CN108957572A - A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing - Google Patents

A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing Download PDF

Info

Publication number
CN108957572A
CN108957572A CN201810479352.8A CN201810479352A CN108957572A CN 108957572 A CN108957572 A CN 108957572A CN 201810479352 A CN201810479352 A CN 201810479352A CN 108957572 A CN108957572 A CN 108957572A
Authority
CN
China
Prior art keywords
choice box
image
box
imaging
terahertz
Prior art date
Application number
CN201810479352.8A
Other languages
Chinese (zh)
Inventor
程良伦
何伟健
吴衡
Original Assignee
广东工业大学
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 广东工业大学 filed Critical 广东工业大学
Priority to CN201810479352.8A priority Critical patent/CN108957572A/en
Publication of CN108957572A publication Critical patent/CN108957572A/en

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01VGEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS
    • G01V8/00Prospecting or detecting by optical means
    • G01V8/005Prospecting or detecting by optical means operating with millimetre waves, e.g. measuring the black losey radiation

Abstract

The invention discloses a kind of terahertz imaging methods, method includes the following steps: carrying out stochastical sampling imaging using back-projection algorithm, obtain stochastical sampling image;The maximum point of gray value is determined in stochastical sampling image, and constructs choice box for maximum point;Choice box is handled according to preset rules, obtains target selection frame;Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image;It is superimposed precise image and stochastical sampling image, obtains terahertz image.High-resolution imaging can be carried out to dangerous material region, guarantee the accuracy of identification of dangerous material, and the region because reducing Precise imaging due to accelerates image taking speed, reduces the waste of computer resource.When being applied in real-time safety check, safety check speed can be accelerated, promote safety check quality.The invention also discloses a kind of terahertz imaging device, equipment and readable storage medium storing program for executing, have corresponding technical effect.

Description

A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing

Technical field

The present invention relates to Terahertz Technology field, more particularly to a kind of terahertz imaging method, device, equipment and readable Storage medium.

Background technique

THz wave has the characteristics that penetration power is strong, harmless, therefore THz wave is widely used in examining safely In looking into.For example, channel-type Terahertz human body safety check instrument can carry out safety check imaging to a wide range of interior crowd constantly passed through, greatly Fast safety check speed.

Conventional synthesis aperture radar imaging algorithm-rear orientation projection used in channel-type Terahertz human body safety check instrument Algorithm model is accurate, without signal form, bandwidth and the limitation for accumulating angle, adapts to non-uniform holes and there are larger random The case where kinematic error.But the algorithm needs to traverse all pixels point of imaging region one by one, when imaging region increases Operand will greatly increase, and operation time becomes very long, cannot achieve real time imagery.And if using stochastical sampling method It carries out imaging or directly using other imaging algorithms, although image taking speed can be accelerated, will lead to imaging resolution reduction, no Conducive to the identification of dangerous material.

In conclusion the problems such as how effectively promoting terahertz imaging, is that current those skilled in the art are badly in need of solving The technical issues of.

Summary of the invention

The object of the present invention is to provide a kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing, to accelerate too Hertz image taking speed.

In order to solve the above technical problems, the invention provides the following technical scheme:

A kind of terahertz imaging method, comprising:

Stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image;

The maximum point of gray value is determined in the stochastical sampling image, and is maximum point building selection Frame;

The choice box is handled according to preset rules, obtains target selection frame;

Precise imaging is carried out to the corresponding target area of the target selection frame, obtains precise image;

It is superimposed the precise image and the stochastical sampling image, obtains terahertz image.

Preferably, described that the choice box is handled according to preset rules, obtain target selection frame, comprising:

Expand the choice box according to default choice box growing strategy;

Choice box after will be enlarged by merges processing;

The choice box that area is greater than preset area threshold value is determined as target selection frame.

Preferably, the choice box that area is greater than preset area threshold value is determined as target selection frame, comprising:

Area is greater than the choice box of preset area threshold value to each boundary extension predetermined coefficient times, and by the choosing after extension It selects frame and is determined as target selection frame.

It is preferably, described to expand the choice box according to default choice box growing strategy, comprising:

The pixel that gray value in the stochastical sampling image is greater than default gray threshold is determined as pixel to be selected;

Expand the choice box, forms choice box undetermined;

When the pixel number to be selected covered in the diff area between the choice box undetermined and the choice box Or the choice box is replaced with the choice box undetermined when being greater than accordingly default judgment threshold by percentage.

Preferably, described to expand the choice box, form choice box undetermined, comprising:

The choice box is expanded to surrounding, forms choice box undetermined.

Preferably, described to expand the choice box, form choice box undetermined, comprising:

Orientation expands the choice box, forms choice box undetermined.

Preferably, it is described will be enlarged by after choice box merge processing, comprising:

Calculate the shortest distance between the every two choice box after expanding;

The choice box that the shortest distance is zero is merged.

A kind of terahertz imaging device, comprising:

Stochastical sampling module obtains stochastical sampling image for carrying out stochastical sampling imaging using back-projection algorithm;

Frame construction module is selected, for determining the maximum point of gray value in the stochastical sampling image, and is institute State maximum point building choice box;

Target selection frame obtains module, for handling according to preset rules the choice box, obtains target selection Frame;

Precise imaging module obtains fine for carrying out Precise imaging to the corresponding target area of the target selection frame Image;

Terahertz image obtains module, for being superimposed the precise image and the stochastical sampling image, obtains Terahertz Image.

A kind of terahertz imaging equipment, comprising:

Memory, for storing computer program;

Processor, the step of above-mentioned terahertz imaging method is realized when for executing the computer program.

A kind of readable storage medium storing program for executing is stored with computer program, the computer program quilt on the readable storage medium storing program for executing The step of processor realizes above-mentioned terahertz imaging method when executing.

Using method provided by the embodiment of the present invention, carry out stochastical sampling imaging using back-projection algorithm, obtain with Machine sampled images;The maximum point of gray value is determined in stochastical sampling image, and constructs choice box for maximum point;According to Preset rules handle choice box, obtain target selection frame;To the corresponding target area of target selection frame carry out finely at Picture obtains precise image;It is superimposed precise image and stochastical sampling image, obtains terahertz image.

When carrying out terahertz imaging, stochastical sampling imaging is carried out first with back-projection algorithm, can obtain and adopt at random Sampled images.Wherein, the Terahertz that the gray value of each of stochastical sampling image pixel has respectively represented corresponding position returns Wave signal strength.The Terahertz echo signal intensity of usual dangerous material is relatively high.That is, dangerous material whereabouts can be Occurs the maximum point of gray value in stochastical sampling image.Then, choice box is established for the maximum point determined, and to choosing Frame is selected to be handled, it is final to obtain target selection frame.The corresponding target area of target selection frame is subjected to Precise imaging, can be obtained Precise image, precise image and Terahertz random image are overlapped, can get terahertz image, i.e., completion Terahertz at Picture.In this way, can either guarantee the accuracy of identification of dangerous material to the high-resolution imaging in key area (dangerous material region), and because Reduce the region of Precise imaging and accelerate image taking speed, reduces the waste of computer resource.It, can when being applied in real-time safety check Accelerate safety check speed, promotes safety check quality.

Correspondingly, the embodiment of the invention also provides terahertz imaging corresponding with above-mentioned terahertz imaging method dresses It sets, equipment and readable storage medium storing program for executing, has above-mentioned technique effect, details are not described herein.

Detailed description of the invention

In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with It obtains other drawings based on these drawings.

Fig. 1 is a kind of implementation flow chart of terahertz imaging method in the embodiment of the present invention;

Fig. 2 is the schematic diagram of maximum point in the embodiment of the present invention;

Fig. 3 is the implementation flow chart of another terahertz imaging method in the embodiment of the present invention;

Fig. 4 is the implementation flow chart of another terahertz imaging method in the embodiment of the present invention;

Fig. 5 is the implementation flow chart of another terahertz imaging method in the embodiment of the present invention;

Fig. 6 is the implementation flow chart of another terahertz imaging method in the embodiment of the present invention;

Fig. 7 is that choice box merges schematic diagram in the embodiment of the present invention;

Fig. 8 is the implementation flow chart of another terahertz imaging method in the embodiment of the present invention;

Fig. 9 is the terahertz image of perfectly fine imaging;

Figure 10 is the terahertz image obtained using terahertz imaging method provided in an embodiment of the present invention;

Figure 11 is a kind of structural schematic diagram of terahertz imaging device in the embodiment of the present invention;

Figure 12 is a kind of structural schematic diagram of terahertz imaging equipment in the embodiment of the present invention.

Specific embodiment

Core of the invention is to provide a kind of terahertz imaging method.In view of traditional back-projection algorithm is needed to hardware The all areas scanned are imaged, and imaging resolution is very high, but due to needing to accurately calculate all pixels point, are consumed Shi Taichang is unfavorable for real-time detection.And channel-type Terahertz human body safety check instrument hardware imaging region is larger, but is actually using When significant component of imaging region will be not present people or the object for needing to carry out safety check, imaging meaning pole is carried out to such region It is small, hardware is but expended significantly calculates cost.It is therefore proposed a kind of terahertz imaging method, first carries out stochastical sampling imaging, then Based on stochastical sampling imaging analyze and count, determine the target area for needing Precise imaging, then only to target area into Row Precise imaging, in this way, can accomplish fast imaging under the premise of ensureing the resolution ratio of key area.It is more advantageous in real time Safety check.

Correspondingly, the embodiment of the invention also provides a kind of terahertz imaging device, equipment and readable storage medium storing program for executing, have Above-mentioned technical effect, details are not described herein.

In order to enable those skilled in the art to better understand the solution of the present invention, with reference to the accompanying drawings and detailed description The present invention is described in further detail.Obviously, described embodiments are only a part of the embodiments of the present invention, rather than Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise Under every other embodiment obtained, shall fall within the protection scope of the present invention.

Embodiment one:

Referring to FIG. 1, Fig. 1 be the embodiment of the present invention in a kind of flow chart of terahertz imaging method, this method include with Lower step:

S101, stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image.

In the present embodiment, a kind of channel-type terahertz imaging system can be used to obtain terahertz signal, which uses 4 Hair 8 receives array, and the hardware in conjunction with circuit orbit, and the extensive area of 2m*2m can be imaged simultaneously.

Stochastical sampling imaging is carried out using back-projection algorithm.Specifically, sample boxes can be set as the rectangle of a*b size Entire image is divided into n such rectangle frames by frame, and wherein n is the positive integer greater than 1.K (0 < is chosen in each sample boxes K < a*b) a point is imaged, such as k=2.Then the signal strength maximum value for calculating each sample boxes, as the sample boxes Signal strength indication.Then, two dimensional image Wavelet Denoising Method can be carried out, to result (the i.e. terahertz of each pixel after denoising Hereby echo signal intensity) it is counted, obtain the maximum value and minimum value of the gray value of SAR image.If after Wavelet Denoising Method Image grayscale Value Data be f (x, y), x, y are respectively horizontal, ordinate, then gray value maximum value is max (f (x, y)), minimum Value is min (f (x, y)).

In the process, current imaging results also be can record and whether each pixel carried out the state of imaging.It needs Illustrate it is that the precision of sampling is related with the pixel number of sample boxes size, selection, and then influences to find target area below Process.Specifically, sampling is finer, image taking speed is accelerated more, but target detection success rate is lower.Therefore, k value can be with Predefine or be determined and adjust according to actual needs, for example, when need image taking speed faster when, settable biggish k Value, when needing to detect success rate, then settable lesser k value.

After obtaining stochastical sampling image, step S102 operation can be performed.

S102, the maximum point that gray value is determined in stochastical sampling image, and choice box is constructed for maximum point.

It should be noted that function is in some minimum section, there are independent variable value x, and exist than its it is big with it is smaller Independent variable, functional value corresponding to these independents variable is respectively less than the corresponding functional value of x.So this functional value is known as maximum. Even there is f (x)≤f (x0) to all x in some neighborhood of point x0, then claims f that there is a maximum, maximum f in x0 (x0)." very big " is the concept of a locality.

Based on the probability of above-mentioned maximum, the maximum point of the gray value in stochastical sampling image is searched.Because should be with Machine sampled images are X-Y scheme, therefore, when bigger than the gray value of the pixel of surrounding there are the gray value of a pixel, then Think that the point is maximum point.Specifically, then the point is maximum when occurring being respectively less than this such as adjacent 8 point of Fig. 2 Point.Specifically, referring to FIG. 2, maximum point is indicated with solid black roundlet, when the gray value for a point occur is than known gray scale When the gray scale of adjacent 8 points (being indicated in figure with white roundlet) of value is big, this point is maximum point.It should be noted that pole The quantity of big value point is related to the stochastical sampling image currently obtained, therefore, the maximum point of different stochastical sampling images Number possibility is identical may also be different.After determining maximum point, in stochastical sampling image with choice box by each Maximum point is framed.That is, there are how many a maximum points, with regard to how many choice box.Specifically, the choice box can Think box.If the upper and lower, left and right boundary of choice box use respectively a, b, c, d represent, then choice box can be denoted as box (a, B, c, d), area S=(b-a+1) * (d-c+1).The side length of the choice box is unit length 1.

S103, choice box is handled according to preset rules, obtains target selection frame.

In the present embodiment, can preset the rule of the processing of choice box, the rule may include to choice box into The processing mode of combination between the one or more of them of processing modes such as row selection, growth expansion, merging.

After maximum point is framed with choice box, it can use the pre-set rule handle to choice box Then, choice box is handled, it is final to obtain target selection frame.That is, further being obtained by handling choice box Obtain target selection frame.The number of target selection frame that is obtained after processing choice box, size can according to the actual situation depending on.Example Such as, if determining 20 maximum points in step s 102, then the number of choice box is also 20, and target selection frame then base It is obtained after this 20 choice boxs are handled, the number of target selection frame still can be 20, or be lower than 20;Target The size of choice box can be identical as the size of choice box, can also be bigger than choice box.

S104, Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image.

After obtaining target selection frame, Precise imaging is carried out to the corresponding target area of target selection frame, can get fine Image.Specifically, all pixels point in the corresponding target area of target selection frame can be traversed one by one, carry out finely at Picture finally obtains precise image.Because in step s101 to have carried out random imaging, i.e., there is the picture for having become picture in target area Vegetarian refreshments can traverse one by one in practical applications only for pixel unimaged in target area, carry out Precise imaging.

S105, superposition precise image and stochastical sampling image, obtain terahertz image.

After obtaining Precise imaging, Precise imaging can be overlapped in stochastical sampling image, finally obtain imaging region Interior complete terahertz image.Resolution ratio in the terahertz image relative to the corresponding position in target area is higher, and for The corresponding position in nontarget area is then lower.

Using method provided by the embodiment of the present invention, carry out stochastical sampling imaging using back-projection algorithm, obtain with Machine sampled images;The maximum point of gray value is determined in stochastical sampling image, and constructs choice box for maximum point;According to Preset rules handle choice box, obtain target selection frame;To the corresponding target area of target selection frame carry out finely at Picture obtains precise image;It is superimposed precise image and stochastical sampling image, obtains terahertz image.When carrying out terahertz imaging, Stochastical sampling imaging is carried out first with back-projection algorithm, stochastical sampling image can be obtained.Wherein, in stochastical sampling image The gray value of each pixel has respectively represented the Terahertz echo signal intensity of corresponding position.The Terahertz of usual dangerous material Echo signal intensity is relatively high.That is, the pole of gray value can occur in stochastical sampling image in dangerous material whereabouts Big value point.Then, choice box is established for the maximum point determined, and choice box is handled, it is final to obtain target choosing Select frame.The corresponding target area of target selection frame is subjected to Precise imaging, precise image can be obtained, by precise image and terahertz Hereby random image is overlapped, and can get terahertz image, i.e. completion terahertz imaging.In this way, can be (dangerous to key area Product region) high-resolution imaging is carried out, guarantee the accuracy of identification of dangerous material, and the region because reducing Precise imaging due to accelerates into As speed, the waste of computer resource is reduced.When being applied in real-time safety check, safety check speed can be accelerated, promote safety check quality.

It should be noted that based on the above embodiment one, the embodiment of the invention also provides be correspondingly improved scheme.Rear Involved in continuous embodiment with can mutually be referred between same steps or corresponding steps in above-described embodiment one, corresponding beneficial effect Can also be cross-referenced, it is no longer repeated one by one in improvement embodiment below.

Embodiment two:

Referring to FIG. 3, Fig. 3 is the flow chart of another terahertz imaging method in the embodiment of the present invention, this method includes Following steps:

S201, stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image.

S202, the maximum point that gray value is determined in stochastical sampling image, and choice box is constructed for maximum point.

S203, expand choice box according to default choice box growing strategy.

In the present embodiment, choice box growing strategy can be preset.For example, when reaching the pixel around choice box Gray value when being greater than default gray threshold, will be enlarged by the choice box, to will be greater than the pixel covering of default gray threshold.

After constructing choice box, expanded according to the growing strategy of default choice box.Because in stochastical sampling image The gray value of pixel is not quite identical, so, the expansion program of each choice box, and shape after expanding may also can not Unanimously.

S204, will be enlarged by after choice box merge processing.

Cause meets default choice box growing strategy, so that it may expand the choice box, it is understood that there may be two or more Choice box mutually covers between choice box after expansion, intersects, each other nonseptate situation.In order to avoid it is subsequent at The pixel for repeating covering to two or more choice boxs as during carries out repeating imaging, it is possible to after will be enlarged by Mutually covering, intersects, nonseptate choice box merges each other.When merging, another, which can be used directly, to cover The minimum choice box for two choice boxs that lid currently mutually covers is replaced, that is to say, that a choice box can be used and replace Change original two and more than two choice boxs.That is, the choice box after merging is than the choice box before merging Quantity it is less.Merge it is primary after, the choice box after merging if it exists meets with other choice boxs again when merging condition, then again It is secondary to merge operation, until merging condition is not satisfied between all choice boxs.Certainly, for being unsatisfactory for merging item The choice box of part then can be without processing.

S205, the choice box that area is greater than preset area threshold value is determined as target selection frame.

In the present embodiment, one can be preset for screening a threshold value of choice box, which is specially one A area threshold.

It is influenced by noise jamming or other accidentalia, may have part choice box and not expanded or given birth to always It is long, be also unsatisfactory for merging and require so that the area of the choice box is smaller, herein it is considered that the choice box it is corresponding greatly Value point may and much be generated due to the factor of dangerous product, but be generated because of other accidentalia, The choice box can be given up.That is, only leaving the choice box that area is greater than preset area threshold value, and give up face Product is less than the choice box of preset area threshold value, in this way, the carry out Precise imaging to meaningless region can be reduced, reduces computer money The waste in source can accelerate image taking speed.

S206, Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image.

S207, superposition precise image and stochastical sampling image, obtain terahertz image.

Embodiment three,

Referring to FIG. 4, Fig. 4 is the flow chart of another terahertz imaging method in the embodiment of the present invention, this method includes Following steps:

S301, stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image.

S302, the maximum point that gray value is determined in stochastical sampling image, and choice box is constructed for maximum point.

S303, expand choice box according to default choice box growing strategy.

S304, will be enlarged by after choice box merge processing.

S305, the choice box by area greater than preset area threshold value will extend to each boundary extension predetermined coefficient times Choice box afterwards is determined as target selection frame.

Specifically, can first remove the choice box that all areas are less than preset area threshold value.Image quality in order to balance is kept away Exempt from lost part fringe region, (λ is constant, can be adjusted according to the actual situation to all existing each λ times of boundary extensions of choice box It is whole), i.e., the choice box box (a, b, c, d) that area is greater than preset area threshold value is replaced are as follows:

box(a-λ(b-a),b+λ(b-a),c-λ(d-c),d+λ(d-c)).And the choice box after extension is determined as target Choice box.

After being extended, intersection judgement can also be carried out to remaining all choice boxs again, by the choosing of all intersections It selects frame to merge, it is ensured that be not overlapped again between final target selection frame region.

S306, Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image.

S307, superposition precise image and stochastical sampling image, obtain terahertz image.

Example IV:

Referring to FIG. 5, Fig. 5 is the flow chart of another terahertz imaging method in the embodiment of the present invention, this method includes Following steps:

S401, stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image.

S402, the maximum point that gray value is determined in stochastical sampling image, and choice box is constructed for maximum point.

S403, the pixel that the gray value in stochastical sampling image is greater than default gray threshold is determined as pixel to be selected Point.

S404, expand choice box, form choice box undetermined.

S405, when the pixel number to be selected or percentage covered in the diff area between choice box undetermined and choice box When greater than corresponding default judgment threshold, choice box is replaced with into choice box undetermined.

It is illustrated for ease of description, below combining tri- steps of above-mentioned steps S403, S404 and S405.

In view of the gray value of the corresponding pixel in the region of dangerous material is greater than the gray scale of the pixel in non-dangerous article region Value, so choosing the pixel that gray value is greater than default gray threshold.Choose all pixels for meeting following formula:

F (x, y) >-W1[max (f (x, y))-min (f (x, y))]+max (f (x, y)), that is to say, that select Pixel f (x, y) to be selected is signal strength W1Preceding pixel.

Then expand choice box, form choice box undetermined, choice box undetermined can be that each boundary value of choice box is past Choice box after one unit of outer expansion, for example, then choice box undetermined is box (a-1, b+ if choice box is box (a, b, c, d) 1,c-1,d+1);It is undetermined select frame can for by one of boundary of choice box toward the choice box after one unit of external expansion, example If choice box is that then choice box undetermined is box (a, b, c, d+1) to box (a, b, c, d).Certainly, the degree for expanding or extending is also It can be multiple parasangs.

Then, the corresponding pixel in region of expansion or dilation is counted.Specifically, if choice box box1 is passed through Choice box box2 to be selected is obtained after crossing extension or expansion, then the corresponding region box2-box1 is counted.Work as box2-box1 The number or percentage of pixel to be selected in pixel in region meet corresponding threshold value, then replace with choice box box1 Pixel box2 to be selected, specifically, if choice box box1 be box1 (a1, b1, c1, d1), choice box undetermined be box2 (a2, b2, C2, d2), then replaced choice box box1 is box1 (a2, b2, c2, d2).

Preferably, in above-mentioned step two, expansion choice box, during forming choice box undetermined, there is the following two kinds expansion One such or two kinds of modes combined can be used in practical applications and expand choice box for big mode.

The first expands mode, and choice box is expanded to surrounding, forms choice box undetermined.

Specifically, being then box to the choice box box2 undetermined after surrounding expansion if choice box box1 is box (a, b, c, d) (a-1,b+1,c-1,d+1)。

Then after obtaining choice box undetermined, all choice box box (a, b, c, d) are judged whether to meet following formula:

Pxpercent (box (a-1, b+1, c-1, d+1)-box (a, b, c, d)) > W2, wherein pxpercent (box) be Selected pixels point accounts for the percentage of selection frame region whole pixel, (box (a-1, b+1, c-1, d+ in the region choice box box 1)-box (a, b, c, d)) indicate that box2-box1 is the part that the region box1 is removed in the region choice box box2, W2To preset Percentage threshold.If it is satisfied, then substituting former choice box with the choice box box (a-1, b+1, c-1, d+1) undetermined after expansion box(a,b,c,d).That is, replaced choice box box2 is box (a-1, b+1, c-1, d+1).

Second of expansion mode, orientation expand choice box, form choice box undetermined.

Specifically, if choice box box1 be box (a, b, c, d), then to orientation expand it is one unit greater after choice box undetermined Box2 has following 4 kinds of forms of expression:

1,box(a-1,b,c,d);

2,box(a,b+1,c,d);

3,box(a,b,c-1,d);

4、box(a,b,c,d+1)。

After obtaining choice box undetermined, following judgement is done:

Judgement formula 1, pxMaxNum (box (a-1, b, c, d)-box (a, b, c, d)) >=W3

Judgement formula 2, pxMaxNum (box (a, b+1, c, d)-box (a, b, c, d)) >=W3

Judgement formula 3, pxMaxNum (box (a, b, c-1, d)-box (a, b, c, d)) >=W3

Judgement formula 4, pxMaxNum (box (a, b, c, d+1)-box (a, b, c, d)) >=W3

Wherein, pxMaxNum (box) is the pixel number that the continuous selected pixels point in a direction is most in the region choice box box Amount, box2-box1 are the part that the region box1 is removed in the region choice box box2, W3For the number threshold value of pixel to be selected.

When judging that formula 1 is set up, former choice box box is substituted with the choice box box (a-1, b, c, d) undetermined after growing up (a,b,c,d)。

It can similarly obtain, when corresponding above formula is set up, respectively correspondingly with choice box box (a, b+1, c, d) undetermined, box (a, B, c-1, d), box (a, b, c, d+1) substitutes former choice box.Circulation is until growth is not satisfied in all boundaries of all choice boxs Condition.

S406, will be enlarged by after choice box merge processing.

S407, the choice box that area is greater than preset area threshold value is determined as target selection frame.

S408, Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image.

S409, superposition precise image and stochastical sampling image, obtain terahertz image.

Embodiment five:

Referring to FIG. 6, Fig. 6 is the flow chart of another terahertz imaging method in the embodiment of the present invention, this method includes Following steps:

S501, stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image.

S502, the maximum point that gray value is determined in stochastical sampling image, and choice box is constructed for maximum point.

S503, expand choice box according to default choice box growing strategy.

The shortest distance between every two choice box after S504, calculating expansion, and the choice box for being zero by the shortest distance It merges.

The shortest distance between every two choice box is the nearest direct range of choice box frame, when two choice boxs intersect Or when adjacent, the shortest distance is 0.Judge the method that whether two choice boxs intersect (containing adjacent): for choice box box (a1, b1,c1,d1) and box (a2,b2,c2,d2), judge whether following formula is true respectively:

|c1+d1-c2-d2|≤|d1+d2-c1-c2|;

|a1+b1-a2-b2|≤|b1+b2-a1-a2|;

When at least meeting one in above-mentioned two formula, two choice box intersections, particularly when any formula equal sign is set up two A choice box is adjacent, and two choice boxs are adjacent in embodiments of the present invention is considered as intersection.

When two choice boxs are non-intersecting, the shortest distance between choice box is calculated using coordinate quadrant method.It is right In calculating choice box box1 (a1,b1,c1,d1) and box2 (a2,b2,c2,d2) distance dis (box1, box2), calculate step It is as follows:

Step 1: taking the central point of choice box box2One 2 dimension coordinate system xy is established for origin, is sentenced Break the central point of another choice box box1In which quadrant of this coordinate system.

Step 2: taking a different points to be compared on choice box frame respectively according to the different quadrant in place, i.e., Take two choice boxs respectively closest to the endpoint of another choice box;If two points got are respectively (x1,y1) and (x2,y2)。

Step 3: the vector when the central point of choice box box1 is in first and third quadrantOtherwise

Step 4: by judging vectorTo calculate the distance of two choice boxs.Calculation formula is as follows:

Choice box merging method: for choice box box (a1,b1,c1,d1) and box (a2,b2,c2,d2), then the choosing after merging Selecting frame is box (min (a1,a2),max(b1,b2),min(c1,c2),max(d1,d2)).Specifically, reference can be made to merging shown in Fig. 7 After box1 and box2, box3 is obtained, it should be noted that, box1 and box2 in Fig. 7 do not intersect, which is only to illustrate to merge Mode.It should be noted that can also be merged when the distance between two choice boxs is not 0.For example, if merging it The number of the pixel to be selected in the region that choice box newly covers afterwards can reach certain numerical value and can also merge.

S505, the choice box by area greater than preset area threshold value will extend to each boundary extension predetermined coefficient times Choice box afterwards is determined as target selection frame.

S506, Precise imaging is carried out to the corresponding target area of target selection frame, obtains precise image.

S507, superposition precise image and stochastical sampling image, obtain terahertz image.

Embodiment six:

For convenient for those skilled in the art understand that technical solution provided by the embodiment of the present invention, below with specific specific Example is provided for the embodiments of the invention technical solution and is described in detail:

As shown in figure 8, method provided by the embodiment of the present invention is a kind of a wide range of Terahertz choice box growth district choosing Select fast imaging method.Existing, the image handled is in perfectly fine imaging as shown in figure 9, photo resolution is 400*400。

The specific implementation steps are as follows:

S601: first with traditional radar imaging method-back-projection algorithm and sample boxes method carry out stochastical sampling at Picture.If sample boxes are the rectangle frame of 3*3 size, entire image is divided into 17689 such sample boxes.In each sample boxes 2 points are chosen to be imaged.Then the signal strength maximum value for calculating each sample boxes, the signal strength as the sample boxes Value.In the process, current imaging results are recorded and whether each pixel carried out the state of imaging.

S602;Two dimensional image Wavelet Denoising Method is carried out, to result (the Terahertz echo-signal of each pixel after denoising Intensity) it is counted, obtain the gray value of SAR (Synthetic Aperture Radar, synthetic aperture radar) image most Big value and minimum value.If the image grayscale Value Data after Wavelet Denoising Method is f (x, y), x, y are respectively horizontal, ordinate, then grey Angle value maximum value is max (f (x, y)), and minimum value is min (f (x, y)).

S603: and then the gray value maximum point on lookup X-Y scheme, it is alternatively that the seed point of frame growth algorithm.Then Each seed point is framed with choice box in two-dimensional imaging result figure.The choice box is box.If choice box it is upper, Under, left and right boundary be respectively a, b, c, d, then choice box is denoted as box (a, b, c, d), area S=(b-a+1) * (d-c+ 1).The side length for defining choice box is unit length 1.The distance between every two choice box be choice box frame it is nearest it is direct away from From when two choice boxs intersect or is adjacent, distance is 0.Judge the method that whether two choice boxs intersect (containing adjacent): for Choice box box (a1,b1,c1,d1) and box (a2,b2,c2,d2), judge respectively

|c1+d1-c2-d2|≤|d1+d2-c1-c2|;

|a1+b1-a2-b2|≤|b1+b2-a1-a2|;

It is whether true.Two choice box intersections when at least meeting one in above-mentioned two formula, particularly when any formula equal sign Two choice boxs are adjacent when establishment, and choice box is adjacent in the method is considered as intersection.

When two choice boxs are non-intersecting, the distance between choice box is calculated using coordinate quadrant method.For meter Calculate choice box box1 (a1,b1,c1,d1) and box2 (a2,b2,c2,d2) distance dis (box1, box2), calculate steps are as follows:

Step 1: taking the central point of choice box box2One 2 dimension coordinate system is established (respectively for origin Represented with x and y), judge the central point of another choice box box1This coordinate system which as Limit.

Step 2: taking a different points to be compared on choice box frame respectively according to the different quadrant in place, i.e., Take two choice boxs respectively closest to the endpoint of another choice box;If two points got are respectively (x1,y1) and (x2,y2)。

Step 3: the vector when the central point of choice box box1 is in first and third quadrantOtherwise

Step 4: by judging vectorTo calculate the distance of two choice boxs.Calculation formula is as follows:

Choice box merging method: for choice box box (a1,b1,c1,d1) and box (a2,b2,c2,d2), then the choosing after merging Selecting frame is box (min (a1,a2),max(b1,b2),min(c1,c2),max(d1,d2))。

S604: W before all signal strengths is chosen1=30% pixel.Choose all pixels for meeting following formula:

F (x, y) >-W1·[max(f(x,y))-min(f(x,y))]+max(f(x,y));Wherein, pxpercent (box) percentage of selection frame region whole pixel is accounted for for selected pixels point in the region choice box box, box2-box1 is choosing The part that the region box1 is removed in the region frame box2 is selected, then following formula is judged to all choice box box (a, b, c, d)

Pxpercent (box (a-1, b+1, c-1, d+1)-box (a, b, c, d)) > W2It is whether true, if so, then use Choice box box (a-1, b+1, c-1, d+1) after expansion substitutes original choice box box (a, b, c, d).W in above formula2=80% is root The threshold value being arranged according to experience and actual conditions.

S605: the pixel of selection is identical as step S604.PxMaxNum (box) is a direction in the region choice box box The most pixel quantity of continuous selected pixels point, box2-box1 is the part that the region box1 is removed in the region choice box box2, then right All choice box box (a, b, c, d) successively judge

pxMaxNum(box(a-1,b,c,d)-box(a,b,c,d))≥W3, if it sets up, if so, then with raw upwards Choice box box (a-1, b, c, d) after length substitutes original choice box box (a, b, c, d).W in above formula3=3 be rule of thumb and real The threshold value of border situation setting.Similarly, then successively judge whether following three formula is true:

pxMaxNum(box(a,b+1,c,d)-box(a,b,c,d))≥W3

pxMaxNum(box(a,b,c-1,d)-box(a,b,c,d))≥W3

pxMaxNum(box(a,b,c,d+1)-box(a,b,c,d))≥W3

If so, box (a, b+1, c, d) respectively correspondingly then is used, box (a, b, c-1, d), box (a, b, c, d+1) substitution Former choice box.Circulation is until growth conditions is not satisfied in all boundaries of all choice boxs.Each pair of choice box carries out growth expansion Afterwards, distance between any two is calculated to all choice boxs, distance is merged for 0 two choice boxs.

S606: condition judgement is merged to all existing choice boxs between any two, meets merging for condition.It closes And condition is as follows: the pixel of selection is identical as step S603, if two original choice boxs are respectivelyWith box2(a2,b2,c2,d2), area S1、S2, the region after merging is box3 (a3,b3,c3,d3), area S2, region 4 There is box4=box3-box2-box1.As shown in fig. 7, choice box box1 and choice box box 2, the then new choice box merged Box3, region 4 are the white portion region in Fig. 7.Then merge condition:

(S1+S2)(1+pxpercent(box4)-0.5*lg(dis(box1,box2)))≥S3.Wherein S3To merge condition Threshold value.Traversal is re-started if having choice box to merge in the process, until all existing choice boxs two-by-two it Between merging condition is not satisfied.

S607: remove all areas less than area threshold W4=25 choice box.Image quality in order to balance avoids losing Part edge region, to all existing each boundary extension of choice box λ=0.05 times, i.e., by all choice box box (a, b, c, d) Replacement are as follows:

Box (a- λ (b-a), b+ λ (b-a), c- λ (d-c), d+ λ (d-c)) then carries out intersection to all choice boxs again and sentences It is disconnected, the choice box of all intersections is merged, it is ensured that be not overlapped again between all selection frame regions.

S608: finally there are the regions of all choice boxs be required interesting target region.Using traditional Back-projection algorithm carries out Precise imaging to all area-of-interests, has carried out the pixel of imaging in step s 601 at this time Point is not repeated to calculate.

S609: Precise imaging is superimposed to obtain final imaging results by pixel with the imaging results of stochastical sampling.Finally Imaging results are as shown in Figure 10.I.e. imaging is as a result, terahertz image namely as described above, black wire frame are by choosing Select the target selection frame obtained after frame is handled.As can be seen that using terahertz imaging side provided by the embodiment of the present invention Imaging results of the terahertz image that method finally obtains in dangerous material i.e. target area are consistent, that is, are not influenced Subsequent identification judging result.And Precise imaging is then reduced for nontarget area, reduce the waste of computer resource, can add Fast image taking speed.

Corresponding to above method embodiment, the embodiment of the invention also provides a kind of terahertz imaging devices, hereafter retouch The terahertz imaging device stated can correspond to each other reference with above-described terahertz imaging method.

Shown in Figure 11, which comprises the following modules:

Stochastical sampling module 101 obtains stochastical sampling figure for carrying out stochastical sampling imaging using back-projection algorithm Picture;

Frame construction module 102 is selected, for determining the maximum point of gray value in stochastical sampling image, and is very big Value point building choice box;

Target selection frame obtains module 103, for handling according to preset rules choice box, obtains target selection Frame;

Precise imaging module 104 obtains fine figure for carrying out Precise imaging to the corresponding target area of target selection frame Picture;

Terahertz image obtains module 105, for being superimposed precise image and stochastical sampling image, obtains terahertz image.

Using device provided by the embodiment of the present invention, carry out stochastical sampling imaging using back-projection algorithm, obtain with Machine sampled images;The maximum point of gray value is determined in stochastical sampling image, and constructs choice box for maximum point;According to Preset rules handle choice box, obtain target selection frame;To the corresponding target area of target selection frame carry out finely at Picture obtains precise image;It is superimposed precise image and stochastical sampling image, obtains terahertz image.When carrying out terahertz imaging, Stochastical sampling imaging is carried out first with back-projection algorithm, stochastical sampling image can be obtained.Wherein, in stochastical sampling image The gray value of each pixel has respectively represented the Terahertz echo signal intensity of corresponding position.The Terahertz of usual dangerous material Echo signal intensity is relatively high.That is, the pole of gray value can occur in stochastical sampling image in dangerous material whereabouts Big value point.Then, choice box is established for the maximum point determined, and choice box is handled, it is final to obtain target choosing Select frame.The corresponding target area of target selection frame is subjected to Precise imaging, precise image can be obtained, by precise image and terahertz Hereby random image is overlapped, and can get terahertz image, i.e. completion terahertz imaging.In this way, can either be to key area (danger Dangerous product region) high-resolution imaging is carried out, guarantee the accuracy of identification of dangerous material, and the region because reducing Precise imaging due to accelerates Image taking speed reduces the waste of computer resource.When being applied in real-time safety check, safety check speed can be accelerated, promote safety check quality.

In a kind of specific embodiment of the invention, target selection frame obtains module 103, specifically includes:

Expanding unit, for expanding choice box according to default choice box growing strategy;

Combining unit merges processing for the choice box after will be enlarged by;

Object selection frame determination unit, the choice box for area to be greater than preset area threshold value are determined as target selection Frame.

In a kind of specific embodiment of the invention, Object selection frame determination unit is specifically used for for area being greater than pre- If the choice box after extension is determined as target selection to each boundary extension predetermined coefficient times by the choice box of area threshold Frame.

In a kind of specific embodiment of the invention, expanding unit is specifically included:

Pixel to be selected determines subelement, for the gray value in stochastical sampling image to be greater than to the picture of default gray threshold Vegetarian refreshments is determined as pixel to be selected;

Expand subelement and forms choice box undetermined for expanding choice box;

Choice box replaces subelement, for when the picture to be selected covered in the diff area between choice box undetermined and choice box When vegetarian refreshments number or percentage are greater than accordingly default judgment threshold, choice box is replaced with into choice box undetermined.

In a kind of specific embodiment of the invention, expand subelement, is specifically used for expanding choice box to surrounding, shape At choice box undetermined.

In a kind of specific embodiment of the invention, expand subelement, be specifically used for orientation expand choice box, formed to Determine choice box.

In a kind of specific embodiment of the invention, combining unit is selected specifically for calculating the every two after expanding The shortest distance between frame, and the choice box that the shortest distance is zero is merged.

Corresponding to above method embodiment, the embodiment of the invention also provides a kind of terahertz imaging equipment, hereafter retouch A kind of terahertz imaging equipment stated can correspond to each other reference with a kind of above-described terahertz imaging method.

Shown in Figure 12, which includes:

Memory D1, for storing computer program;

Processor D2 realizes the step of the terahertz imaging method of above method embodiment when for executing computer program Suddenly.

Corresponding to above method embodiment, the embodiment of the invention also provides a kind of readable storage medium storing program for executing, are described below A kind of readable storage medium storing program for executing can correspond to each other reference with a kind of above-described terahertz imaging method.

A kind of readable storage medium storing program for executing is stored with computer program on readable storage medium storing program for executing, and computer program is held by processor The step of terahertz imaging method of above method embodiment is realized when row.

The readable storage medium storing program for executing be specifically as follows USB flash disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), the various program storage generations such as random access memory (Random Access Memory, RAM), magnetic or disk The readable storage medium storing program for executing of code.

Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other The difference of embodiment, same or similar part may refer to each other between each embodiment.For being filled disclosed in embodiment For setting, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is referring to method part Explanation.

Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These Function is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Profession Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered Think beyond the scope of this invention.

The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology In any other form of storage medium well known in field.

Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said It is bright to be merely used to help understand technical solution of the present invention and its core concept.It should be pointed out that for the common of the art , without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for technical staff, these Improvement and modification are also fallen within the protection scope of the claims of the present invention.

Claims (10)

1. a kind of terahertz imaging method characterized by comprising
Stochastical sampling imaging is carried out using back-projection algorithm, obtains stochastical sampling image;
The maximum point of gray value is determined in the stochastical sampling image, and constructs choice box for the maximum point;
The choice box is handled according to preset rules, obtains target selection frame;
Precise imaging is carried out to the corresponding target area of the target selection frame, obtains precise image;
It is superimposed the precise image and the stochastical sampling image, obtains terahertz image.
2. terahertz imaging method according to claim 1, which is characterized in that it is described according to preset rules to the selection Frame is handled, and target selection frame is obtained, comprising:
Expand the choice box according to default choice box growing strategy;
Choice box after will be enlarged by merges processing;
The choice box that area is greater than preset area threshold value is determined as target selection frame.
3. terahertz image imaging method according to claim 2, which is characterized in that described that area is greater than preset area The choice box of threshold value is determined as target selection frame, comprising:
Area is greater than the choice box of preset area threshold value to each boundary extension predetermined coefficient times, and by the choice box after extension It is determined as target selection frame.
4. terahertz imaging method according to claim 2, which is characterized in that described according to default choice box growing strategy Expand the choice box, comprising:
The pixel that gray value in the stochastical sampling image is greater than default gray threshold is determined as pixel to be selected;
Expand the choice box, forms choice box undetermined;
When the pixel number to be selected covered in the diff area between the choice box undetermined and the choice box or hundred When dividing than being greater than corresponding default judgment threshold, the choice box is replaced with into the choice box undetermined.
5. terahertz imaging method according to claim 4, which is characterized in that it is described to expand the choice box, formed to Determine choice box, comprising:
The choice box is expanded to surrounding, forms choice box undetermined.
6. terahertz imaging method according to claim 4, which is characterized in that it is described to expand the choice box, formed to Determine choice box, comprising:
Orientation expands the choice box, forms choice box undetermined.
7. according to the described in any item terahertz imaging methods of claim 2 to 6, which is characterized in that it is described will be enlarged by after choosing It selects frame and merges processing, comprising:
Calculate the shortest distance between the every two choice box after expanding;
The choice box that the shortest distance is zero is merged.
8. a kind of terahertz imaging device characterized by comprising
Stochastical sampling module obtains stochastical sampling image for carrying out stochastical sampling imaging using back-projection algorithm;
Frame construction module is selected, for determining the maximum point of gray value in the stochastical sampling image, and is the pole Big value point building choice box;
Target selection frame obtains module, for handling according to preset rules the choice box, obtains target selection frame;
Precise imaging module obtains precise image for carrying out Precise imaging to the corresponding target area of the target selection frame;
Terahertz image obtains module, for being superimposed the precise image and the stochastical sampling image, obtains terahertz image.
9. a kind of terahertz imaging equipment characterized by comprising
Memory, for storing computer program;
Processor realizes the terahertz imaging method as described in any one of claim 1 to 7 when for executing the computer program The step of.
10. a kind of readable storage medium storing program for executing, which is characterized in that be stored with computer program, the meter on the readable storage medium storing program for executing It is realized when calculation machine program is executed by processor as described in any one of claim 1 to 7 the step of terahertz imaging method.
CN201810479352.8A 2018-05-18 2018-05-18 A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing CN108957572A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810479352.8A CN108957572A (en) 2018-05-18 2018-05-18 A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810479352.8A CN108957572A (en) 2018-05-18 2018-05-18 A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing

Publications (1)

Publication Number Publication Date
CN108957572A true CN108957572A (en) 2018-12-07

Family

ID=64499271

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810479352.8A CN108957572A (en) 2018-05-18 2018-05-18 A kind of terahertz imaging method, device, equipment and readable storage medium storing program for executing

Country Status (1)

Country Link
CN (1) CN108957572A (en)

Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040062441A1 (en) * 2000-12-06 2004-04-01 Jerome Meniere Method for detecting new objects in an illuminated scene
US20040095484A1 (en) * 2002-11-14 2004-05-20 Beardsley Paul A. Object segmentation from images acquired by handheld cameras
CN101286233A (en) * 2008-05-19 2008-10-15 重庆邮电大学 Fuzzy edge detection method based on object cloud
CN101556695A (en) * 2009-05-15 2009-10-14 广东工业大学 Image matching method
CN101916373A (en) * 2010-08-11 2010-12-15 西安电子科技大学 Road semiautomatic extraction method based on wavelet detection and ridge line tracking
CN102621548A (en) * 2012-04-17 2012-08-01 中南大学 Multi-scale backward projection imaging method for ground penetrating radar
CN102629386A (en) * 2012-03-28 2012-08-08 浙江大学 Region segmentation method for colorful textile texture images
CN105335960A (en) * 2014-08-13 2016-02-17 温州大学 Image segmentation method combining edge detection algorithm with watershed algorithm
CN106250895A (en) * 2016-08-15 2016-12-21 北京理工大学 A kind of remote sensing image region of interest area detecting method
CN106308795A (en) * 2016-08-31 2017-01-11 北京农业信息技术研究中心 Warning device for foot-and-mouth diseases of animals based on terahertz imaging technology and system and method thereof
CN106447659A (en) * 2016-09-27 2017-02-22 电子科技大学 Region growth detection method based on multiple judgments
CN106683076A (en) * 2016-11-24 2017-05-17 南京航空航天大学 Texture feature clustering-based locomotive wheelset tread damage detection method
CN106952315A (en) * 2017-03-22 2017-07-14 广东工业大学 A kind of method that image quick reconfiguration is carried out to Terahertz complex-valued data based on BFGS
CN107092040A (en) * 2017-06-01 2017-08-25 上海理工大学 Terahertz imaging rays safety detection apparatus and video procession method
CN107255641A (en) * 2017-06-06 2017-10-17 西安理工大学 A kind of method that Machine Vision Detection is carried out for GRIN Lens surface defect
CN107665486A (en) * 2017-09-30 2018-02-06 深圳绰曦互动科技有限公司 A kind of method for automatically split-jointing, device and terminal device applied to radioscopic image

Patent Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040062441A1 (en) * 2000-12-06 2004-04-01 Jerome Meniere Method for detecting new objects in an illuminated scene
US20040095484A1 (en) * 2002-11-14 2004-05-20 Beardsley Paul A. Object segmentation from images acquired by handheld cameras
CN101286233A (en) * 2008-05-19 2008-10-15 重庆邮电大学 Fuzzy edge detection method based on object cloud
CN101556695A (en) * 2009-05-15 2009-10-14 广东工业大学 Image matching method
CN101916373A (en) * 2010-08-11 2010-12-15 西安电子科技大学 Road semiautomatic extraction method based on wavelet detection and ridge line tracking
CN102629386A (en) * 2012-03-28 2012-08-08 浙江大学 Region segmentation method for colorful textile texture images
CN102621548A (en) * 2012-04-17 2012-08-01 中南大学 Multi-scale backward projection imaging method for ground penetrating radar
CN105335960A (en) * 2014-08-13 2016-02-17 温州大学 Image segmentation method combining edge detection algorithm with watershed algorithm
CN106250895A (en) * 2016-08-15 2016-12-21 北京理工大学 A kind of remote sensing image region of interest area detecting method
CN106308795A (en) * 2016-08-31 2017-01-11 北京农业信息技术研究中心 Warning device for foot-and-mouth diseases of animals based on terahertz imaging technology and system and method thereof
CN106447659A (en) * 2016-09-27 2017-02-22 电子科技大学 Region growth detection method based on multiple judgments
CN106683076A (en) * 2016-11-24 2017-05-17 南京航空航天大学 Texture feature clustering-based locomotive wheelset tread damage detection method
CN106952315A (en) * 2017-03-22 2017-07-14 广东工业大学 A kind of method that image quick reconfiguration is carried out to Terahertz complex-valued data based on BFGS
CN107092040A (en) * 2017-06-01 2017-08-25 上海理工大学 Terahertz imaging rays safety detection apparatus and video procession method
CN107255641A (en) * 2017-06-06 2017-10-17 西安理工大学 A kind of method that Machine Vision Detection is carried out for GRIN Lens surface defect
CN107665486A (en) * 2017-09-30 2018-02-06 深圳绰曦互动科技有限公司 A kind of method for automatically split-jointing, device and terminal device applied to radioscopic image

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
张馨 等: ""被动式人体太赫兹成像系统的图像重构算法研究"", 《中国激光》 *
王林华 等: ""太赫兹安检系统人体图像边缘物体识别"", 《红外与激光工程》 *

Similar Documents

Publication Publication Date Title
Zhang et al. Adaptive and quality quadrilateral/hexahedral meshing from volumetric data
Saito et al. Comprehensible rendering of 3-D shapes
Graham et al. Automated sizing of coarse-grained sediments: image-processing procedures
Zuiderveld et al. Acceleration of ray-casting using 3-D distance transforms
US5166876A (en) System and method for detecting internal structures contained within the interior region of a solid object
Saha et al. Three‐dimensional digital topological characterization of cancellous bone architecture
US8040350B2 (en) Method and system for adaptive direct volume rendering
Wang et al. Volume sampled voxelization of geometric primitives
US7280121B2 (en) Image processing apparatus and method of same
EP1304651A2 (en) Background-based image segmentation
Provenzi et al. Random spray retinex: a new retinex implementation to investigate the local properties of the model
US6791540B1 (en) Image processing apparatus
US7680314B2 (en) Devices, systems, and methods for improving image consistency
EP1091320A2 (en) Processing multiple digital images
US7038697B2 (en) Color gradient paths
Russ Computer-assisted microscopy: the measurement and analysis of images
US7936922B2 (en) Method and apparatus for segmenting images
EP0424912A2 (en) Region extracting method and three-dimensional display method
US7801361B2 (en) Analyzing pixel data using image, thematic and object layers of a computer-implemented network structure
TWI362007B (en) Apparatus and methods for identifying regions within an image having similar continuity values, and blob identifier
US20080100621A1 (en) System and method for coronary segmentation and visualization
JP2005228342A (en) Method and system for segmenting scanned document
JPH07170406A (en) Adaptive digital picture signal filter
JP3684017B2 (en) Image processing apparatus and method
JP3759971B2 (en) How to shade a 3D image

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