CN100452110C - Automobile video frequency discrimination speed-testing method - Google Patents
Automobile video frequency discrimination speed-testing method Download PDFInfo
- Publication number
- CN100452110C CN100452110C CNB2006100471946A CN200610047194A CN100452110C CN 100452110 C CN100452110 C CN 100452110C CN B2006100471946 A CNB2006100471946 A CN B2006100471946A CN 200610047194 A CN200610047194 A CN 200610047194A CN 100452110 C CN100452110 C CN 100452110C
- Authority
- CN
- China
- Prior art keywords
- image
- license plate
- car
- car plate
- plate
- 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 vehicle video identification and speed measuring method. The concrete steps are as follows: (a). obtain a vehicle picture; (b). search the partial threshold value of vehicle license plate on the vehicle picture to obtain the license picture; (c). redress the inclination of the license picture; (d). search the frame of the license plate on the redressed picture and calculate the height of characters; (e). standardize characters on the license plate to obtain a number of partial blocks; (f). carry out the binary system for the partial blocks of each character to obtain two-digit pictures; (g). identify characters on the two-digit pictures to obtain characters on the license plate; (h). track the moving route of the license plate through results of step b and d; (i). calculate the moving distance of the vehicle body; and (j). calculate the running speed. The invention has the advantages of high precision of identifying vehicle licenses and low impact on light reflection of license plate, easy realization, and high practical value.
Description
Technical field
The present invention relates to a kind of method that whether the monitoring stream motor vehicle breaks rules and regulations on highway, specifically a kind of automobile video frequency discrimination speed-testing method.
Background technology
Along with improving constantly of developing of automobile industry and people's living standard, vehicle has become indispensable walking-replacing tool, all there is every day a large amount of vehicles to pour into the streets, no matter this has proposed new problem to traffic administration. at common road or on highway, capital vehicles peccancy such as occur driving over the speed limit, can't guarantee traffic safety to greatest extent. in order to alleviate this nervous situation, vehicle supervision department has installed and used automatic monitoring system on highway, this system is mostly by carrying out edge extracting to license plate image, Hough transformation, license plate number is discerned in the calculating of histogram analysis and morphological operator, this automatic monitoring is applicable to that the speed of a motor vehicle is certain, under the stable situation of vehicle flowrate, but for precipitate variation, as chasing the vehicle that quickens to pass through of escaping and then have the low problem of discrimination for hiding police car; Many owing to this detection method operand again, as utilize Hough transformation to vote, the zone for determining of triumph though this method is more accurate, requires bigger storage space and computing time, is not suitable for real-time application.
Summary of the invention
For solve above-mentioned of the prior art low to the automobile storage passed through fast at discrimination, be not suitable for real-time application problem, no matter the object of the present invention is to provide the automobile video frequency discrimination speed-testing method that all can detect constantly at a kind of daytime or night.
For achieving the above object, the technical solution used in the present invention is:
This method has following steps:
A. obtain the car body image;
B. the local threshold of search car plate on above-mentioned car body image obtains license plate image;
C. above-mentioned license plate image is carried out the pitch angle revisal;
D. on the license plate image after the revisal of inclination angle, search the car plate frame region, calculate word height;
E. the trade mark word graph control gauge in the license plate image is formatted, obtain a plurality of local fritters;
F. the local fritter to each literal carries out binarization, the output bianry image;
G. in the enterprising style of writing word identification of above-mentioned bianry image, obtain the license plate number literal;
H. connect above-mentioned steps d, utilize the result of step b and d in one group of consecutive image, to follow the trail of the car plate mobile route;
I. calculate car body displacement;
J. calculate running speed.
Described step b searches for car plate on above-mentioned car body image local threshold also has following process:
B1. license plate area and background area are cut apart;
B2. whole license plate images are carried out rim detection, then license plate image is divided into the fritter of identical size by unit picture element (present embodiment is unit with 8*8);
B3. calculate reference value for each fritter, all license plate image is a unit with a fritter, obtains the local ranks reference value of license plate image;
B4. according to above-mentioned ranks reference value the license plate image binaryzation, obtain the bianry image of car plate.
Described step c carries out the pitch angle revisal to license plate image and has following steps:
C1. be partitioned into character area at license plate area;
C2. above-mentioned character area ask its line direction to disperse and;
C3. obtain in whole number plate zones that line direction disperses and;
C4. the line direction that calculates non-legible zone on whole car plates divides bulk density;
C5. repeat c1~c4 process in the car plate image rotating between [5 °, 5 °], the line direction of obtaining non-legible zone in each rotated image divides bulk density;
C6. when d value hour, finally the pitch angle of car plate is the car plate rotation angle of the revisal of wanting.
Described steps d is searched the car plate frame region, calculates word height and may further comprise the steps:
D1. license plate image is carried out intensive treatment;
D2. utilize the specification information of car plate that each character area is separated, the decision text point, promptly under the prerequisite that satisfies the car plate ratio, search white space brightness on the license plate image and the minimum value zone determine white space between literal and the literal, further definite character area;
D3. the standard specification information according to car plate obtains the literal height;
Described step e is specially license plate number literal graphic standardization: the literal size specification is turned to 16*16 local fritter.
Described step f carries out binarization to the local fritter of each literal, and the output bianry image also comprises:
F1. the local fritter that will obtain in step e is divided into a plurality of squares;
F2. 4 that obtain the same square of inclusion are the square group of its twice length of side;
F3. obtain the Zone Full reference value in the square group of 4 twice length of sides;
F4. determine the reference value of square with the mean value of above-mentioned Zone Full reference value;
F5. on the character image of local fritter, in kind obtain the regional reference value of other each squares;
F6. obtain threshold matrix according to The above results about the character image of local fritter;
F7. according to this threshold matrix, each character image is carried out 2 systemizations, obtains bianry image.
In step g, utilize Artificial Neural Network to carry out literal identification to the bianry image that in step f, obtains.
Described step h determines that in one group of consecutive image the car plate mobile route may further comprise the steps:
H1. mistiming of mistiming of present frame and former frame in the real time record image column, present frame and later frame, and obtain two edge images;
H2. two edge images binarization respectively of obtaining, obtain two bianry images;
H3. above-mentioned two bianry images are carried out logical operation;
H4. in above-mentioned operation result, judge the size of the external figure of car body by histogrammic method;
H5. when car body shows that on image its external figure is maximum, obtain car plate and be the ratio of pixel and the height pixel of whole locomotive body image of the height of benchmark with the whole locomotive body image, this ratio is the 1st ratio;
H6. obtain the ratio of car plate reference position width pixel and whole locomotive body picture traverse pixel on the whole locomotive body image, this ratio is the 2nd ratio;
H7. calculate the ratio of height (or width) pixel just entered the car body image behind the camera coverage and height (or width) pixel of rolling the car body image before the camera coverage away from, this ratio is the 3rd ratio;
H8. obtain the coordinate of car plate in the entire image of the upper left corner by the 1st~3 ratio;
H9. on all frames, obtain the centre coordinate of car plate, list the tracks of decision car plate at all images.
Described step I is calculated car body displacement and is specially: when the free focal length lens of camera head immobilizes, obtain camera head to the distance between the car plate according to the pixel quantity of car plate length direction.
The present invention has following beneficial effect and advantage:
1. the accuracy rate height of identification license plate number. the mode identification technology of picture signal treatment technology and artificial neural network is cut apart etc. in the inventive method utilization to pitch angle revisal, the character area of license plate image, with utilize the Hough transform method and compare, improved the accuracy rate that the car plate discrimination and the speed of a motor vehicle detect, the car plate discrimination is daytime 99.95%, at night 98%, the speed of a motor vehicle detects error range and is ± 3Km/h;
2. reduced influence to the reflection of car plate light.The present invention has adopted the method for local binaryzation to license plate image, and efficiently solving because of the reflection of light car plate light influence the problem that literal is discerned;
3. be suitable for real-time processing.The present invention effectively utilizes the information resources of vehicle arrangement such as car plate specification characteristic to carry out the identification of license plate number, so recognition speed is very fast, the vehicle movement track is searched in real time and the car plate literal is discerned, can in time monitor out for over-speed vehicles, the speed of a motor vehicle identification processing time is in 200ms, and, can catch definite vehicles peccancy rapidly by license plate number identification;
4. be easy to realize, implementation is strong. and only utilize picture signal because of the inventive method and discern or detect, the device of using the inventive method does not need infrared transceiver, no matter daytime or night, do not need to select to be provided with, even under vehicle is turned on light the situation of head-on sailing, also can accurately discern the car plate literal.
Description of drawings
Fig. 1 is a structured flowchart of the present invention;
Result images and the data of Fig. 2 (a)~(e) in image processing process, export for the present invention;
The car plate specification information synoptic diagram of Fig. 3 for using in the inventive method;
Fig. 4 (a)~(e) is a license plate number literal graphic standard process synoptic diagram in the inventive method;
Fig. 5 is for calculating the actual displacement synoptic diagram of car body.
Embodiment
The inventive method realizes by the control program that is installed in the hardware device, as shown in Figure 1,
The inventive method comprises the steps:
A. obtain the car body image;
B. the local threshold of search car plate on above-mentioned car body image obtains license plate image;
C. above-mentioned license plate image is carried out the pitch angle revisal;
D. on the license plate image after the revisal of inclination angle, search the car plate frame region, calculate the literal height;
E. to the trade mark literal graphic standardization in the license plate image, be that local fritter is divided on the basis with each literal;
F. the local fritter to each literal carries out binarization, the output bianry image;
G. in the enterprising style of writing word identification of above-mentioned bianry image, obtain the license plate number literal;
H. connect above-mentioned steps d, utilize the result of step b and d in one group of consecutive image, to follow the trail of the car plate mobile route;
I. calculate the actual displacement of car body according to the car plate mobile route;
J. calculate running speed.
Described step b searches for car plate on above-mentioned car body image local threshold also has following process:
B1. utilize and ask extremum method that license plate area and background area are cut apart, present embodiment adopts the method for minimizing point that license plate area and background area are cut apart;
B2. whole license plate images are carried out rim detection, then license plate image is divided into the fritter of identical size by unit picture element (present embodiment is unit with 8*8);
B3. calculate luminance reference value (threshold value) for each fritter, all license plate image is a unit with a fritter, obtains the local ranks reference values (local luminance threshold matrix) of license plate image, and promptly the piece for each 8*8 calculates 1 luminance reference value:
T1
i,j=M
i,j+2.2б
i,j (1)
M wherein
I, jBe the average brightness of whole license plate images, б
I, jLuminance standard deviation for 8*8BLOCK; Because of the top repeatedly process of each 8*8 fritter, be unit for car body image 8*8, the threshold matrix (Threshold matrix) after can obtaining to localize;
B4. according to above-mentioned threshold matrix (Threshold matrix) the license plate image binaryzation, obtain the bianry image of car plate, promptly according to all license plate image binarizations of above-mentioned threshold matrix handle, the occurrence rate height of license plate area 1 on binary picture, the blank characteristic that can correctly reflect car plate, so utilize the specification characteristic of car plate to be easy to detect the car plate field, concrete grammar is as follows:
Carry out morphology (morphology) computing of unit matrix (element matrix) for (3,3) size for the binary picture of obtaining, the computing form is open (opening); Carry out unit matrix (element matrix) then and be expand (dilation) for (16,3) big or small morphology (morphology) computing computing form; Utilize the characteristic (laterally with size longitudinally, laterally with longitudinally than, the black of number plate characteristic) of number plate to obtain the number plate field on the image that in the end obtains with white.
Shown in Fig. 2 (a), be the license plate area image that obtains after license plate area and background area are cut apart.
Described step c carries out the pitch angle revisal to license plate image and has following steps:
C1. be partitioned into character area at license plate area;
On the license plate image of gathering, need carry out Region Segmentation to literal, owing to often run into license plate sloped situation, meeting is very difficult and can reduce discrimination.So before the literal subregion, the pitch angle revisal of the car plate that need tilt. the maximum inclination angle is limited in [5 ° ,+5 °] in the present embodiment; Basic thought is that car plate is rectified, during horizontal positioned, on the blank bar zone of car plate upper and lower part, line direction divides the bulk density minimum; Described line direction divide bulk density be with each behavior unit obtain dispersion value (VARIANCE) with after again divided by line number.
Pixel wide with font weight is that unit carries out the line direction differential, obtains the line direction differential matrix; The line direction differential matrix is added up, obtain differential and S
i(SUMMATION OFDIFFERNCE), this differential and comprised column vector (COLUMN DIRECTION) information;
S
i=∑
N j=1(P
ij-P
ij+a)
Wherein: S
iFor the capable differential of i and; N is picture traverse (unit is a pixel); A is a set width least unit pixel; P
IjBe (i, j) brightness value of pixel; P
Ij+aFor (i, j+a) brightness value of pixel.
Obtain the average brightness of column vector,, think that the zone of this section column vector correspondence is a character area when the average brightness of column vector during continuously greater than this mean value.
C2. ask its line direction to disperse and (ROW DIRECTIONSUMMATION OF VARIANCE line direction variance and) at above-mentioned character area;
Vi=1/N(∑
N j=1(M
i-P
ij)
2)
1/2
M
i=1/N∑
N j=1(P
ij/N)
V
0=∑
H0 i=1V
i
Wherein: H0 is the height (unit is a pixel) in literal bar zone; Vi is the capable dispersion value (variance yields) of i; Mi is the capable average brightness of i; V
0Be the capable dispersion value of i and;
C3. whole license plate areas obtain that line direction disperses and;
Vi=1/N(∑
N j=1(M
i-P
ij)
2)
1/2
M
i=1/N∑
N j=1(P
ij/N)
V=∑
H i=1V
i
Wherein: V be in the car plate Zone Full dispersion value and; H is the whole height of car plate.
C4. the line direction that calculates non-legible zone on whole car plates divides bulk density:
d=(V-V
0)/(H-H
0)
Wherein: d is that the line direction in non-legible zone divides bulk density;
C5. repeat c1~c4 process in the car plate image rotating between [5 °, 5 °], the line direction of obtaining non-legible zone in each rotated image divides bulk density;
C6. when d value hour, finally the pitch angle of car plate is the car plate rotation angle of the revisal of wanting:
[d,α]=min(d(K);
Wherein: α is the angle of inclination of car plate, and span is [5 ° ,+5 °], and when d (k) is hour angle of rotation, promptly car plate needs the pitch angle of revisal;
D (K) is that the line direction in non-legible zone divides bulk density in the K time image rotating;
Shown in Fig. 2 (b), for license plate image through the pitch angle revisal and carry out the image that obtains after character area is cut apart.
Described steps d is searched the car plate frame region, calculates word height and may further comprise the steps:
D1. license plate image is carried out intensive treatment;
D2. utilize the specification information of car plate that each character area is cut apart, the decision text point, promptly under the prerequisite that satisfies car plate ratio characteristic, by search brightness on the license plate image and the minimum value zone determine white space between literal and the literal, further definite character area;
The specification information that has utilized car plate cut apart in literal, list at dynamic image, only utilize a scanning field (FIELD of each frame, be equivalent to field), so license plate image dwindles half than actual license plate specification at vertical direction, basic thought is when satisfying the ratio characteristic of car plate specification, seek white space brightness between literal and the literal and the location point of minimum.List a scanning field that only utilizes each frame (frame) at dynamic image, so the fore-and-aft distance than actual license plate dwindles half, therefore the license plate image of being imported dwindles half than actual license plate specification side ratio (side ratio), the actual license plate side ratio is 440: 130, and the car plate side ratio on the license plate image is 440: 65.
As shown in Figure 3, above the relation between word height and the car plate length is as follows for car plate:
L=(9+34/45)H0
Wherein: L is a car plate length; H0 is the literal height on the car plate;
With the conversion between 0.8~1.2 times of the above-mentioned literal height H of obtaining 0, decision can reduce the error of image document for the car plate specification size of each literal height through this conversion;
On Fig. 3 corresponding to the pixel of black part, as the comparison point of car plate specification and real image.
Brightness on the white space of search car plate standard specification on the license plate image and be hour position.
With the lower-left point of car plate as true origin, rectangular coordinate system is set up as two coordinate axis in both sides with car plate, distance from true origin to the search starting point is P0, search height H a=a*H0, a is value in 0.8~1.2 scope, and the comparison point that is used for searching for the car plate standard specification is that the black region of image is determined as follows:
m1=60/45*a*H0;n1=72/45*a*H0;
m2=117/45*a*H0;n2=129/45*a*H0;
m3=174/45*a*H0;n3=208/45*a*H0;
m4=253/45*a*H0;n4=265/45*a*H0;
m5=310/45*a*H0;n5=322/45*a*H0;
m6=367/45*a*H0;n6=379/45*a*H0;
Wherein: m1~m6 is respectively the reference position horizontal ordinate of the 1st~6 black region in coordinate system in the license plate image; N1~n6 is respectively the final position horizontal ordinate of the 1st~6 black region in coordinate system in the license plate image.
If the brightness of the pixel of the white space of license plate image standard specification and be S,
Sa,p
0=∑
6 k=1∑
a*H0 i=0∑
nk j=mkP
ij
When brightness and S hour obtain corresponding a and the value of P0:
S
min=min
0.8<a<1.2{min
0<p0<Δp{S
a,p0}}
Wherein: Δ p is the step-length sum of conversion after every run-down on the license plate image, and a step-length is 1 picture in the present embodiment;
According to car plate starting position in the P0 value decision license plate image of this moment, the suitable literal altitude range of a value decision according to this moment, final decision on the white space of car plate standard specification brightness and for hour, corresponding literal height H O determines each car plate text point on the license plate image according to the car plate standard specification again. shown in Fig. 2 (c).
D3. again according to the standard specification information (ratio characteristic) of car plate, finally obtain the literal height pixel on license plate image;
Described step e formats to license plate number word graph control gauge and may further comprise the steps:
E1. be literal size determined pixel the local fritter of 16*16, shown in Fig. 4 (a); Present embodiment adopts interpolation (INTERPOLATION).
Described step f carries out binarization to the local fritter of the 16*16 of each literal, and the output bianry image comprises;
On the character image of the local fritter of the 16*16 that f1. will obtain in step e, decomposing size is the 4*4 square, shown in Fig. 4 (a);
F2. 4 that obtain to comprise same 4*4 square are the square group of its twice length of side, i.e. 4 8*8 pieces (BLOCK) are shown in Fig. 4 (b)~(e);
F3. utilize the OTSU method to obtain 4 Zone Full reference values (threshold value) in the 8*8 piece
F4. determine the reference value of 4*4 square with the mean value of above-mentioned Zone Full reference value;
F5. on the character image of 16*16, in kind obtain the regional reference value of other each 4*4 squares;
F6. obtain threshold matrix according to The above results about the character image of 16*16;
F7. according to this threshold matrix, each character image is carried out binarization.
In step g, the bianry image that obtains among the step f is carried out literal identification, obtain the license plate number literal, shown in Fig. 2 (d); Present embodiment adopts artificial neural network, the biological neural function of simulation, data are classified, the present invention utilizes 4 layers of artificial neural network for English words and digit recognition, for Chinese Character Recognition is utilized 3 layers of artificial neural network, wherein 4 layers of artificial neural network structure are the inlet number of plies 256, hide the neurocyte (neuron) several 160 of layer 1, it is several 80 to hide the neurocyte (neuron) of layer 2, the outlet number of plies 34.The renewal of strength of joint and threshold value has utilized minimum anxious reinforcement (gradient-descent method gradient gradient method)
For discrimination and conclusion (generalization) ability that improves artificial neural network, make mode of learning (LEARN PATTERN) the quantity unanimity of each literal, increased artificial mode image and statistics formula noisy image.
Described step h determines that in one group of consecutive image the car plate mobile route may further comprise the steps:
H1. the mistiming of the mistiming of present frame and former frame, present frame and later frame in the real time record image column, present embodiment utilizes the PREWITTS operator to obtain two edge image E
1, E
2:
E1=Prewitt(F
b-F
c)
E
2=Prewitt(F
f-F
c)
Wherein: F
cBe present frame (FRAME); F
bBe former frame; F
fBe later frame;
H2. two edge images binarization respectively of obtaining, obtain two bianry image Th1, Th2:
Th1=Threshold(E1);
Th2=Threshold(E2);
The binarization process has been utilized for the standard value of whole threshold values of all images as follows:
T=M+36;
Wherein: M is the average brightness of edge image E1, E2; 6 is the luminance standard deviation of edge image E1, E2;
H3. above-mentioned two bianry images are carried out and (AND) computing, obtain the external figure (present embodiment is a boundary rectangle) of car body image;
Th=(Th1)AND(Th2);
H4. in above-mentioned boundary rectangle, judge the size of the external figure of car body image by histogrammic method;
H5. when car body shows that on image its external figure is maximum, obtain car plate and be the ratio of pixel and the height pixel of whole locomotive body image of the height of benchmark with the whole locomotive body image, this ratio is the 1st ratio;
H6. obtain the ratio of car plate reference position width pixel and whole locomotive body picture traverse pixel on the whole locomotive body image, this ratio is the 2nd ratio;
H7. calculate the ratio of height (or width) pixel just entered the car body image behind the camera coverage and height (or width) pixel of rolling the car body image before the camera coverage away from, this ratio is the 3rd ratio;
H8. obtain the coordinate of car plate in the entire image of the upper left corner by the 1st~3 ratio;
H9. on all frames, obtain the centre coordinate of car plate, list the tracks of decision car plate at all images.
Described step j calculates car body displacement and also comprises:
I1. when the free focal length lens of camera head immobilizes, obtain camera head to the distance between the car plate according to the pixel quantity of car plate length direction;
In order to judge car speed, must know the displacement of this vehicle in the certain hour section; At first obtain this vehicle ' in the present image distance, the free focal length lens state of gamma camera will be known every pixel corresponding distance on real space then.
The present invention has utilized the length of car plate for measuring distance.Under the fixing state of the free focal length lens of video camera, according to the pixel quantity of car plate length direction, the distance between obtaining from the video camera to the license plate number,
As shown in Figure 5, video camera is located at O ' point, its over the ground setting height(from bottom) OO ' be h, if the setting angle of video camera is α, visible angle is α ', and the centre coordinate of car plate is P in the car body image, C ' for vehicle enters the optional position of camera coverage, and what ordered apart from O in this position is y apart from OC ', then:
Wherein :-w/2<p<w/2, w are the height pixel count on the display screen;
J2. calculate the speed of a motor vehicle
In the car body image, the pixel p 1 that vehicle moves (i1, j1)-p2 (i2, j2) Dui Ying actual range is obtained as follows:
V=Y/t
Y=abs(y(i1)-y(i2))
T=N/ frame per second
Time t is the frame number N decision of moving from p1 to p2 according to the pixel that vehicle moves, and V is a Vehicle Speed, and Y is the distance of the actual process of vehicle in time t, shown in Fig. 2 (e), is the speed of a motor vehicle and the recognition time of present embodiment output.
Claims (9)
1. automobile video frequency discrimination speed-testing method is characterized in that having following steps:
A. obtain the car body image;
B. the local threshold of search car plate on above-mentioned car body image obtains license plate image;
C. above-mentioned license plate image is carried out the pitch angle revisal;
D. on the license plate image after the revisal of pitch angle, search the car plate frame region, calculate word height;
E. the trade mark word graph control gauge in the license plate image is formatted, obtain a plurality of local fritters;
F. the local fritter to each literal carries out binarization, the output bianry image;
G. in the enterprising style of writing word identification of above-mentioned bianry image, obtain the license plate number literal;
H. connect above-mentioned steps d, utilize the result of step b and d in one group of consecutive image, to follow the trail of the car plate mobile route;
I. calculate car body displacement;
J. calculate running speed.
2. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described step b searches for car plate on above-mentioned car body image local threshold also has following process:
B1. license plate area and background area are cut apart;
B2. whole license plate images are carried out rim detection, then license plate image is divided into the fritter of identical size by unit picture element;
B3. calculate reference value for each fritter, all license plate image is a unit with a fritter, obtains the local ranks reference value of license plate image;
B4. according to above-mentioned ranks reference value the license plate image binaryzation, obtain the bianry image of car plate;
3. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described step c carries out the pitch angle revisal to license plate image and has following steps:
C1. be partitioned into character area at license plate area;
C2. above-mentioned character area ask its line direction to disperse and;
C3. obtain in whole number plate zones that line direction disperses and;
C4. the line direction that calculates non-legible zone on whole car plates divides bulk density;
C5. repeat c1~c4 process in the car plate image rotating between [5 °, 5 °], the line direction of obtaining non-legible zone in each rotated image divides bulk density;
C6. when the line direction in non-legible zone divided bulk density minimum, the pitch angle of final car plate was the car plate rotation angle of the revisal of wanting.
4. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described steps d is searched the car plate frame region, calculates word height and may further comprise the steps:
D1. license plate image is carried out intensive treatment;
D2. utilize the specification information of car plate that each character area is separated, the decision text point, promptly under the prerequisite that satisfies the car plate ratio, search white space brightness on the license plate image and the minimum value zone determine white space between literal and the literal, further definite character area;
D3. the standard specification information according to car plate obtains the literal height.
5. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described step e is specially license plate number literal graphic standardization: the literal size specification is turned to 16*16 local fritter.
6. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described step f carries out binarization to the local fritter of each literal, and the output bianry image also comprises:
F1. the local fritter that will obtain in step e is divided into a plurality of squares;
F2. obtain 4 of comprising same square and be the square group of its twice length of side;
F3. obtain the Zone Full reference value in the square group of 4 twice length of sides;
F4. determine the reference value of square with the mean value of above-mentioned Zone Full reference value;
F5. on the character image of local fritter, in kind obtain the regional reference value of other each squares;
F6. obtain threshold matrix according to The above results about the character image of local fritter;
F7. according to this threshold matrix, each character image is carried out 2 systemizations, obtain bianry image.
7. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: in step g, utilize Artificial Neural Network to carry out literal identification the bianry image that in step f, obtains.
8. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that: described step h follows the trail of the car plate mobile route and may further comprise the steps in one group of consecutive image:
H1. mistiming of mistiming of present frame and former frame in the real time record image column, present frame and later frame, and obtain two edge images;
H2. two edge images binarization respectively of obtaining, obtain two bianry images;
H3. above-mentioned two bianry images are carried out logical operation;
H4. in above-mentioned operation result, judge the size of the external figure of car body by histogrammic method;
H5. when car body shows that on image its external figure is maximum, obtain car plate and be the ratio of pixel and the height pixel of whole locomotive body image of the height of benchmark with the whole locomotive body image, this ratio is the 1st ratio;
H6. obtain the ratio of car plate reference position width pixel and whole locomotive body picture traverse pixel on the whole locomotive body image, this ratio is the 2nd ratio;
H7. calculate the height just entered the car body image behind the camera coverage or width pixel and the height that rolls the car body image before the camera coverage away from or the ratio of width pixel, this ratio is the 3rd ratio;
H8. obtain the coordinate of car plate in the entire image of the upper left corner by the 1st~3 ratio;
H9. on all frames, obtain the centre coordinate of car plate, list the tracks of decision car plate at all images.
9. by the described automobile video frequency discrimination speed-testing method of claim 1, it is characterized in that described step I calculating car body displacement is specially: when the free focal length lens of camera head immobilizes, obtain camera head to the distance between the car plate according to the pixel quantity of car plate length direction.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2006100471946A CN100452110C (en) | 2006-07-14 | 2006-07-14 | Automobile video frequency discrimination speed-testing method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2006100471946A CN100452110C (en) | 2006-07-14 | 2006-07-14 | Automobile video frequency discrimination speed-testing method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101105893A CN101105893A (en) | 2008-01-16 |
CN100452110C true CN100452110C (en) | 2009-01-14 |
Family
ID=38999776
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB2006100471946A Expired - Fee Related CN100452110C (en) | 2006-07-14 | 2006-07-14 | Automobile video frequency discrimination speed-testing method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100452110C (en) |
Families Citing this family (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB2472793B (en) * | 2009-08-17 | 2012-05-09 | Pips Technology Ltd | A method and system for measuring the speed of a vehicle |
CN101833863B (en) * | 2010-05-17 | 2013-09-11 | 瑞斯康达科技发展股份有限公司 | Method and device for detecting vehicle flow speed, as well as method and system for controlling traffic lights |
CN101937614A (en) * | 2010-06-12 | 2011-01-05 | 北京中科卓视科技有限责任公司 | Plug and play comprehensive traffic detection system |
CN102339531B (en) * | 2010-07-14 | 2014-11-05 | 数伦计算机技术(上海)有限公司 | Road traffic detection system |
RU2419884C1 (en) * | 2010-07-20 | 2011-05-27 | Общество С Ограниченной Ответственностью "Технологии Распознавания" | Method of determining vehicle speed |
CN101976340B (en) * | 2010-10-13 | 2013-04-24 | 重庆大学 | License plate positioning method based on compressed domain |
CN102156480A (en) * | 2010-12-30 | 2011-08-17 | 清华大学 | Unmanned helicopter independent landing method based on natural landmark and vision navigation |
CN102332209B (en) * | 2011-02-28 | 2015-03-18 | 王志清 | Automobile violation video monitoring method |
CN102147858B (en) * | 2011-03-31 | 2012-08-01 | 重庆大学 | License plate character identification method |
CN102436744B (en) * | 2011-09-20 | 2014-12-10 | 中盟智能科技(苏州)有限公司 | Method and device for locking license plate picture |
US9083856B2 (en) * | 2012-03-02 | 2015-07-14 | Xerox Corporation | Vehicle speed measurement method and system utilizing a single image capturing unit |
CN103605960B (en) * | 2013-11-15 | 2016-09-28 | 长安大学 | A kind of method for identifying traffic status merged based on different focal video image |
WO2015147764A1 (en) * | 2014-03-28 | 2015-10-01 | Kisa Mustafa | A method for vehicle recognition, measurement of relative speed and distance with a single camera |
CN104318782B (en) * | 2014-10-31 | 2016-08-17 | 浙江力石科技股份有限公司 | The highway video frequency speed-measuring method of a kind of facing area overlap and system |
CN107292932B (en) * | 2016-04-07 | 2020-06-09 | 上海交通大学 | Head-on video speed measurement method based on image expansion rate |
CN107730614A (en) * | 2017-01-05 | 2018-02-23 | 西安艾润物联网技术服务有限责任公司 | Parking charge method and device |
CN107392313B (en) * | 2017-06-12 | 2020-09-29 | 五邑大学 | Steel rail identification method based on deep learning |
CN108052866A (en) * | 2017-11-17 | 2018-05-18 | 克立司帝控制系统(上海)有限公司 | Car license recognition learning method and system based on artificial neural network |
CN108320531A (en) * | 2018-04-04 | 2018-07-24 | 武汉市技领科技有限公司 | A kind of speed measuring equipment and velocity-measuring system |
CN109615879B (en) * | 2018-12-28 | 2020-12-11 | 成都路行通信息技术有限公司 | Vehicle speed abnormity early warning model and method based on Internet of vehicles and model construction method |
CN109767629A (en) * | 2019-02-01 | 2019-05-17 | 深圳市润泽材料科技有限公司 | A kind of camera system and its speed-measuring method of integrated speed measuring function |
CN110632339A (en) * | 2019-10-09 | 2019-12-31 | 天津天地伟业信息系统集成有限公司 | Water flow testing method of video flow velocity tester |
CN110705493A (en) * | 2019-10-10 | 2020-01-17 | 深圳市元征科技股份有限公司 | Method and system for detecting vehicle running environment, electronic device and storage medium |
CN112199545B (en) * | 2020-11-23 | 2021-09-07 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
CN113030506B (en) * | 2021-03-25 | 2022-07-12 | 上海其高电子科技有限公司 | Micro-area speed measurement method and system based on multi-license plate calibration library |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0668376A (en) * | 1992-08-17 | 1994-03-11 | Omron Corp | Speed violation warning device |
JPH0883393A (en) * | 1994-09-14 | 1996-03-26 | Hitachi Ltd | Vehicle speed measuring instrument |
JP2001126184A (en) * | 1999-10-29 | 2001-05-11 | Matsushita Electric Ind Co Ltd | Automatic license plate recognizing device and vehicle speed measuring method |
JP2003228793A (en) * | 2002-02-04 | 2003-08-15 | Matsushita Electric Ind Co Ltd | Reader for front and rear license plates |
US20050008194A1 (en) * | 1999-05-28 | 2005-01-13 | Satoshi Sakuma | Apparatus and method for image processing |
CN1737578A (en) * | 2004-08-19 | 2006-02-22 | 昆明利普机器视觉工程有限公司 | Road vehicle speed measuring method realized only by video |
-
2006
- 2006-07-14 CN CNB2006100471946A patent/CN100452110C/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0668376A (en) * | 1992-08-17 | 1994-03-11 | Omron Corp | Speed violation warning device |
JPH0883393A (en) * | 1994-09-14 | 1996-03-26 | Hitachi Ltd | Vehicle speed measuring instrument |
US20050008194A1 (en) * | 1999-05-28 | 2005-01-13 | Satoshi Sakuma | Apparatus and method for image processing |
JP2001126184A (en) * | 1999-10-29 | 2001-05-11 | Matsushita Electric Ind Co Ltd | Automatic license plate recognizing device and vehicle speed measuring method |
JP2003228793A (en) * | 2002-02-04 | 2003-08-15 | Matsushita Electric Ind Co Ltd | Reader for front and rear license plates |
CN1737578A (en) * | 2004-08-19 | 2006-02-22 | 昆明利普机器视觉工程有限公司 | Road vehicle speed measuring method realized only by video |
Also Published As
Publication number | Publication date |
---|---|
CN101105893A (en) | 2008-01-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100452110C (en) | Automobile video frequency discrimination speed-testing method | |
CN105005771B (en) | A kind of detection method of the lane line solid line based on light stream locus of points statistics | |
KR100969995B1 (en) | System of traffic conflict decision for signalized intersections using image processing technique | |
CN109670376B (en) | Lane line identification method and system | |
CN103324930B (en) | A kind of registration number character dividing method based on grey level histogram binaryzation | |
CN104029680B (en) | Lane Departure Warning System based on monocular cam and method | |
Wu et al. | Applying a functional neurofuzzy network to real-time lane detection and front-vehicle distance measurement | |
CN103824452A (en) | Lightweight peccancy parking detection device based on full view vision | |
CN103065138A (en) | Recognition method of license plate number of motor vehicle | |
CN103295420A (en) | Method for recognizing lane line | |
CN110689724B (en) | Automatic motor vehicle zebra crossing present pedestrian auditing method based on deep learning | |
Hechri et al. | Robust road lanes and traffic signs recognition for driver assistance system | |
CN110335467B (en) | Method for realizing highway vehicle behavior detection by using computer vision | |
CN102902983B (en) | A kind of taxi identification method based on support vector machine | |
CN112001216A (en) | Automobile driving lane detection system based on computer | |
CN103593981A (en) | Vehicle model identification method based on video | |
CN114898296A (en) | Bus lane occupation detection method based on millimeter wave radar and vision fusion | |
CN112084900A (en) | Underground garage random parking detection method based on video analysis | |
CN105574502A (en) | Automatic detection method for violation behaviors of self-service card sender | |
CN104574993A (en) | Road monitoring method and device | |
CN107918775B (en) | Zebra crossing detection method and system for assisting safe driving of vehicle | |
CN103440785A (en) | Method for rapid lane departure warning | |
KR101347886B1 (en) | Method and Apparatus for Road Lane Recognition by Surface Region and Geometry Information | |
CN113011331B (en) | Method and device for detecting whether motor vehicle gives way to pedestrians, electronic equipment and medium | |
Ren et al. | Automatic measurement of traffic state parameters based on computer vision for intelligent transportation surveillance |
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: 20090114 Termination date: 20120714 |