold segmentation.
The method that " point of interest " is detected in the step of methods described three, including:
Step A:It is in seed region, with interest point detecting method proposed by the present invention, close based on caster axle color probability
Degree, finds the point of interest window including caster axle, and the window describes the position of caster axle and size;
Step B:Due to the difference of caster axle imaging angle, caster axle is presented dumbbell shape, calabash shape or sub-circular, wherein
Dumbbell shape, can form 1 " point of interest " in each split, according to syntagmatic, emerging apart from similar two of close, size
Interest point combines, and forms 1 caster axle;
Step C:Eliminate not prominent enough the point of interest of those point of interest correlation properties.
The feature of the caster axle is gray scale, material behavior and the shape of caster axle, and the caster axle characteristic is caster axle
Using light alloy material.Fig. 1 gives the method flow diagram that the present invention carries out castor detection in radioscopic image.Such as accompanying drawing 2
With the typical castor form of luggage case shown in Fig. 3, luggage castor is mainly comprising two classes:One class is that caster axle is embedded in luggage case
In base plate, as shown in Figure 2;Another kind of is that castor is suspended from outside casing, can be rotated freely, as shown in figure 3, the latter is as casing
Outside isolate part, can be excluded with morphological operation with comparalive ease, thus the present invention identification main target be before
Person.As caster axle is the most distinct characteristic of castor, so the present invention first has to recognize caster axle, using simple material information
Feature, obtains the seed region of caster axle;Then the presence of caster axle is searched near seed region;Finally again in caster axle
Round-looking scan castor itself.In order to find the seed region of caster axle, the region easily occurred in castor scans for, so first
Determine 1 overall search scope, be " body region " that this first obtains Luggage trunk body, step includes:
1. trunk area is split, this forms 0-1 binary map with simple gray level threshold segmentation method;
2. the binary map obtained in step 1., micro is 1/12 original size, and the opening operation for carrying out some wheels is eliminated
Details, you can remove the subsidiary stand-alone component of casing surrounding;
3. the micro binary map through opening operation is enlarged into into 12 times again, obtains the body region of casing.
Castor on luggage case is usually present near four sides of luggage case body region, from the habit of safety check execute-in-place
For used, maximum probability is castor in the front end of luggage case, next to that the rear end of luggage case, luggage case or so two ends can
Energy property is less.So, when castor is searched for, with front end, rear end, left and right two ends as order, near corresponding luggage case border
Scan for, in radioscopic image, caster axle generally has relatively low gray scale, and in RGB pseudocolour pictures, G represents green
Component map, which can highlight the scope of subject of caster axle, and this is primarily due to caster axle and is generally made up of the alloy of lightweight.For
This, high energy gray scale is less than certain threshold valueAnd G components are less than certain threshold valuePixel, can be used as caster axle
Sub-pixel, sub-pixel adjacent to each other is coupled together to form UNICOM region, as the seed region of caster axle, and is filtered
The too small seed region of elemental area.Caster axle seed region provides preliminary hunting zone, next with seed region is
Guide, take the method that " point of interest " is detected to determine the region at complete caster axle place.So-called " point of interest ", in this enforcement
The position that the pixel of caster axle color is gathered is referred in example, generally with Score (x, y)=P (cx,y) represent a pixel
" interest " fraction, P (c) is equivalent to caster axle color probability density table, cx,yRepresent the RGB color vector of picture position (x, y).
Formula (1) calculates the dependency of shades of colour vector c and caster axle.The present embodiment acquires E caster axle in advance
Window its comprising caster axle main part, if having N in example eeIndividual pixel,Represent in e-th caster axle window
The RGB color vector of j point,The image coordinate of the point is represented, θ is normalization factor.Formula (1) is the change of Parzen window methods
Shape, this is a kind of method of probability density in space for estimate vector, and in bracket, the factor on the left side is Parzen window methods
Basis, is a Gauss distance function as shown in formula (2), is representedThe probability for belonging to caster axle color to color point c is close
The contribution of degree;AndThen reflectImportance in current window --- castor is represented closer to center
The characteristic of axle, wherein peRepresent the image coordinate of the center of e-th example, ReRepresent the radius of e-th example window.One
" point of interest " region is thick with the pixel related to caster axle, it is meant that its internal average Score (x, y) value is sufficiently large, is this
Need to calculate Mean (x, y, σs) --- point (x, y) is nearby with σsFor the meansigma methodss of the Score (x, y) in the square region of radius.
In order to quickly calculate, the method for taking Integral Image establishes corresponding S tables and ii tables, such as formula (4), (5) institute
Show, wherein ii (x, y) represents the summation of the Score (x, y) from image upper left corner the to point in (x, y) this rectangular area, so as to
Mean (x, y, σ can quickly be calculated with formula (6)s),
S (x, y)=S (x, y-1)+Score (x, y) (4)
Ii (x, y)=ii (x-1, y)+S (x, y) (5)
Caster axle has the slightly weak pixel of interestingness score, can weaken Mean (x, y, σ near the position of peripherys) value,
So that Mean (x, y, σs) maximum σsWindow can not include complete caster axle.For this purpose, each point in seed region is calculated
Each σsDoM (x, y, σ in ranks), such as formula (7), σsS represent the rank of window radius, the present embodiment is provided with about 10 levels
Not, and σs/σs-1=20.25, yardstick sampling property it is enough;The meaning of formula (7) is to find suitable window radius sigmas, investigating scope
Beyond Mean (x, y, σ after this radiuss) significant reduction is there occurs at once, make DoM (x, y, σs) maximum xmax、ymax、σmaxMark
Define the real point of interest region i.e. window for including caster axle main body.
DoM(x,y,σs)=Mean (x, y, σs-1)-Mean(x,y,σs) (7)
In practice, it has been found that due to the difference of imaging angle, caster axle may be presented dumbbell shape, calabash shape or approximate circle
Shape, such as shown in the figure a in Fig. 4, figure b and figure c, wherein dumbbell shape often in each split forms 1 " point of interest ", is
This needs gets up two point of interest window combinations apart from close, similar radius according to syntagmatic, forms 1 caster axle
Window.The above-mentioned color merely with caster axle is the suspicion position that material behavior just have found some caster axles, in order to exclude row
Some metals in Lee's case, for example, the interference of household small electric appliance, in addition it is also necessary to which other shapes and positional information are being sentenced
Disconnected, the wrong report for next carrying out to eliminate majority by the following method is the situation of non-castor:
1) DoM (x of " point of interest "max,ymax,σmax) and Mean (xmax,ymax,σmax) require more than certain threshold value.
2) in the window of the suspicion position of each caster axle for obtaining, which is generally made up of single or 2 " points of interest ", is looked for
To axis of symmetry, and by Score (x, the y) upright projection of each point in the region on axis of symmetry, projection accumulation histogram is formed
Hist;Specifically, it is assumed that the intersection point of point (x, y) to axis of symmetry is (a, b), and (a, b) is in the h lattice of axis of symmetry;If
(x, y) is accumulated in rectangular histogram Hist1 (h) positioned at the upside of axis of symmetry, then Score (x, y), is otherwise accumulated to rectangular histogram
In Hist2 (h), and there is Hist (h)=Hist1 (h)+Hist2 (h);It is required that Hist curves have that both sides are high, middle low become
Gesture, this meets the grown form of caster axle, while requiringIt is sufficiently small, i.e., point of interest region along
Axis of symmetry has preferable symmetry.
3) for the castor shaft position for existing, 1 ellipticity object, i.e. castor itself, method are found in its surrounding
It is to be converted by Hough (Hough);Elliptic equation isThere are 2 parameters --- oval X-axis radius
With Y-axis radius, in caster axle window surrounding certain limit, with Tuscany (Canny) algorithm detected edge points, calculate further according to Hough
Regulation is then put into each marginal point in Hough (A, B) table, and in table, the value of each position is referred to as Hough transformation value, obtains Hough
Maximum combination (the A of transformed valuemax,Bmax), and ensure that two parameter has rational magnitude relationship.Due to chest main shaft and image master
Axle is commonly present certain angle, and such as chest is just put not, so need to be by above-mentioned Hough transformation repeatedly, by oval X-axis/Y-axis
Rotated in the picture, every time such as 10 degree of change, to finding maximum Hough transformation value;Maximum Hough transformation value, also needs
Certain threshold value is greater than, otherwise must not believe that periphery is present oval.
4) as the castor of luggage is that occur in pairs, and the angle of both lines is approximately perpendicular to luggage case main shaft, right
In obvious irrational " castor " position in isolated exist or position, eliminated.
Fig. 5 gives the example of testing process, and figure d is luggage case artwork, and figure e is luggage case segmentation result figure, and figure f is given
The hunting zone of caster axle, wherein can confirm that castor is located at upper and lower ends according to the length-width ratio of parcel cut zone, figure g gives
The point of interest window for detecting, i.e. picture material in black box are gone out, figure h is given in the castor for detecting, i.e. black box
Picture material;The wherein luggage case metal parts of image upper end also with the color similar to caster axle, so also serving as " emerging
Interesting point " is detected, and does not meet yet with its shape, positional information and castor, 1) is washed in a pan to the method described in 4) by above-mentioned
Eliminate;Then it is detected in two real caster axles of image base, wherein the caster axle on right side defines 1 point of interest, and
The caster axle in left side defines 2 points of interest, but defines 1 caster axle by syntagmatic.The invention provides a kind of X is penetrated
In line Security Inspection Equipments image, luggage castor knows method for distinguishing, and the method can find the position of castor in luggage image and model
Enclose, so as to the content analysis to luggage provide extracted information, and reduce the spurious alarm that castor is mistaken for explosive, improve safety
The efficiency of inspection.
The present invention is not limited to above-mentioned embodiment, in the case of the flesh and blood without departing substantially from the present invention, this area skill
Any deformation that art personnel are contemplated that, improvement, replacement each fall within the scope of the present invention.
Claims (6)
1. in a kind of X-ray safety check equipment luggage castor detection method, it is characterised in that comprise the following steps:
Step one:Set up luggage caster axle point of interest detection model;
Step 2:When luggage is entered in the working region of X-ray safety check equipment, according to the position characteristic of caster for trunk,
The body region of Luggage trunk body is determined as overall search scope, the region easily occurred in castor scans for, and obtains described
The seed region of caster axle;
Step 3:With seed region to guide, take the method that point of interest is detected to determine the region at caster axle place;Using step
The rapid one caster axle point of interest detection model set up, searches the point of interest of caster axle color dense distribution near seed region
Region, locks position and the size of caster axle point of interest;
Step 4:The shape obtained according to point of interest or positional information, judge around point of interest with the presence or absence of caster axle, method
It is as follows:In the region that each caster axle suspicion position for obtaining is made up of single or 2 point of interest windows, find symmetrical
Axle, and by the interestingness score upright projection of each pixel in the region on axis of symmetry, it is desirable to the curve two that projection accumulation is formed
Bian Gao, the middle low grown form for meeting caster axle, while requiring that point of interest region has preferable symmetry along axis of symmetry;
Step 5:Castor is determined whether there is in the surrounding of caster axle, method is as follows:
Step A:It is for the castor shaft position for existing, oval in its surrounding searching 1, endpoint detections are first carried out, then is led to
Parameter i.e. X-axis radius and Y-axis radius that Hough transformation determines elliptic equation are crossed, and ensures to close with rational size between parameter
System, it is determined that there is elliptoid castor;
Step B:Luggage case main shaft is commonly present certain angle with image main shaft, and Hough transformation is needed repeatedly, by oval X-axis/Y
The multiple angles that axle rotate in the picture repeat to process, and find maximum Hough transformation value, and the maximum Hough transformation value is greater than one
Determine threshold value, otherwise must not believe that periphery is present oval;
Step C:Eliminate the obvious irrational castor position of isolated presence or position.
2. according to claim 1 in X-ray safety check equipment luggage castor detection method, it is characterised in that it is described
In method and step one, prepare the caster axle example image of a number of manual mark in advance, estimated with the window function estimation technique
Color point belongs to the probability density of caster axle color.
3. according to claim 1 in X-ray safety check equipment luggage castor detection method, it is characterised in that it is described
Method and step two is comprised the following steps:
Step A:With gray level threshold segmentation method Luggage trunk body region segmentation out, formed 0-1 binary map;
Step B:The 0-1 binary map obtained in the step A, by way of first micro 12 does opening operation again again, eliminate thin
Section, removes the subsidiary stand-alone component of Luggage trunk body surrounding, then micro binary map is enlarged into original size again, obtain luggage
The body region of casing;
Step C:With the front end of luggage case, rear end, left and right two ends as order, scan near corresponding luggage case border;
D steps:In radioscopic image, the specific gray scale having using the caster axle of luggage case and RGB pseudo-colourss scopes are used
The seed region of caster axle is split by the mode of Threshold segmentation.
4. according to claim 1 in X-ray safety check equipment luggage castor detection method, it is characterised in that it is described
The method of point of interest detection in the step of method three, including:
Step A:In seed region, the method detected using point of interest, based on caster axle color probability density, is found including foot
The point of interest window of wheel shaft, the window describe the position of caster axle and size;
Step B:Due to the difference of caster axle imaging angle, caster axle is presented dumbbell shape, calabash shape or sub-circular, wherein dumbbell
Type, can form 1 point of interest, in each split according to syntagmatic, apart from two similar point of interest groups of close, size
Altogether, form 1 caster axle;
Step C:Eliminate not prominent enough the point of interest of those point of interest correlation properties.
5. according to claim 1 in X-ray safety check equipment luggage castor detection method, it is characterised in that it is described
The feature of caster axle is gray scale, material behavior and the shape of caster axle.
6. according to claim 1 in X-ray safety check equipment luggage castor detection method, it is characterised in that it is described
Caster axle characteristic adopts light alloy material for caster axle.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410378395.9A CN104166963B (en) | 2014-08-01 | 2014-08-01 | The detection method of luggage castor in a kind of X-ray safety check equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410378395.9A CN104166963B (en) | 2014-08-01 | 2014-08-01 | The detection method of luggage castor in a kind of X-ray safety check equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104166963A CN104166963A (en) | 2014-11-26 |
CN104166963B true CN104166963B (en) | 2017-03-29 |
Family
ID=51910756
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410378395.9A Active CN104166963B (en) | 2014-08-01 | 2014-08-01 | The detection method of luggage castor in a kind of X-ray safety check equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104166963B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113361446B (en) * | 2021-06-22 | 2024-04-16 | 公安部第一研究所 | X-ray package image extraction method |
-
2014
- 2014-08-01 CN CN201410378395.9A patent/CN104166963B/en active Active
Non-Patent Citations (2)
Title |
---|
《双能X 射线旅行箱图像走轮及拉杆识别方法》;黄加翼等;《核电子学与探测技术》;20130831;第33卷(第8期);1021-1026 * |
《双能X射线安检图像的包裹特征剔除》;张娴等;《计算机光盘软件与应用》;20140401;第17卷(第7期);20-22 * |
Also Published As
Publication number | Publication date |
---|---|
CN104166963A (en) | 2014-11-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102028482B (en) | Human body detection image processing method and human body detection apparatus | |
CN105404847B (en) | A kind of residue real-time detection method | |
Patil et al. | Cancer cells detection using digital image processing methods | |
CN106600625A (en) | Image processing method and device for detecting small-sized living thing | |
CN106204510B (en) | A kind of infrared polarization and intensity image fusion method based on structural similarity constraint | |
CN107464252A (en) | A kind of visible ray based on composite character and infrared heterologous image-recognizing method | |
CN104268877B (en) | A kind of infrared image sea horizon self-adapting detecting method | |
CN103400156A (en) | CFAR (Constant False Alarm Rate) and sparse representation-based high-resolution SAR (Synthetic Aperture Radar) image ship detection method | |
CN106682641A (en) | Pedestrian identification method based on image with FHOG- LBPH feature | |
CN104484667A (en) | Contour extraction method based on brightness characteristic and contour integrity | |
CN102722891A (en) | Method for detecting image significance | |
CN105512622B (en) | A kind of visible remote sensing image sea land dividing method based on figure segmentation and supervised learning | |
CN103020959B (en) | Gravity model-based oceanic front information extraction method | |
CN103810716A (en) | Image segmentation method based on grey scale flitting and Renyi entropy | |
Bai et al. | Multi scale top-hat transform based algorithm for image enhancement | |
CN102663723A (en) | Image segmentation method based on color sample and electric field model | |
CN107944392A (en) | A kind of effective ways suitable for cell bayonet Dense crowd monitor video target mark | |
CN102609728A (en) | Method for detecting special pornographic image | |
CN104166963B (en) | The detection method of luggage castor in a kind of X-ray safety check equipment | |
CN104036461A (en) | Infrared complicated background inhibiting method based on combined filtering | |
Zhang et al. | Citrus canker detection based on leaf images analysis | |
CN106508046B (en) | One kind is based on multiple dimensioned bilateral optimized detection method of small target | |
CN105068134A (en) | Method for detecting dangerous goods hidden in shoe by using X-ray multi-view image | |
CN103279960B (en) | A kind of image partition method of human body cache based on X-ray backscatter images | |
CN110097524A (en) | SAR image object detection method based on fusion convolutional neural networks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CB03 | Change of inventor or designer information | ||
CB03 | Change of inventor or designer information |
Inventor after: Wang Yushi Inventor after: Yang Lirui Inventor after: Cha Yanli Inventor after: Kong Weiwu Inventor before: Wang Yushi Inventor before: Yang Lirui Inventor before: Li Baolei Inventor before: Kong Weiwu Inventor before: Cha Yanli |