CN101840003B - Green channel open vehicle radar detection method for metal contraband articles - Google Patents
Green channel open vehicle radar detection method for metal contraband articles Download PDFInfo
- Publication number
- CN101840003B CN101840003B CN2010101201245A CN201010120124A CN101840003B CN 101840003 B CN101840003 B CN 101840003B CN 2010101201245 A CN2010101201245 A CN 2010101201245A CN 201010120124 A CN201010120124 A CN 201010120124A CN 101840003 B CN101840003 B CN 101840003B
- Authority
- CN
- China
- Prior art keywords
- data
- row
- value
- submatrix
- echo
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention discloses a green channel open vehicle radar detection method for metal contraband articles. The method comprises the following steps of: recording radar echo data of an open vehicle; then removing coupled waves and intercepting a carriage area; finding out an abnormal sub-matrix position; removing the influence of the tail echo; and finally offering a placement range of the metal contraband articles. By referring to the placement range of the metal contraband articles, the detecting personnel can judge whether the green channel open vehicle carries the metal contraband articles and manually detect the corresponding position of the vehicle. The method in the invention has the advantages of high speed, time saving, labor saving, high detection accuracy and the like.
Description
Technical field
The invention belongs to radar detection and technical field of image processing, be specifically related to a kind of green channel open car radar detecting method to the metal contraband goods.
Background technology
In order to serve agriculture, rural areas and farmers; Further reduce the transportation cost of fresh and live agricultural product; Stablize the market price of agricultural product, strengthen support dynamics, Ministry of Communications's regulation agricultural production; To the legal loading fresh and live agricultural product of the car load haulage vehicle that on " green channel " of the whole nation " five vertical two horizontal strokes " highway, goes, exempt from toll.
When national policy was supported agriculture, rural areas and farmers, the promotion people's livelihood, part green channel vehicle was but secured personal gain national policy, means such as utilization loads in mixture, camouflage; Load a large amount of non-agricultural product at vehicle interior, only cover with a spot of fresh and live agricultural product the outside, escapes to pay freeway toll; Caused part high speed toll to run off; And this phenomenon is growing trend, has brought very big loss to country, has influenced the sound development of green channel policy.
In order to ensure that country's " green channel " policy effectively carries out, reduce the loss of state-owned assets, must sternly hit the illegal activities that highway toll is paid in frauds escapes such as utilization loads in mixture, camouflage.But freeway management department still lacks effective green channel vehicle detection means at present.
The green channel vehicle detection mainly relies on manual method inspection at present.A kind of method is to be got into by the staff to detect by an unaided eye in the compartment and manually move goods inspection.Concrete operations are; After waiting of passing through checks the green channel vehicle to stop; The removable staircase that the staff promotes to have wheel is to rear view of vehicle; Ascending stair gets into interior and checks, manually moves a part of goods on surface in case of necessity, and whether inspection exists the article that do not meet the current standard of green channel.This method inspection speed is slow, and staff's labour intensity is big.Big and goods is piled up when higher when carload, and staff's manual examination (check) can only be checked the goods top layer, is not easy to check the goods depths.And when vehicle was filled fully, the staff can't get into inspection in the compartment, and therefore the detection coverage rate of this method is low.Second method is to use drill rod to detect; After the vehicle to be detected that passes through stops; Whether the staff penetrates any part of a long drill rod that has spiral lamination from vehicle both sides or rear, extract the residue taken out of in the back inspection drill rod helicla flute out and be consistent with the characteristic of lade.This method subjectivity is big, inspection speed is slow, labour intensity is big and loss is very high, can play a role to jubilee wagon.But to container-type lorry large-scale, overlength, drill rod can't can only insert from the rear from both sides, and probing depth is far smaller than car length when inserting from the rear, and it is limited to the effect that the container-type lorry detects that drill rod is detected method.Drill rod is detected method and may be produced certain destruction to shipping goods in addition, when adopting the drill rod method of detecting to detect to the green channel vehicle that transports high price, fragile agricultural product, tends to cause being discontented with of shipping driver.Therefore, a kind of quick, efficient, lossless detection means are badly in need of in the green channel checkpoint.
Known another kind is considering that the detection method of implementing is a hydrometer method.This method uses the embedded platform scale in ground to come check weighing, when platform scale left by vehicle, records complete vehicle weight, and it is exactly the shared weight of vehicle loading goods that complete vehicle weight deducts the car body weight that obtains according to vehicle.Method with infrared acquisition can obtain the compartment volume then; Goods weight and compartment volume are divided by and obtain a rate of specific gravity; The standard specific gravity value of this value and vehicle loading cargo type is compared, see whether both conform to, if difference is excessive; The vehicle that then decidable detected possibly be mounted with different types of goods, needs emphasis to detect.But because check weighing is in the process of vehicle movement, dynamically to carry out in this method, what obtain is an instantaneous weight value, all inequality to the possibility of result that same car repeatedly measured, so the measurement accuracy of goods weight can not guarantee; The sparse degree that dissimilar in addition goods is piled up is also different, and when some cargo type was placed in the compartment, the space between individuality and the individuality was bigger, and the accumulation degree is sparse, like pumpkin; And other cargo types are when placing in the compartment, and the space between individuality and the individuality is less, and the accumulation degree is closely knit, and like green pepper, the standard specific gravity value of lade type is difficult to given.And on a car quite a few kinds of dissimilar green channel goods possibly are housed during the green channel vehicle transport in the reality, this brings difficulty also for the accurate measurement of hydrometer method.
The method that can realize the inner structure lossless detection mainly contains ultrasonic Detection Method, X-ray check method and nuclear detection method.Ultrasonic detection method mainly utilizes hyperacoustic reflection, refraction effect to survey the unevenness of interior of articles; Normally for surveying its discontinuous degree with a kind of medium even, compact structure; Be used for metal defect detection more; And fresh and live agricultural product out-of-shape, accumulation degree are sparse, so the ultrasonic listening method is not suitable for agricultural product and detects; X-ray check method and nuclear detection method resolution are high, imaging is directly perceived, but equipment operation is with high costs, and, poor stability strong to human body radiation, thereby also are difficult to be applied in the bigger highway of stream of people's traffic volume.
Summary of the invention
Open car detects to green channel to the objective of the invention is to utilize the electromagnetic wave of radar antenna emission; After the open car echo data handled; Judge whether open car is entrained with the metal contraband goods, if the placement scope of carrying secretly then providing the metal contraband goods that open car carries secretly is arranged.
The detections of radar method can realize the inner structure lossless detection, utilization be the variation difference characteristic of electromagnetic wave when in the medium of differing dielectric constant, propagating.Electromagnetic wave can see through agricultural product transmission such as trees, vegetables and melon and fruit; In the electromagnetic transmission process; Fresh and live agricultural product has served as electromagnetic transmission medium; When electromagnetic wave is propagated in medium, its path, electromagnetic intensity and waveform will with electrical property and geometric shape through medium change, this characteristic that detections of radar equipment of the present invention is propagated in medium according to electromagnetic wave is just come work.Detections of radar equipment can be according to electromagnetic variation difference; Reflect the interface between occurrence patterns, size and each dielectric of dielectric to be detected; In addition; The speed that electromagnetic wave is propagated in different medium there are differences, and different medium is also variant to absorption of electromagnetic wave and reflection potential.Different these characteristics of the image that utilizes different medium in detections of radar equipment, to appear can disclose the different rules that influence of filling medium to radar image in the investigative range.When electromagnetic wave was propagated in conducting medium, the penetration capacity of radar wave was limited.The high more decay of conductivity is big more, and reflection echo is just strong more.When metal object receives electromagnetic wave irradiation, will produce strong reflection, whether green passage open car on this basis is entrained with the metal contraband goods.
The inventive method comprises following step:
Step 1: to open car admission radar return data;
The hardware device of green passage open car is made up of antenna support frame, radar antenna, radar host computer and control computer.The antenna support frame is " door " font antenna holder; When open car is detected; With the radar antenna mouth down, be lifted on the centre position of antenna support frame, radar antenna links to each other with control computer with the radar host computer of placing on antenna support frame next door through cable.When open car passes through green channel vehicle detection point; Ordering about open car at the uniform velocity crosses from the antenna support frame below that radar antenna is housed with 3～8 kilometers/hour speed; Control computer is through radar host computer control radar antenna launching electromagnetic wave and receive the echo data of open car; After open car passed through the radar illumination zone fully, data recording was accomplished.
Step 2: uncoupling ripple and intercepting go out carriage area;
The radar return data of admission are arranged in the data matrix that the capable N of M is listed as by the sequencing of time, are shown as raw data image F, and wherein M is a sampling number that the reception echo comprises, and N is the total number that receives echo in the data recording time.Data among the image F have passed through delivery and normalization is handled, and span is 0～1.Because transmitter and receiver splits and works the coupling influence that the received echoed signal of receiver can be transmitted simultaneously in the radar antenna.Transmit and be coupled to the beginning part that the coupled waves that receives in the signal concentrates on the reception signal, invalid information, subsequent treatment should be removed for ease.Coupled waves is shown as the horizontal stripe belt-like zone of the top in raw data image F.The weak signal zone that is close to this below, zone is the aerial echo of radar wave, and the difference of the data in the data in this weak signal zone and coupled waves zone is greater than 0.6.Value through analyzing coupled waves zone and weak signal zone distributes; Roughly determine the estimated value H of the dead line position of coupled waves; Try to achieve the average I of all data that list in 1～H line range then; Wherein the value of H is relevant with radar parameter, can estimate to obtain through observing raw data image F, and average I is value in 0.6～1.Is that all list the average of data in the interval acquiring different rows interval since the 1st row with W; Just obtain the average of W * N data successively; Till average was less than I, this moment, this average was exactly the dead line position X of coupled waves in raw data image F less than the maximum row position that the capable interval of I comprises
1, wherein W rounds numerical value in 10～50.Active data partly is the carriage area echo data of open car during detection; In the admission data procedures when open car is not in the range of exposures at radar wave; Radar wave shines directly into ground after through air borne; Receive the strong reflection ripple that echo comprises ground, echo data does not at this moment comprise the effective information of open car.The ground echo data are greater than 0.8, and are distributed on the multirow, remove the echo data that comprises ground echo, at first will obtain ground echo line position D, and the value of D is by the distance decision to radar antenna of radar SF and ground.Each lists in D-100～D+100 line range the data number greater than 0.8 among the statistics raw data image F, and the columns series arrangement by corresponding obtains the sequence Y that length is N, and getting among the Y peaked 1/2～1/3 is the first thresholding T
1, the first thresholding T
1Round numerical value, as the number thresholding of search ground echo end column position and begin column position.The head end that lists the sequence Y of data number from correspondence the 1st begins caudad to do search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the end column position Y of ground echo
1, corresponding, the tail end that lists the sequence Y of data number from corresponding N begins head-end and does search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the begin column position Y of ground echo
2Echo line position X at the bottom of definite then compartment
2, end to the D-200 behavior from capable beginning of D, find out Y
1～Y
2Each lists first line position less than 0.3 data place in the row scope, if the Y that obtains
2-Y
1+ 1 line position difference each other is all less than 100, and the minimum value of getting in this N line position is X
2If, there is one in the difference greater than 100, then set X
2=D.The dead line position X that combines coupled waves again
1End column position Y with ground echo
1With begin column position Y
2, intercepting goes out carriage area image G from raw data image F.
Step 3: find out unusual submatrix position;
The metal contraband goods of carrying secretly in the open car can have the strong reflection effect to the electromagnetic wave of radar emission; Because the relative motion between open car and the radar antenna; Receive echo data and only include the strong reflection echo data at electromagnetic wave irradiation during to the metal contraband goods, reception echo data at this moment and electromagnetic wave have very big-difference when not shining the metal contraband goods.Therefore will analyze and handle carriage area image G, judge whether to have the unusual submatrix position by the echo data generation of metal contraband goods, if there is unusual submatrix position, but this open car of preliminary judgement be entrained with the metal contraband goods.When finding out unusual submatrix position; At first in carriage area image G, confirm the estimation line position L of goods surface echo; Obtain the data formation initial matrix E of all line positions,, divide submatrix successively from the 1st row and the 1st row beginning of initial matrix E greater than L; The size of submatrix is the capable V row of U, is the k of element with the submatrix what obtain then
CRow k
BIn the new matrix of row, a row division row are interval at interval, and a value is no more than k
B/ 10.If the second thresholding T
2Value in 0.5～0.7, the average thresholding during as the unusual submatrix of searching position.Obtain the maximum average of the interval interior submatrix of each row, wherein greater than the second thresholding T
2The position, the lower left corner of the corresponding submatrix of maximum average be exactly unusual submatrix position, so just obtained the distribution of preliminary unusual submatrix position.
Step 4: get rid of the influence of tailstock echo;
When open car rear portion lade is discontented; The tailstock metal baffle of goods surface can produce the strong reflection echo; Therefore the tailstock echo data of open car is obscured with metal contraband goods echo data sometimes mutually; Whether the unusual submatrix position that tentatively obtains above therefore needing to judge belongs to tailstock echo data, so that get rid of the influence of tailstock echo.The all corresponding data of each point among the carriage area image G; Then a plurality of adjacent datas constitute line segment; At this moment tailstock echo data is shown as many bold curve sections of extending to the upper right side from the image bottom in carriage area image G, extends to the position of goods surface echo always, successively each unusual submatrix position is handled; Judge whether to belong to tailstock echo data; If then reject this unusual submatrix position, obtained final unusual submatrix position at last, all belong to the influence of metal contraband goods echo data.When a unusual submatrix position was handled, the average of at first establishing the corresponding submatrix in this unusual submatrix position was m
p, set 0.8 * m
pBe the 3rd thresholding T
3, as the threshold value of conversion bianry image.Extracting then with this unusual submatrix position is the maximum rectangle data image O in the lower left corner, according to the 3rd thresholding T
3O transfers bianry image to image, utilizes the refinement computing in the mathematical morphology again, to obtaining refined image S behind the bianry image " refinement ".It is at first definite in refined image S that nearest value is 1 starting point q from this unusual submatrix position
1, note Q is the set of the point that searches, Q initially is made as empty set, earlier with starting point q
1Be included into Q, then to starting point q
1Carry out point search, the next value of search is 1 point in the R of hunting zone, and it is the rectangle region of search in the lower left corner with current starting point that hunting zone R is one, and counting of the length of region of search and wide correspondence rounds numerical value in 20～50 scope.If search next value and be 1 some q
2, need to judge q
2With among the Q whether have a few on a curve, basis for estimation is the coefficient of three rank fitting of a polynomials.If not on a curve with regard to cut-off point q
2, the next value of continuation search is 1 point, otherwise with q
2Be included into Q, and with q
2For new starting point continues in the R of the hunting zone next value of search is 1 point, by that analogy, is 1 point up to searching for less than next one value, then to starting point q
1Point search finish, just to unusual submatrix position (i
p, j
p) processing finish, last point that is included into is q in this moment Q
nBecause being the tailstock metal baffle of goods surface, tailstock echo data produces; Topmost the line position of data should be very approaching with the line position of goods surface echo data for tailstock echo; But because the estimation line position L of the goods surface echo that obtains is an estimated value; And consider the resolution of radar return and the accurate rate problem of point search, q is just put in the tailstock echo data top line position that search obtains
nThe estimation line position L of line position and goods surface echo between can have difference, need provide the 4th thresholding T
4Judge when the unusual submatrix position of pre-treatment and whether belong to tailstock echo data.According to engineering experience and concrete test operation, in 100～200 scope, get a round values as the 4th thresholding T
4, the 4th thresholding T
4Value more little, the false dismissal probability during actual detected is more little, but the probability of wrong report can increase and the 4th thresholding T
4Value when increasing, the probability of wrong report reduces, but the probability of omission can increase.If q
nThe absolute value of difference of line position and L less than T
4Explain that the unusual submatrix position of handling this moment belongs to tailstock echo data, directly reject this unusual submatrix position, otherwise keep this unusual submatrix position; After each unusual submatrix position that tentatively obtains handled, what remain at last was exactly final unusual submatrix position.
Step 5: the placement location that provides the metal object contraband goods;
According to the distribution of final unusual submatrix position, judge whether open car is entrained with the metal contraband goods, if carry secretly, obtain the roughly placement location of metal contraband goods in the compartment, and in control computer display result.
The invention has the advantages that:
(1) the present invention utilizes radar equipment to come the open car of green passage, through the open car radar return data that analysis receives, judges whether open car is entrained with the metal contraband goods;
(2) the inventive method has fast, the time saving and energy saving and detection high accuracy for examination of speed.
Description of drawings
Fig. 1 is the green channel open car radar detecting method process flow diagram that the present invention is directed to the metal contraband goods;
Fig. 2 is the green channel open car detections of radar device synoptic diagram that the present invention is directed to the metal contraband goods;
Fig. 3 is the raw data image of open car of the present invention;
Fig. 4 is the carriage area image that intercepting of the present invention goes out;
Fig. 5 is the unusual submatrix location drawing that the present invention tentatively obtains;
Fig. 6 is the corresponding refined image of carriage area image of the present invention;
The matched curve figure of Fig. 7 after to be the present invention to a unusual submatrix position that belongs to metal contraband goods echo data handle;
The matched curve figure of Fig. 8 after to be the present invention to a unusual submatrix position that belongs to tailstock echo data handle;
Fig. 9 is the final unusual submatrix location drawing of the present invention.
| Among the figure: | 1. |
2. |
3. |
| 4. |
5. open |
6. metal contraband goods |
Embodiment
Below in conjunction with accompanying drawing and embodiment the present invention is explained further details.
The present invention enrolls echo datas through radar equipment to the green channel open car 5 of running on expressway, provides the roughly placement scope of the metal contraband goods 6 of carrying secretly after data are handled.Provided the process flow diagram to the green channel open car radar detecting method of metal contraband goods among Fig. 1, processing procedure comprises following five steps:
Step 1: to open car admission radar return data;
The hardware device of green passage open car is made up of antenna support frame 1, radar antenna 2, radar host computer 3 and control computer 4.With radar antenna 2 mouth downs, be lifted on the centre position of the antenna support frame 1 of " door " font when open car 5 is detected, the radar host computer 3 that radar antenna 2 is placed through cable and next door links to each other with control computer 4.When open car 5 passes through green channel vehicle detection point; Ordering about open car 5 at the uniform velocity crosses from antenna support frame 1 below that radar antenna 2 is housed with 3～8 kilometers/hour speed; Control computer 4 is through radar host computer 3 control radar antennas, 2 launching electromagnetic waves and receive the echo data of open car 5; After open car 5 passed through the radar illumination zone fully, data recording was accomplished.
As shown in Figure 2, antenna support frame 1 is " door " font support, stands vertically on the ground.Radar antenna 2 mouth downs are lifted on the centre position of antenna support frame 1; Link to each other with the radar host computer 3 that place on the next door from the radar antenna 2 controls mouthful cable that even goes out; Communicate through USB (Universal Serial Bus abbreviates USB as) interface between radar host computer 3 and the control computer 4.As green channel open car 5 when radar antenna 2 belows are passed through; Control computer 4 sending controling instructions; Make radar antenna 2 launching electromagnetic wave and receive echo data under the parameters of corresponding open car 5, open car 5 is fully through obtaining the radar return data of vehicle various piece after the range of exposures of radar antenna 2.
Step 2: uncoupling ripple and intercepting go out carriage area;
After echo data has been enrolled in 2 pairs of open cars 5 that pass through from the below of radar antenna; Just form a radar return data matrix; Processing is afterwards all carried out around this radar return data matrix; All as a pixel, then the radar return data matrix constitutes a width of cloth raw data image, is designated as F with each data in the radar return data matrix.
The raw data image F of open car 5 that has typically carried metal contraband goods 6 secretly is as shown in Figure 3.The data recording time T that this raw data image is corresponding
RBe 12.4s, pulse repetition rate PRF is 128, time of reception window length T
WBe 68.3ns, SF F
SBe 30GHz, therefore enrolled 1587 row echo pulse signals altogether, the sampling number of every row echo pulse signal is 2049.As can be seen from the figure horizontal ordinate is represented 1～1587 columns scope, and ordinate is represented 1～2049 line number scope.Data in the image are value in 0～1 scope, and the scale strip on image the right has indicated the gray scale of different value representatives.Demonstrated the radar return data of open car 5 among Fig. 3, can see that zone at the bottom of first half in the compartment is near the compartment has demonstrated the echo data greater than 0.8 metal contraband goods 6 from headstock to tailstock various piece.
Because transmitter and receiver splits and works the coupling influence that the received echoed signal of receiver can be transmitted simultaneously in the radar antenna 2.Transmit and be coupled to the beginning part that the coupled waves that receives in the signal concentrates on the reception signal, belong to invalid information.As can be seen from Figure 3; Coupled waves is shown as the horizontal stripe belt-like zone of the top in raw data image F, each band all occupies the N row, is distributed on the several rows; And the data on the band are all greater than 0.8; Have 3 bands among Fig. 3, occupy 1～10 row between each coupled waves band in addition, and data are less than 0.3 be interrupted.The weak signal zone that is close to this horizontal stripe belt-like zone below is the aerial echo data of launching electromagnetic wave, and little of the value of the data that the value of the data in this weak signal zone is more regional than coupled waves are generally less than 0.3 more.When radar antenna 2 parameters fixedly the time, the dead line position X of the coupled waves of reception in raw data image F
1Constant, in the reality because radar parameter is different, coupled waves dead line position X
1Value indefinite.Coupled waves dead line position X can be set earlier
1Estimated value H.Because H is an estimated value, do not require accurately, can obtain through observing raw data image F.Observe last band in the coupled waves zone behind the enlarged image F; The rough span of confirming the dead line position of this band; The value of H is greater than the maximal value that this span comprises, but can not be greater than the minimum row position that the vehicle echo data occurs, as can be seen from Figure 3 data greater than the dead line position of last band in 0.8 coupled waves zone between 50～150; And the line position that the vehicle echo data occurs is greater than 200, and then H can be made as 200.Try to achieve the average I of all the H * N that lists data in 1～H line range then.In order to obtain more accurate coupled waves dead line position than H, be the line range of dividing 1～H at interval with W, suppose can be divided into k at most
WIndividual row is interval, then (k
W+ 1) * W＞H.Because each coupled waves bar interband also has data value less than 0.3, occupy being interrupted of 1～10 row, so W rounds numerical value in 10～50, and can not be greater than H/3.Obtain the interval 1～W of row successively, W+1～2 * W, 2 * W+1～3 * W ..., (k
W-1) * W+1～k
WThe average of all data that list the in * W is the average that the interval obtains W * N data successively with W since the 1st row just, and establishing first average is k less than the capable interval of I
iIndividual row is interval, then the maximal value k that comprises of this row interval
i* W is exactly coupled waves dead line position, is designated as X
1, X
1Get the round values in 1～M.Concerning raw data image shown in Figure 3, the average I that tries to achieve is 0.4322, and the value of W is made as 10, and it is interval to have divided 20 row, and first average is that the 10th row is interval less than the capable interval of I, then X
1Value be 100.
Active data partly is the carriage area echo data of open car 5 during processing; In the admission data procedures when open car 5 is not in the range of exposures at radar wave; The electromagnetic wave of radar antenna 2 emission shines ground after through air borne, receives the direct strong reflection ripple that echo has comprised ground, is shown as between Fig. 3 two ends ordinate 900～1100 row greater than the strong echo data in 0.8 ground; At this moment echo data does not include effective information, need remove.After radar antenna 2 stationkeeping, ground and radar antenna 2 distances are constant, and the position of ground echo in Fig. 3 is also constant, can confirm ground echo line position D according to actual conditions, and the value of D is 2 * R
D* F
S/ C removes the round values after the fraction part, wherein R
DBe the vertical range of radar antenna 2 to ground, F
SBe SF, C is the light velocity.For raw data image shown in Figure 3, R
DBe 5 meters, then ground echo line position D is 1000.As can be seen from Figure 3, when open car 5 passed through radar antenna 2 belows, the ground echo data have at the image middle part was interrupted, and what discontinuities showed is exactly the echo data of open car 5, and the column position at discontinuities two ends is respectively ground echo end column position Y
1With begin column position Y
2Each lists the data number greater than 0.8 in D-100～D+100 line range among the statistics raw data image F, and by corresponding columns series arrangement, having obtained length is the sequence Y of N, and getting among the Y peaked 1/2～1/3 is the first thresholding T
1, the first thresholding T
1Round numerical value, as the number thresholding of end of search ground echo and begin column position.The head end that lists the sequence Y of data number from correspondence the 1st begins caudad to do search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the end column position Y of ground echo
1, corresponding, the tail end that lists the sequence Y of data number from corresponding N begins head-end and does search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the begin column position Y of ground echo
2, Y
1And Y
2All get the round values in 1～N, and Y
1≤Y
2The Y that raw data image among Fig. 3 is corresponding
1Be 670 row, Y
2Be 1325 row.
Also need confirm the echo line position X at the bottom of the compartment at last
2, in raw data image F, be benchmark with ground echo line position D, end to the D-200 behavior from capable beginning of D, at Y
1～Y
2Find out in every row first line position in the row scope, obtained Y less than 0.3 data
2-Y
1+ 1 line position is if these line positions difference is each other then got this Y all less than 100
2-Y
1Minimum value in+1 line position is the echo line position X at the bottom of the compartment
2But sometimes because the propagation attenuation of echo in goods is strong, and the echo data at the bottom of the compartment is less than 0.3, show not obvious, the Y that then obtains
2-Y
1One of difference between+1 line position maybe be greater than 100, at this moment with the echo line position X of ground echo line position D at the bottom of as the compartment
2Carry out subsequent treatment.This moment in raw data image F intercepting X
1～X
2OK, Y
1～Y
2All data between the row, the image G that obtains is effective compartment area image of open car 5.Capable B is listed as if the matrix that carriage area image G is corresponding is A, wherein A=X
2-X
1+ 1, B=Y
2-Y
1+ 1.Echo data among Fig. 3 at the bottom of the compartment shows and is not obvious, then establishes X
2=D=1000, in conjunction with before the Y that obtains
1, Y
2And X
1Value, intercepting 100～1000 row, the carriage area image G that all data between 670～1325 row constitute provides in Fig. 4, corresponding A value is 901, the B value is 656.
Step 3: find out unusual submatrix position;
The metal contraband goods 6 of carrying secretly in the open car 5 can have the strong reflection effect to the electromagnetic wave of radar emission; Because the relative motion between open car 5 and the radar antenna 2; Only include the strong reflection echo data during to metal contraband goods 6 in the echo data that radar antenna 2 receives at electromagnetic wave irradiation; Echo data when not shining metal contraband goods 6 with electromagnetic wave has very big-difference, in carriage area image G, is shown as strong value zone.Therefore carriage area image G is analyzed and handles, judge whether the unusual submatrix position that exists metal contraband goods echo data to produce.
For the strong value zone of finding metal contraband goods echo data in carriage area image G, to show; At first should confirm the estimation line position L of goods surface echo; Each row among the carriage area image G all find first greater than 0.8 data since the 1st row; Write down the line position at this data place, obtain each at last and list the minimum row position greater than 0.8 data, maximal value wherein is B
MWith 1～B
MLine range to be divided into size be that several row of 100 are interval, if B
MCan be divided exactly by 100, then be divided into B
M/ 100 row are interval, if aliquant, B just
M=b
k* 100+b
r, b
r＜100, b wherein
kAnd b
rAll round numerical value, then still be divided into b
kIndividual row is interval, but preceding b
k-1 interval size of row all is 100, and the interval size of last row is 100+b
rThe statistics front obtain each list the distribution number of minimum row position in each row is interval greater than 0.8 data, the central value in the interval of distribution most number is made round after, what obtain is exactly the estimation line position of goods surface echo, is designated as L.On the metal contraband goods 6 of carrying secretly in the open car 5 generally is placed at the bottom of the compartment, away from the goods surface.When therefore seeking unusual submatrix position, can in carriage area image G, constitute the initial matrix E of the capable B row of C, wherein C=A-L greater than all data of L by the intercepting line position.Since the 1st row and the 1st row, initial matrix E is divided into the submatrix that several sizes are listed as for the capable V of U, submatrix is big or small identical and be arranged in order; Row between adjacent submatrix with list allow a part overlapping; Wherein U and V round numerical value in 10～30, but all can not surpass min (C, B)/10.When supposing to divide submatrix, between adjacent submatrix on line direction overlapping O
UOK, overlapping O on the column direction
VRow, wherein 0≤O
U＜U, 0≤O
V＜V, O
UAnd O
VThe accuracy rate as a result of the more little then subsequent processing steps of value high more, but calculated amount can increase.Initial matrix E is divided into k
C* k
BIndividual submatrix, k
CValue is (C-U)/Q
URemove the round values after the fraction part, k
BValue is (B-U)/Q
VRemove the round values after the fraction part, constituted k
CRow k
BThe new matrix of row, new entry of a matrix element is a submatrix.It is interval to divide row at new matrix interval a row then, and obtaining being listed as interval number is k
B/ a removes the round values after the fraction part, and all there is k in each row interval
C* a submatrix, wherein a value is no more than k
B/ 10, the value of a is more little, and the unusual submatrix position of confirming is accurate more, but calculated amount can increase.If the second thresholding T
2Value in 0.5～0.7, the average thresholding during as the unusual submatrix of searching position.Obtain the maximum average of the interval interior submatrix of each row successively, if maximum average is greater than the second thresholding T
2, to set the corresponding submatrix of this maximum average and exist unusually, column position and the line position of data in G of demarcating this submatrix lower left corner is unusual submatrix position, so just tentatively obtained the distribution of unusual submatrix position.Carriage area image to Fig. 4; Fig. 5 has provided the unusual submatrix location drawing that tentatively obtains; The horizontal ordinate of each point and ordinate have been represented column position and the line position of unusual submatrix position in carriage area image G respectively among the figure; Can find out that the unusual submatrix position that produces except metal contraband goods echo data, the unusual submatrix position that also has tailstock echo data to produce should be rejected.A unusual submatrix position (796,81) of metal contraband goods echo data generation and the unusual submatrix position (797,410) that tailstock echo data produces have been indicated with small arrow respectively among the figure.
Step 4: get rid of the influence of tailstock echo;
Because open car 5 is at the uniform velocity to pass through radar antenna 2 belows, the echo data of the headstock and the tailstock is shown as two long cross curves in carriage area image G, and wherein the influence of the tailstock is especially obvious.As can be seen from Figure 4, the tailstock echo data of open car 5 is obscured with metal contraband goods echo data greater than 0.8 easily mutually.Therefore need handle each unusual submatrix position that tentatively obtains, judge whether to belong to tailstock echo data, so that get rid of the influence of tailstock echo.The all corresponding data of each point among the carriage area image G; Then a plurality of adjacent datas constitute line segment; At this moment tailstock echo data is shown as many bold curve sections in carriage area image G, handles each unusual submatrix position successively, to a unusual submatrix position (i
p, j
p) when handling, concrete steps are: establish unusual submatrix position (i
p, j
p) average of corresponding submatrix p is m
p, i wherein
pBe the line position of data in carriage area image G in the submatrix p lower left corner, j
pBe the column position of data in carriage area image G in the submatrix p lower left corner.At first extract among the image G all line positions smaller or equal to i
pAnd column position is more than or equal to j
pThe data area formed of point; Being the maximum rectangle data image O in the lower left corner with unusual submatrix position among the carriage area image G just, is convenient and image G comparison, and the line position of data and column position are not reset among the image O; With identical in G, that is to say the data (i among the G
p-i
k, j
p+ j
k) in O, still be designated as data (i
p-i
k, j
p+ j
k), this operation does not influence subsequent step.Set the 3rd thresholding T
3Be 0.6 * m
p, as the threshold value of conversion bianry image.According to the 3rd thresholding T
3O transfers bianry image to image, will be more than or equal to the 3rd thresholding T
3Data become 1, will be less than the 3rd thresholding T
3Data become 0.Utilize the refinement computing of deriving by switch change-over in the mathematical morphology then, this bianry image is carried out " refinement ", obtain refined image S at last.Provided the corresponding refined image of carriage area image G among Fig. 6, comparison diagram 4 can find out that " refinement " operation has been refined into thin segment of curve with the bold curve section of tailstock echo data, is convenient to next step point search.
Then will be in refined image S, from unusual submatrix position (i
p, j
p) beginning finds out the position and be (i in unusual submatrix p
1, j
1) starting point q
1, the value of this point is 1, and is 1 point with respect to other value among the submatrix p, | i
1-i
p|+| j
1-j
p| value minimum.Then to starting point q
1Carry out point search, it is the rectangle region of search in the lower left corner with current starting point that hunting zone R is one, and counting of the length of region of search and wide correspondence rounds numerical value in 20～50 scope.
To starting point q
1Carry out point search, establish earlier record searching to the set of point be Q, Q is initially empty set, at first with starting point q
1Be included into Q, the next value of search is 1 point in the R of hunting zone then, and the point search that carries out in the R of hunting zone is in proper order: at first press each row of columns sequential search from small to large, each is gone in each row, to press line number sequential search from big to small.If search next value and be 1 some q
2, need to judge q
2With among the Q whether have a few on a curve, basis for estimation is the coefficient of three rank fitting of a polynomials.Respectively pair set Q with comprise q
2Set { Q, q
2Carry out three rank fitting of a polynomials, obtain two groups of fitting coefficients, every group has four fitting coefficients respectively, asks absolute value after corresponding fitting coefficient subtracts each other, with four absolute value additions that obtain, and if greater than 0.1 with regard to cut-off point q
2, the next value of continuation search is 1 point, otherwise with q
2Be included into Q, and with q
2For new starting point continues in the R of the hunting zone next value of search is 1 point, by that analogy, and till search is at 1 less than next one value, then at this moment to unusual submatrix position (i
p, j
p) processing finish, write down the some q that last is included among the Q at last
nThe position be (i
n, j
n).
Because being the tailstock metal baffle of goods surface, tailstock echo produces; The minimum row position of tailstock echo data should be very approaching with the line position of goods surface echo data; But because the estimation line position L of the goods surface echo that obtains is an estimated value; And consider the resolution of radar return and the accuracy problem of point search, the minimum row position of the tailstock echo data that search obtains, just q
nThe estimation line position L of line position and goods surface echo between can have difference, need provide the 4th thresholding T
4Judge when the unusual submatrix position of pre-treatment and whether belong to tailstock echo data.According to engineering experience and concrete test operation, in 100～200 scopes, get a round values as the 4th thresholding T
4, the 4th thresholding T
4Value more little, the false dismissal probability during actual detected is more little, but the probability of wrong report can increase simultaneously, and the 4th thresholding T
4Value when increasing, the probability of wrong report reduces, but the probability of omission can increase.If | L-i
n|＜T
4, the unusual submatrix position (i of this moment is described
p, j
p) belonging to tailstock echo data, should directly reject, be final unusual submatrix position otherwise keep this unusual submatrix position.
To the carriage area image among Fig. 4, the 4th thresholding T
4Be made as 170, the estimation line position L of the goods surface echo that in step 3, obtains is 285.The unusual submatrix position (796 that belongs to metal contraband goods echo data that left side small arrow among Fig. 5 is indicated; 81) handle after; Observe for convenient, the institute in the Q is carried out three rank fitting of a polynomials a little, the matched curve that obtains is presented among Fig. 7; The minimum row position that can find out the matched curve mid point is 723, | L-723|＞T
4, then keeping this unusual submatrix position is final unusual submatrix position.To the unusual submatrix position (797 that belongs to tailstock echo data that indicates of small arrow, the right among Fig. 5; 410) handle after; Observe for convenient, equally the institute in the Q is carried out three rank fitting of a polynomials a little, the matched curve that obtains is presented among Fig. 8; The minimum row position of matched curve mid point is 401, | L-401|＜T
4, then reject this unusual submatrix position.
Handle unusual submatrix position through all are tentatively obtained, and after the influence of eliminating tailstock echo data, it is as shown in Figure 9 to obtain final unusual submatrix position.
Step 5: the placement location that provides the metal object contraband goods;
According to the distribution of final unusual submatrix position, judge whether open car 5 is entrained with metal contraband goods 6, if carry secretly, obtain the roughly placement location of metal contraband goods 6 in open car 5, and in control computer 4 display result.
To final unusual submatrix position, the line position through adding up each unusual submatrix position and the distribution of column position have obtained the maximal value i of line position
MaxWith minimum value i
Min, the maximal value j of column position
MaxWith minimum value j
Min, | A-(i
Max+ i
Min)/2|/A is the distance of center at the bottom of the compartment and the ratio of height of wagon of metal contraband goods 6, | (j
Min+ j
Max)/2|/B is the ratio of distance and car length of center pronotum to the compartment of metal contraband goods 6; Detect for ease at last; What provide is the roughly placement location of metal contraband goods 6 in the compartment, and two ratio results of front are shown as denominator in control computer 4 be 10 fractional form.To the carriage area image among Fig. 4, the center of metal contraband goods 6 is at the bottom of the compartment 0.89/10, from compartment pronotum 4.8/10 place.If to not obtaining final unusual submatrix position after the carriage area Flame Image Process, then in control computer 4, show " the metal contraband goods is not carried in open car secretly ".
When green channel open car 5 is detected; Through after the processing of above five steps; Can judge whether open car 5 is entrained with metal contraband goods 6,, can do further hand inspection to open car 5 according to this result if the placement location of carrying secretly then providing metal contraband goods 6 is arranged.
The present invention uses radar equipment to come green passage open car 5; Through analyzing and handle the radar return data; Judge whether open car 5 is entrained with metal contraband goods 6, can alleviate green channel staff's workload greatly, accelerate detected vehicular velocity; Improve to detect accuracy rate, for preventing to escape to pay the behavior of freeway toll and play positive role through carrying non-agricultural product secretly.
Claims (8)
1. the green channel open car radar detecting method to the metal contraband goods is characterized in that this method comprises the steps:
Step 1: to open car admission radar return data;
The hardware device of green passage open car is made up of antenna support frame (1), radar antenna (2), radar host computer (3) and control computer (4);
When open car (5) is detected with radar antenna (2) mouth down; Be lifted on the centre position of the antenna support frame (1) of " door " font, radar antenna (2) links to each other with control computer (4) with the radar host computer of placing on antenna support frame (1) next door (3) through cable;
When open car (5) passes through green channel vehicle detection point; Ordering about open car (5) crosses from antenna support frame (1) below that radar antenna (2) is housed with 3～8 kilometers/hour speed; Control computer (4) is through radar host computer (3) control radar antenna (2) launching electromagnetic wave and receive the echo data of open car (5); After open car (5) passed through the radar illumination zone fully, data recording was accomplished;
Step 2: uncoupling ripple and intercepting go out carriage area;
After radar antenna (2) has been enrolled echo data to the open car (5) that passes through from its below; Just form a radar return data matrix; All as a pixel, then the radar return data matrix constitutes a width of cloth raw data image, is designated as F with each data in this radar return data matrix;
Value through analyzing raw data image F the top horizontal stripe belt-like zone distributes, and confirms the distribution range of coupled waves, obtains the dead line position X of coupled waves in raw data image F
1
Value through analyzing the strong echo in ground distributes, and finds out the end column position Y of open car (5) through ground echo before and after the radar antenna (2)
1With begin column position Y
2, the echo line position X at the bottom of definite then compartment
2
Intercepting X in raw data image F at last
1～X
2OK, Y
1～Y
2All data between the row, the carriage area image G that obtains is effective compartment area image of open car (5);
Capable B is listed as if the matrix that carriage area image G is corresponding is A, wherein A=X
2-X
1+ 1, B=Y
2-Y
1+ 1;
Step 3: find out unusual submatrix position;
At first in carriage area image G, all find first greater than 0.8 data since the 1st row to each row, write down the line position at this data place, obtained each at last and listed the minimum row position greater than 0.8 data, maximal value wherein is B
M
Then in carriage area image G with 1～B
MLine range to be divided into size be 100 b
kIndividual row is interval: if B
MCan be divided exactly by 100, then divide b
k=B
M/ 100 row are interval; If aliquant, B just
M=b
k* 100+b
r, b
r＜100, b wherein
kAnd b
rAll round numerical value, then still be divided into b
kIndividual row is interval, but preceding b
kThe size in-1 interval all is 100, and last interval size is 100+b
rThe statistics front obtain each list the distribution number of minimum row position in each row is interval greater than 0.8 data, behind the central value round with the interval of distribution most number, what obtain is exactly the estimation line position of goods surface echo, is designated as L;
When seeking unusual submatrix position, at first the intercepting line position constitutes the initial matrix E of the capable B row of C, wherein C=A-L greater than all data of L in carriage area image G; Since the 1st row and the 1st row, initial matrix E is divided into the submatrix that several sizes are listed as for the capable V of U, submatrix is big or small identical and be arranged in order; Row between adjacent submatrix with list allow a part overlapping; Wherein U and V round numerical value in 10～30, but all can not surpass min (C, B)/10; When supposing to divide submatrix, between the adjacent submatrix on line direction overlapping O
UOK, overlapping O on the column direction
VRow, wherein 0≤O
U＜U, 0≤O
V＜V, O
UAnd O
VThe accuracy rate as a result of the more little then subsequent processing steps of value high more, but calculated amount increases; Initial matrix E is divided into k
C* k
BIndividual submatrix, k
CBe (C-U)/O
URemove the round values after the fraction part, k
BBe (B-U)/O
VRemove the round values after the fraction part, constituted k
CRow k
BThe new matrix of row, new entry of a matrix element is a submatrix;
It is interval to divide row at new matrix interval a row then, and obtaining being listed as interval number is k
B/ a removes the round values after the fraction part, and all there is k in each row interval
C* a submatrix, wherein a value is no more than k
B/ 10, the value of a is more little, and the unusual submatrix position of confirming is accurate more, but calculated amount increases; If the second thresholding T
2Value in 0.5～0.7, the average thresholding during as the unusual submatrix of searching position; Obtain the maximum average of the interval interior submatrix of each row successively, if maximum average is greater than the second thresholding T
2, to set the corresponding submatrix of this maximum average and exist unusually, column position and the line position of data in G of demarcating this submatrix lower left corner is unusual submatrix position;
Step 4: get rid of the influence of tailstock echo;
All need whether belong to the judgement of tailstock echo data to each unusual submatrix position that tentatively obtains, to a unusual submatrix position (i
p, j
p), the concrete steps of processing are:
If unusual submatrix position (i
p, j
p) average of corresponding submatrix p is m
p, i wherein
pBe the line position of data in carriage area image G in the submatrix p lower left corner, j
pBe the column position of data in carriage area image G in the submatrix p lower left corner;
All corresponding data of each point among the carriage area image G are at first extracted among the carriage area image G all line positions smaller or equal to i
pAnd column position is more than or equal to j
pThe data area formed of point, be the maximum rectangle data area O in the lower left corner with unusual submatrix position among the carriage area image G just; Set the 3rd thresholding T
3Be 0.6 * m
p, and with the 3rd thresholding T
3Threshold value as the conversion bianry image; According to the 3rd thresholding T
3O transfers bianry image to image, will be more than or equal to the 3rd thresholding T
3Data become 1, will be less than the 3rd thresholding T
3Data become 0; Utilize the refinement computing of deriving by switch change-over in the mathematical morphology then, this bianry image is carried out " refinement ", obtain refined image S at last;
Then in refined image S, from unusual submatrix position (i
p, j
p) beginning finds out the position for (i in submatrix p
1, j
1) starting point q
1, the value of this point is 1, and is 1 point with respect to other value among the submatrix p, | i
1-i
p|+| j
1-j
p| value minimum;
To starting point q
1Carry out point search, establish earlier record searching to the set of point be Q, Q is initially empty set, at first with starting point q
1Be included into Q, the next value of search is 1 point in the R of hunting zone then;
If search next value and be 1 some q
2, need to judge q
2With among the Q whether have a few on a curve, basis for estimation is the coefficient of three rank fitting of a polynomials, respectively pair set Q with comprise q
2Set { Q, q
2Carry out three rank fitting of a polynomials, obtain two groups of fitting coefficients, every group has four fitting coefficients respectively, asks absolute value after corresponding fitting coefficient subtracts each other, with four absolute value additions that obtain, and if greater than 0.1 with regard to cut-off point q
2, the next value of continuation search is 1 point, otherwise with q
2Be included into Q, and with q
2For new starting point continues in the R of the hunting zone next value of search is 1 point; By that analogy, till search was at 1 less than next one value, then this moment was to unusual submatrix position (i
p, j
p) processing finish, write down the some q that last is included among the Q at last
nThe position be (i
n, j
n);
Use the 4th thresholding T
4Judge when the unusual submatrix position of pre-treatment and whether belong to tailstock echo data; According to engineering experience and concrete test operation, in 100～200 scope, get a round values as the 4th thresholding T
4, the 4th thresholding T
4Value more little, the false dismissal probability during actual detected is more little, but the probability of wrong report increases simultaneously, and the 4th thresholding T
4Value when increasing, the probability of wrong report reduces, but false dismissal probability increases;
If | L-i
n|＜T
4, explaining that the unusual submatrix position of this moment belongs to tailstock echo data, should directly reject; Otherwise keep this unusual submatrix position is final unusual submatrix position;
Step 5: the placement location that provides the metal object contraband goods;
To final unusual submatrix position, the line position through adding up each unusual submatrix position and the distribution of column position have obtained the maximal value i of line position
MaxWith minimum value i
Min, the maximal value j of column position
MaxWith minimum value j
Min, | A-(i
Max+ i
Min)/2|/A is the distance of center at the bottom of the compartment and the ratio of height of wagon of metal contraband goods (6), | (j
Min+ j
Max)/2|/B is the ratio of distance and car length of center pronotum to the compartment of metal contraband goods (6); Detect for ease at last, what provide is the roughly placement location of metal contraband goods (6) in the compartment, and two ratio results of front are shown as denominator in control computer (4) be 10 fractional form;
If to not obtaining final unusual submatrix position after the carriage area image G processing, then in control computer (4), show " the metal contraband goods is not carried in open car secretly ".
2. based on the described a kind of green channel open car radar detecting method of claim 1 to the metal contraband; It is characterized in that the radar return data matrix described in the step 2 is to do the row alignment arrangement by a plurality of echo pulse signals that form a line by the sequencing of time of reception to form;
Data after the sampling are plural numbers, and all the data elder generation deliverys in the radar return data matrix adopt linear function conversion carrying out normalization to handle then, and the span of data is 0～1 after the normalization;
If received N row echo pulse signal altogether during to open car (5) admission data, the sampling number of every row echo pulse signal is M, then obtains the radar return data matrix of the capable N row of M; The data of the capable j row of the i of this radar return data matrix are designated as data, and (i, j), corresponding is i sampled value of j echo pulse signal, wherein N and actual data recording time T
RPRF is relevant with pulse repetition rate, and value is T
R* PRF removes the round values after the fraction part, M and time of reception window length T
WWith SF F
SRelevant, value is T
W* F
SRemove the round values after the fraction part, the i value is got the round values in 1～M scope, and j value is got the interior round values of 1～N scope;
In the radar return data matrix, line number 1～M arranges from top to bottom, and columns 1～N arranges from left to right; The value of the data of each particular locations all is the sampled value of echo pulse signal, and the position that is to say data (1,1) is the 1st row the 1st row; Be positioned at the upper left corner of data matrix; The value of these data is the 1st sampled values of the 1st echo pulse signal, and data (M, position N) is the capable N row of M; Be positioned at the lower right corner of data matrix, the value of these data is M sampled values of N echo pulse signal.
3. according to a kind of green channel open car radar detecting method described in the claim 1, it is characterized in that the distribution range of the coupled waves described in the step 2 concentrates on the beginning part that receives signal, belongs to invalid information to the metal contraband goods; Coupled waves is shown as the horizontal stripe belt-like zone of the top in raw data image F, each band all occupies the N row, and the data on each band are all greater than 0.8.
4. according to a kind of green channel open car radar detecting method described in the claim 1, it is characterized in that the dead line position X of the coupled waves described in the step 2 in raw data image F to the metal contraband goods
1Definite method following:
When radar antenna (2) parameter fixedly the time, the dead line position X of the coupled waves of reception in raw data image F
1Constant; Coupled waves dead line position X at first is set
1Estimated value H; H is an estimated value; Obtain through observing raw data image F: observe last band in the coupled waves zone after amplifying raw data image F; The rough span of confirming the dead line position of this band, the maximal value that the value of H comprises greater than this span, but be not more than the minimum row position that the vehicle echo data occurs;
Try to achieve the average I of all the H * N that lists data in 1～H line range then; In order to obtain more accurate coupled waves dead line position X than H
1, be the line range of dividing 1～H at interval with W, suppose to be divided at most k
WIndividual row is interval, then (k
W+ 1) * W＞H; W rounds numerical value in 10～50, and is not more than H/3; Obtain the interval 1～W of row successively, W+1～2 * W, 2 * W+1～3 * W ..., (k
W-1) * W+1～k
WThe average of all data that list the in * W is the average that the interval obtains W * N data successively with W since first row, and establishing first average is k less than the capable interval of I
iIndividual row is interval, then the maximal value k that comprises of this row interval
i* W is exactly coupled waves dead line position X
1, X
1Get the round values in 1～M.
5. according to a kind of green channel open car radar detecting method described in the claim 1, it is characterized in that the end column position Y of the ground echo described in the step 2 to the metal contraband goods
1With begin column position Y
2Definite method following:
After radar antenna (2) stationkeeping, ground and radar antenna (2) distance is constant, and the position of ground echo data in raw data image F is also constant, confirms ground echo line position D according to actual conditions, and the value of D is 2 * R
D* F
S/ C removes the round values after the fraction part, wherein R
DBe the vertical range of radar antenna (2) to ground, F
SBe SF, C is the light velocity;
When open car (5) passed through radar antenna (2) below, the ground echo data have at the data image middle part was interrupted, and what discontinuities showed is exactly the echo data of open car (5), and the column position at discontinuities two ends is respectively ground echo end column position Y
1With begin column position Y
2Among the statistics raw data image F in D-100～D+100 line range in each row greater than the number of 0.8 data, the columns series arrangement by corresponding obtains the sequence Y that length is N, getting among the Y peaked 1/2～1/3 is the first thresholding T
1, the first thresholding T
1Round numerical value, as the number thresholding of end of search ground echo and begin column position; The head end that lists the sequence Y of data number from correspondence the 1st begins caudad to do search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the end column position Y of ground echo
1, corresponding, the tail end that lists the sequence Y of data number from corresponding N begins head-end and does search, and first that obtains is less than the first thresholding T
1The corresponding column position of sequence member be the begin column position Y of ground echo
2, Y
1And Y
2All get the round values in 1～N, and Y
1≤Y
2
6. according to a kind of green channel open car radar detecting method described in the claim 1, it is characterized in that the echo line position X at the bottom of the compartment described in the step 2 to the metal contraband goods
2Definite method following:
In raw data image F, be benchmark with ground echo line position D, end to the D-200 behavior from capable beginning of D, at Y
1～Y
2Find out in every row first line position in the row scope, obtain Y less than 0.3 data
2-Y
1+ 1 line position;
If these line positions difference each other, is then got minimum value in these line positions echo line position X at the bottom of as the compartment all less than 100
2
If one of these line positions difference each other is greater than 100, at this moment directly with the echo line position X of ground echo line position D at the bottom of as the compartment
2
7. according to a kind of green channel open car radar detecting method described in the claim 1 to the metal contraband goods; It is characterized in that; It is the rectangle region of search in the lower left corner with current starting point that hunting zone R described in the step 4 is one, and counting of the length of region of search and wide correspondence rounds numerical value in 20～50 scope.
8. according to a kind of green channel open car radar detecting method described in the claim 1 to the metal contraband goods; It is characterized in that; The order of in the R of hunting zone, carrying out point search described in the step 4 is: at first press each row of columns sequential search from small to large, each is gone by line number sequential search from big to small in each row.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101201245A CN101840003B (en) | 2010-03-05 | 2010-03-05 | Green channel open vehicle radar detection method for metal contraband articles |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101201245A CN101840003B (en) | 2010-03-05 | 2010-03-05 | Green channel open vehicle radar detection method for metal contraband articles |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101840003A CN101840003A (en) | 2010-09-22 |
| CN101840003B true CN101840003B (en) | 2012-07-25 |
Family
ID=42743502
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010101201245A Expired - Fee Related CN101840003B (en) | 2010-03-05 | 2010-03-05 | Green channel open vehicle radar detection method for metal contraband articles |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101840003B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9746553B2 (en) * | 2012-12-19 | 2017-08-29 | Sony Corporation | Method for generating an image and handheld screening device |
| CN109765548B (en) * | 2017-11-10 | 2023-09-08 | 富士通株式会社 | Article detection method and device |
| CN109765549B (en) * | 2017-11-10 | 2022-09-27 | 富士通株式会社 | Data cutting method and device and article detection method and device |
| CN109613108B (en) * | 2018-12-27 | 2022-09-20 | 帝沃检测技术(上海)有限公司 | Threshold selection method and equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE29719600U1 (en) * | 1997-11-05 | 1998-01-02 | Rgi Industriemesgeraete Gmbh | Device for detecting inhomogeneities, in particular foreign bodies, in a material layer to be checked |
| EP0872742A1 (en) * | 1997-04-18 | 1998-10-21 | Jacques Dory | Method and system for processing signals representing reflected waves, transmitted or diffracted by a volumetric structure, for the purpose of effecting an investigation and an analysis of this structure |
| RU2212042C1 (en) * | 2002-03-05 | 2003-09-10 | Федеральное государственное унитарное предприятие "Центральный научно-исследовательский институт "Морфизприбор" | Device processing radar images |
| CN1802676A (en) * | 2003-06-11 | 2006-07-12 | 昆腾磁性元件公司 | Screening checkpoint for passengers and baggage |
| CN201141981Y (en) * | 2008-01-02 | 2008-10-29 | 张敏 | Continuous wave Terahertz real-time imaging device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7492861B2 (en) * | 2006-10-13 | 2009-02-17 | Tsinghua University | Apparatus and method for quick imaging and inspecting moving target |
-
2010
- 2010-03-05 CN CN2010101201245A patent/CN101840003B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0872742A1 (en) * | 1997-04-18 | 1998-10-21 | Jacques Dory | Method and system for processing signals representing reflected waves, transmitted or diffracted by a volumetric structure, for the purpose of effecting an investigation and an analysis of this structure |
| DE29719600U1 (en) * | 1997-11-05 | 1998-01-02 | Rgi Industriemesgeraete Gmbh | Device for detecting inhomogeneities, in particular foreign bodies, in a material layer to be checked |
| RU2212042C1 (en) * | 2002-03-05 | 2003-09-10 | Федеральное государственное унитарное предприятие "Центральный научно-исследовательский институт "Морфизприбор" | Device processing radar images |
| CN1802676A (en) * | 2003-06-11 | 2006-07-12 | 昆腾磁性元件公司 | Screening checkpoint for passengers and baggage |
| CN201141981Y (en) * | 2008-01-02 | 2008-10-29 | 张敏 | Continuous wave Terahertz real-time imaging device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101840003A (en) | 2010-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107314741A (en) | Measurement of cargo measuring method | |
| CN101840003B (en) | Green channel open vehicle radar detection method for metal contraband articles | |
| US7142150B2 (en) | Method and system for detecting an object using a composite evidence grid | |
| DE102015100134B4 (en) | Method for detecting and tracking objects using multiple radar sensors | |
| US6750805B1 (en) | Full polarization synthetic aperture radar automatic target detection algorithm | |
| CN102508219B (en) | Turbulent current target detection method of wind profiler radar | |
| CN110488393A (en) | X-band dual-polarization radar measurement method and system | |
| US20130082858A1 (en) | Classification of subsurface objects using singular values derived from signal frames | |
| CN108197610A (en) | A kind of track foreign matter detection system based on deep learning | |
| CN110520896A (en) | Weeding system and method, railway weed killing waggon | |
| CN111123212B (en) | Signal processing method of scene surveillance radar based on complex clutter background | |
| CN103149599A (en) | System and method for cab avoidance of vehicles of highway green detection channel | |
| CN107957259B (en) | Wheelmark cross direction profiles measuring system and measurement method | |
| CN101839978B (en) | Freeway green channel open vehicle load factor obtaining method | |
| KR100923963B1 (en) | Portable vehicle detector system using complex sensor and control method for processing the same | |
| CN103438794B (en) | A kind of engineering structure impact position based on piezoelectric sensing network determines method | |
| CN114266892A (en) | Pavement disease identification method and system for multi-source data deep learning | |
| CN108535730A (en) | A kind of Doppler weather radar solution velocity ambiguity method and system | |
| CN107202986A (en) | Radar installations and target object detection method | |
| CN104280566A (en) | Low altitude wind shear wind speed estimation method based on space-time amplitude and phase estimation | |
| CN207946146U (en) | A kind of green detecting system that is open to traffic | |
| Thompson et al. | Comparing satellite and helicopter-based methods for observing crevasses, application in East Antarctica | |
| WO2018008637A1 (en) | Method for estimating abnormal part of object being surveyed | |
| CN103714604B (en) | Based on the bus passenger flow statistical system of ultrasonic sensor | |
| CN113310851B (en) | Real-time turbid medium target identification system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120725 Termination date: 20130305 |