CN107292260A - The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp - Google Patents
The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp Download PDFInfo
- Publication number
- CN107292260A CN107292260A CN201710452155.2A CN201710452155A CN107292260A CN 107292260 A CN107292260 A CN 107292260A CN 201710452155 A CN201710452155 A CN 201710452155A CN 107292260 A CN107292260 A CN 107292260A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- car light
- vehicle
- car
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Geometry (AREA)
- Lighting Device Outwards From Vehicle And Optical Signal (AREA)
Abstract
The invention discloses a kind of thick fog day vehicle checking method for associating pairing with fog lamp based on vehicle head lamp, comprise the following steps：1) image sequence is obtained by original thick fog day Traffic Surveillance Video, extracts the G component maps of image；2) adaptive threshold calculations are carried out according to the brightness of car light under image G component maps, morphologic processing and the extraction of boundary rectangle frame is carried out to the vehicle lamp area being partitioned into；3) screening pairing is carried out by the relation between car light under the vehicle lamp area boundary rectangle frame of extraction；Then by matching the detection for associating the pairing completed between two groups of car lights pair, completing vehicle between headlamp and pairing fog lamp；4) rejecting of reflected light is carried out using the neighborhood characteristics of reflected light, eliminates its interference to vehicle detection.The present invention can be matched under thick fog day adverse circumstances by the pairing based on car light and the association of headlamp and fog lamp, intactly detect the vehicle in thick fog day video image.
Description
Technical field
The present invention relates to vehicle testing techniques, more particularly to a kind of thick fog for associating pairing with fog lamp based on vehicle head lamp
Its vehicle checking method.
Background technology
Under foggy environment, because visibility is relatively low, driver does not see the traffic sign, pavement facilities and pedestrian of surrounding
Deng so as to easily cause the generation of traffic accident.Meanwhile, during travelling in fog day, speed is generally slower, in height on and off duty
During peak, vehicle flowrate is larger, easily causes congested with cars.Therefore, detecting and tracking is carried out to vehicle under foggy environment, extracts corresponding
Traffic parameter, formulation traffic guidance policy timely to relevant departments, carry out Traffic flux detection and rescue with high safety have it is non-
Chang Chong great meaning.
In the case of thick fog, the gray value between moving target and background differs very little, the various features such as profile of vehicle
Profile, color, texture etc. are difficult intactly to be presented in frame of video, and background modeling is complicated, and object of reference is few, and background subtraction is no longer
It is feasible.And road speed is slow in the case of thick fog, the detection for carrying out vehicle using frame differential method can not detect the complete of vehicle
Whole profile even can not detect vehicle.Therefore, in current vehicle testing techniques, be badly in need of construction one kind can thick fog day this
Plant the vehicle checking method under inclement weather conditions.
The content of the invention
The technical problem to be solved in the present invention is to be based on vehicle head lamp there is provided one kind for defect of the prior art
The thick fog day vehicle checking method of pairing is associated with fog lamp.
The technical solution adopted for the present invention to solve the technical problems is：One kind is associated based on vehicle head lamp with fog lamp matches somebody with somebody
To thick fog day vehicle checking method, comprise the following steps：
1) image sequence is obtained by original thick fog day Traffic Surveillance Video, the characteristics of degraded image strong according to thick fog day
Image preprocessing, the best G component maps of extraction effect are carried out, tri- kinds of R, G, B that described image pretreatment includes extracting image leads to
Road image, and the indexs such as the definition of each channel image, noise size are contrasted, choose the best channel image of effect and carry out subsequently
Processing；The thick fog day is less than 0.1 kilometer of weather conditions for visibility；
2) adaptive threshold calculations, reinforcing car light area are carried out according to the brightness of car light under image G component maps
The feature in domain is simultaneously split, and is then carried out morphologic processing to the vehicle lamp area being partitioned into, is carried out again afterwards external
The extraction of rectangle frame；
3) screening pairing, the car light are carried out by the relation between car light under the vehicle lamp area boundary rectangle frame of extraction
Including headlamp and fog lamp；Then by match headlamp and pairing fog lamp between associate completion two groups of car lights pair between match somebody with somebody
It is right, complete the detection of vehicle；
4) rejecting of reflected light is carried out using the neighborhood characteristics of reflected light, eliminates its interference to vehicle detection.
By such scheme, the step (2) includes：
2.1) grey level histogram of traversing graph picture, and all troughs therein are labeled first, then according to thick fog
The highlighted characteristic of its car light, chooses the threshold value that the maximum trough of gray value is split as car light, the foreground image being partitioned into；
2.2) when carrying out Morphological scale-space to foreground image, an opening operation is carried out to image using 3*3 structure first,
The noise spot in binary image is eliminated, then chooses identical structural element and carries out a closed operation, the sky of filling car light part
Hole；
2.3) the boundary rectangle frame of car light is extracted using region-growing method, while recording the positional information and length and width of car light
Than.
By such scheme, the step 2.3) in record car light positional information and length-width ratio it is specific as follows：
A car light chained list List1 is set up, boundary rectangle frame C is stored respectivelyiThe lower left corner and upper right angular coordinate (xLD,i,
yLD,i), (xRU,i,yRU,i), and its length-width ratio is CKi, and the connected region is numbered, the car light quantity in each region is set
It is set to 1, wherein CKiMeet following formula,
In formula：CKiFor car light length-width ratio；(xLD,i,yLD,i) it is rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i) it is square
Shape frame CiUpper right angular coordinate；
Center point coordinate (the h of minimum enclosed rectangle framei,wi) meet following formula：
hi=(xLD,i+xRU,i)/2,wi=(yLD,i+yRU,i)/2
In formula：hiFor minimum enclosed rectangle frame CiCentral point abscissa；wiFor minimum enclosed rectangle frame CiCentral point is vertical to be sat
Mark；(xLD,i,yLD,i) it is rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i) it is rectangle frame CiUpper right angular coordinate.
By such scheme, the step 3) in vehicle detection specific method it is as follows：
3.1) calculating of the matching rate of car light pairing：
In formula：E is car light matching rate；α is the weight of car light symmetry；EangleFor the inclination of two car lights in the horizontal direction
Degree；βFor the weight of car light spacing；EdistFor the matching rate of car light spacing；γ is the weight of car light similitude；EAreaFor two cars
The normalization difference of lamp area；
Each car light has and only another car light is matching, and the car light of matching degree highest two is successful matching；
3.2) after the match is successful, a car light pairing chained list List2, the car light numbering and its seat of storage successful matching are set up
Information is marked, and the car light quantity of the car light centering is updated to 2.If car light CiWith CjSuccessful matching, and stored according in List1
Information understand, CiLower-left angular coordinate be (xLD,i,yLD,i), car light CjUpper right angular coordinate be (xRU,j,yRU,j), then set up one
Individual new rectangle frame M, and deposit in List2 its lower left corner and upper right angular coordinate (xLD,i,yLD,i), (xRU,j,yRU,j), M is represented
One car light pair；
3.3) Chinese herbaceous peony illuminator group and fog lamp group associate matching.
By such scheme, the step 3.1) in car light match feature calculation it is as follows：
The inclined degree E of two car lights in the horizontal directionangleFollowing formula should then be met：
In formula：EangleFor the inclined degree of two car lights in the horizontal direction；(h1,w1) be car light 1 to be matched central point
Coordinate；(h2,w2) be car light 2 to be matched center point coordinate；
If dthFor distance between the car light of setting, then between car light distance matching rate EdistMeet following formula：
In formula：EdistFor the matching rate of distance between car light；dthFor distance between the car light of setting；
If A1、A2The respectively area of two car lights, EAreaFor the normalization difference of two car light areas, then two car light sizes
Similarity is shown below：
In formula：EAreaFor the normalization difference of two car light areas；A1For the area of car light 1 to be matched；A2For car light to be matched
2 area.
By such scheme, the step 3.3) Chinese herbaceous peony illuminator group and fog lamp group to associate matching specific as follows：
The size of the centroid position of two groups of car lights pair, fore-and-aft distance and minimum external matrix frame is matched as association
Standard, its process is as follows：
Its centroid position is calculated according to the coordinate position of the boundary rectangle frame of each car light pair, if car light is to MiThe lower left corner sit
It is designated as (xLD,i,yLD,i), upper right angular coordinate is (xRU,i,yRU,i), then its centroid position (xc,i,yc,i) meet following formula：
In formula：(xc,i,yc,i) it is i-th group of car light centroid position coordinate；(xLD,i,yLD,i) for car light to MiThe lower left corner sit
Mark；(xRU,i,yRU,i) for car light to MiUpper right angular coordinate；MiFor in above-mentioned steps 3.2) defined in car light pair, this interval scale
I-th of car light pair；
Compare the size of two rectangle frames first, when the area of two rectangle frames is similar, then certain threshold value be set,
Compare the abscissa of its centroid position and the difference of ordinate, when the absolute value of the difference is in threshold range, that is, think this
Two groups of car lights are to belonging to same vehicle.It is specifically shown in following formula：
In formula：S (i, j) is group car light pair of judged result, i.e., i-th and jth group car light to whether belonging to same vehicle；TxFor
Two rectangle frame centroid position abscissa differences；TyFor two rectangle frame centroid position ordinate differences；(xc,i,yc,i) it is i-th
Group car light centroid position coordinate；(xc,j,yc,j) it is jth group car light centroid position coordinate；
By experiment, suitable T is chosenx, TyValue, precision is matched to obtain higher association.For association successful matching
Car light is to Mi, Mj, its minimum enclosed rectangle frame is extracted, and a car is characterized with this rectangle frame, vehicle detection knot is obtained with this
Really, the numbering of each car light and the boundary rectangle frame coordinate of vehicle are deposited while setting up a vehicle chained list List3, and by car light
Quantity be updated to 3.If the coordinate in the rectangle frame lower left corner and the upper right corner is respectively (XLD,k, YLD,k)、(XRU,k, YRU,k)。
It can be seen from priori, left side fog lamp is always in headlamp lower right in the picture, and right side fog lamp is then in headlamp
Lower left, from pairing chained list List2 in obtain Mi, MjThe upper left corner and upper right corner coordinate position (xLD,i, yLD,i)、(xLD,j,
yLD,j)、(xRU,i, yRU,i)、(xRU,j, yRU,j), then the coordinate of each vehicle meets following formula in vehicle chained list：
In formula：(XLD,k, YLD,k) it is the lower-left angular coordinate for representing vehicle rectangle frame；(XRU,k, YRU,k) it is to represent vehicle rectangle
Upper right angular coordinate (the x of frameLD,i,yLD,i) for car light to MiLower-left angular coordinate；(xRU,i,yRU,i) for car light to MiThe upper right corner sit
Mark；MiFor the car light pair defined in above-mentioned steps (3.2), i-th of car light pair of this interval scale.
By such scheme, the step 4) according to region-growing method search for binary image in foreground target, when this
When place closer to the distance has another foreground target below target, then the target and lower section target are car light；If conversely, the car
Then it is reflected light when lamp is individually present, and the pixel value in reflected light region is set to 0, realizes the rejecting of reflected light.
The beneficial effect comprise that：
The present invention, when vehicle is not being travelled during traveling according to lane line, can be led under thick fog day adverse circumstances
Cross the features such as the symmetry based on car light and enter the pairing of driving lamp and the association pairing of headlamp and fog lamp, intactly detect
Vehicle in thick fog day video image.The complete detection to vehicle is realized under dense foggy environment, traffic control department can therefrom carry
The traffic parameters such as corresponding traffic flow are taken out, and then corresponding traffic parameter is analyzed, the analysis result can be traffic control portion
Door carries out follow-up traffic safety and traffic monitoring management brings theories integration, is lured while for relevant departments traffic can also be formulated in time
Policy, progress Traffic flux detection and rescue with high safety offer one qualitatively data supporting are provided.
Brief description of the drawings
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing：
Fig. 1 is the thick fog day vehicle checking method flow that the embodiment of the present invention associates pairing based on vehicle head lamp with fog lamp
Figure.
Fig. 2 is the flow for the demarcation for carrying out Morphological scale-space and connected region in the embodiment of the present invention to foreground image
Figure.
Fig. 3 is to the flow chart for associating matching of the pairing of car light and headlight with fog lamp in the embodiment of the present invention.
Fig. 4 is entire flow experimental result picture of the embodiment of the present invention.
In Fig. 4：(a) thick fog day source images；(b) source images G component maps；(c) Threshold segmentation figure；(d) Morphological scale-space is imitated
Fruit is schemed；(e) car light minimum enclosed rectangle frame extraction effect figure；(f) car light pairing design sketch；(g) headlamp is associated with fog lamp and matched somebody with somebody
To design sketch；(h) reflected light based on neighborhood rejects design sketch.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that specific embodiment described herein is not used to limit only to explain the present invention
The fixed present invention.
As shown in figure 1, including following step based on the thick fog day vehicle checking method that vehicle head lamp associates pairing with fog lamp
Suddenly：
The sequence of video images shot under step S100, the dense foggy environment of input, the video image can be sent out from traffic control department
Corresponding picture frame sequence is intercepted in the road running situation video of the thick fog weather urban traffic intersection of cloth.
Step S200, color space conversion is carried out to original thick fog day Traffic Surveillance Video during, extract respectively
R, G, channel B image in the sequence of video images, contrast definition, noise size of each channel image etc., each by contrast
The G component maps of image are further processed the factors such as definition, the noise size of channel image, final choice.
Step S300, traversing graph picture grey level histogram, and all troughs therein are labeled, then according to thick fog
The highlighted characteristic of its car light, chooses the threshold value that the maximum trough of gray value is split as car light, the foreground image being partitioned into.
Step S400, main include carrying out Morphological scale-space and company to the foreground image after step S300 is treated
The demarcation in logical region, as shown in Fig. 2 specifically including following two steps：
Step S401, the foreground image progress Morphological scale-space to step S300 after treated.Using 3*3 structure to figure
As carry out an opening operation, eliminate binary image in noise spot, then choose identical structural element carry out a closed operation,
Fill the cavity of car light part.
Step S402, the image progress connected region demarcation treated to step S401.Extracted using region-growing method
The boundary rectangle frame of car light, while recording the positional information and length-width ratio of car light, specific practice is：
A car light chained list List1 is set up, respectively storage rectangle frame CiThe lower left corner and upper right angular coordinate (xLD,i,yLD,i),
(xRU,i,yRU,i), and its length-width ratio is CKi, and the connected region is numbered, the car light quantity in each region is set to 1,
Wherein CKiMeet following formula.
In formula：CKi- car light length-width ratio；(xLD,i,yLD,i)-rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i)-square
Shape frame CiUpper right angular coordinate；
Center point coordinate (the h of minimum enclosed rectangle framei,wi) meet following formula：
hi=(xLD,i+xRU,i)/2,wi=(yLD,i+yRU,i)/2
In formula：hi- minimum enclosed rectangle frame CiCentral point abscissa；wi- minimum enclosed rectangle frame CiCentral point is vertical to be sat
Mark；(xLD,i,yLD,i)-rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i)-rectangle frame CiUpper right angular coordinate.
Step S500, the pairing of car light and headlight associate matching with fog lamp.As shown in figure 3, mainly including following
Two steps：
Step S501, the car light pairing based on multiple features, multiple features refer mainly to the following aspects：
A) symmetry.Because video camera is therefore two car light central points of same vehicle in the horizontal direction when shooting
It is approximately at same horizontal line.Coordinate (the h of two car light central points1,w1), (h2,w2), the inclination of two car lights in the horizontal direction
Degree EangleFollowing formula should then be met：
In formula：EangleThe inclined degree of-two car lights in the horizontal direction；(h1,w1The center point coordinate of)-car light 1；
(h2,w2The center point coordinate of)-car light 2；
Work as EangleValue gets over hour, and the position of two car lights is closer to same horizontal line, and the possibility of successful matching is bigger.When two
When car light is in same level line position, weighed with maximum priority match.
B) between car light distance size.The distance between the headlamp of vehicle two of different model difference, but distance value is one
Determine in interval range, if dthFor distance between the car light of setting, then between car light distance matching rate EdistMeet following formula：
In formula：EdistThe matching rate of distance between-car light；dthDistance between the car light of-setting；
EdistIt is smaller, then, the E higher apart from matching rate between two car lightsdistDetermine of two car lights in same horizontal line
Two priority match.
C) similitude.Size with a pair of car lights is approximately the same, if A1、A2The respectively area of two car lights, EAreaFor
The normalization difference of two car light areas, then the similarity of two car light sizes be shown below：
In formula：EAreaThe normalization difference of-two car light areas；A1The area of-car light 1；A2The area of-car light 2；dth—
Distance between the car light of setting；
Work as EAreaSmaller, the similitude of car light area is higher, EAreaDetermine that immediate two car light of size has the
Three pairing priority.
D) uniqueness.Each car light has and only another car light is matching, and the car light of matching degree highest two is pairing
Success.
Four properties during comprehensive car light pairing, assign different weights integrate pairing, matching rate E calculating side
Method is shown below：
In formula：E-car light matching rate；The weight of α-symmetry；EangleThe inclination journey of-two car lights in the horizontal direction
Degree；βThe weight of-following distance；EdistThe matching rate of distance between-car light；The weight of γ-similitude；EArea- two car light areas
Normalization difference；
E values are smaller, and the probability that the match is successful is bigger.By experiment, optimal characteristic value weight is chosen, so as to obtain higher
Matching precision；
After the match is successful, a car light pairing chained list List2, the car light numbering and its coordinate letter of storage successful matching are set up
Breath, and the car light quantity of the car light centering is updated to 2.If car light CiWith CjSuccessful matching, and according to the letter stored in List1
Knowable to breath, CiLower-left angular coordinate be (xLD,i,yLD,i), car light CjUpper right angular coordinate be (xRU,j,yRU,j), then set up one newly
Rectangle frame M, and deposit in List2 its lower left corner and upper right angular coordinate (xLD,i,yLD,i), (xRU,j,yRU,j), M represents one
Car light pair.
Step S502, Chinese herbaceous peony illuminator group associate matching with fog lamp group.Centroid position mainly by two groups of car lights pair, longitudinal direction
The standard that the size of distance and minimum external matrix frame is matched as association, its process is as follows：
Its centroid position is calculated according to the coordinate position of the boundary rectangle frame of each car light pair, if car light is to MiThe lower left corner sit
It is designated as (xLD,i,yLD,i), upper right angular coordinate is (xRU,i,yRU,i), then its centroid position (xc,i,yc,i) meet following formula：
In formula：(xc,i,yc,i)-i-th group car light centroid position coordinate；(xLD,i,yLD,i)-car light is to MiThe lower left corner sit
Mark；(xRU,i,yRU,i)-car light is to MiUpper right angular coordinate；Mi- the car light pair defined in above-mentioned (3-1), this interval scale i-th
Individual car light pair；
Compare the size of two rectangle frames first, when the area of two rectangle frames is similar, then certain threshold value be set,
Compare the abscissa of its centroid position and the difference of ordinate, when the absolute value of the difference is in threshold range, that is, think this
Two groups of car lights are to belonging to same vehicle.It is specifically shown in following formula：
In formula：Whether group car light pair of s (i, j)-judged result, i.e., i-th and jth group car light are to belonging to same vehicle；Tx—
Two rectangle frame centroid position abscissa differences；Ty- two rectangle frame centroid position ordinate differences；(xc,i,yc,i)-i-th
Group car light centroid position coordinate；(xc,j,yc,j)-jth group car light centroid position coordinate；
By experiment, suitable T is chosenx, TyValue, precision is matched to obtain higher association.By experiment, draw
One class value, works as Tx, TyWhen taking 5 and 15 respectively, association pairing precision highest.For associating the car light of successful matching to Mi, Mj, carry
Its minimum enclosed rectangle frame is taken, and a car is characterized with this rectangle frame, vehicle detection result is obtained with this.Set up one simultaneously
Individual vehicle chained list List3, deposits the numbering of each car light and the boundary rectangle frame coordinate of vehicle, and the quantity of car light is updated into 3.
If the coordinate in the rectangle frame lower left corner and the upper right corner is respectively (XLD,k, YLD,k)、(XRU,k, YRU,k)。
It can be seen from priori, left side fog lamp is always in headlamp lower right in the picture, and right side fog lamp is then in headlamp
Lower left.M is obtained from pairing chained list List2i, MjThe upper left corner and upper right corner coordinate position (xLD,i, yLD,i)、(xLD,j,
yLD,j)、(xRU,i, yRU,i)、(xRU,j, yRU,j), then the coordinate of each vehicle meets following formula in vehicle chained list：
In formula：(XLD,k, YLD,kThe lower-left angular coordinate of)-represent vehicle rectangle frame；(XRU,k, YRU,k)-represent vehicle rectangle
Upper right angular coordinate (the x of frameLD,i,yLD,i)-car light is to MiLower-left angular coordinate；(xRU,i,yRU,i)-car light is to MiThe upper right corner sit
Mark；Mi- the car light pair defined in above-mentioned (3-1), i-th of car light pair of this interval scale.
Step S600, using reflected light neighborhood characteristics carry out reflected light rejecting, eliminate its interference to vehicle detection.
Its specific practice is：The foreground target searched for according to region-growing method in binary image, the place closer to the distance below the target
When there is another foreground target, then the target and lower section target are car light；If conversely, when the car light is individually present, for
Reflected light, and the pixel value in reflected light region is set to 0, realize the rejecting of reflected light.
Said process is the thick fog day whole technology of vehicle checking method for associating pairing with fog lamp based on vehicle head lamp
Scheme, then carries out experiment test, Fig. 4 (a) is thick fog day source images, as can be seen from this figure under dense foggy environment, vehicle
Information seems very fuzzy, is difficult detection.Fig. 4 (b) is the G component maps of source images, from this figure, it can be seen that G component maps are in reduction
During background image, vehicle lamp area is also highlighted.Fig. 4 (c) is the Threshold segmentation figure obtained after step S300 is treated, from this
Figure is it can be seen that headlight can be come out with fog lamp by independent detection, and the area of reflected light is smaller.Fig. 4 (d) is to pass through step
The image obtained after S401 is treated, vehicle head lamp can be separated well with fog lamp as can be seen from this figure.Fig. 4 (e)
For the image obtained after step S402 is treated, car light can be preferably extracted using this method as can be seen from this figure
Minimum enclosed rectangle frame.Fig. 4 (f) is the image obtained after step S501 is treated, as can be seen from the figure for car
Headlamp and fog lamp can preferably successful matching, and simultaneously also with the right rectangular area formed by reflected light
Frame.Fig. 4 (g) is that the vehicle head lamp obtained after step S502 is treated associates pairing design sketch with fog lamp, from the figure
It can be seen that vehicle head lamp and fog lamp have carried out good pairing, and due to reflected light formation rectangle frame not with before vehicle
Illuminator and fog lamp are matched.Fig. 4 (h) is that the reflected light obtained after step S600 is treated rejects vehicle head lamp and mist
Lamp matches design sketch, it will be apparent from this figure that it has not only been sufficiently reserved vehicle head lamp and fog lamp unpaired message, also simultaneously
Reflected light can be rejected well.
It should be appreciated that for those of ordinary skills, can according to the above description be improved or converted,
And all these modifications and variations should all belong to the protection domain of appended claims of the present invention.
Claims (7)
1. a kind of thick fog day vehicle checking method for associating pairing with fog lamp based on vehicle head lamp, it is characterised in that including with
Lower step：
1) image sequence is obtained by original thick fog day Traffic Surveillance Video, G components is extracted according to the strong degraded image of thick fog day
Figure；The thick fog day is less than 0.1 kilometer of weather conditions for visibility；
2) adaptive threshold calculations are carried out according to the brightness of car light under image G component maps, reinforcing vehicle lamp area
Feature is simultaneously split, and is then carried out morphologic processing to the vehicle lamp area being partitioned into, is carried out boundary rectangle again afterwards
The extraction of frame；
3) screening pairing is carried out by the relation between car light under the vehicle lamp area boundary rectangle frame of extraction, the car light includes
Headlamp and fog lamp；Then the pairing completed between two groups of car lights pair is associated between headlamp and pairing fog lamp by matching,
Complete the detection of vehicle；
4) rejecting of reflected light is carried out using the neighborhood characteristics of reflected light, eliminates its interference to vehicle detection.
2. thick fog day vehicle checking method according to claim 1, it is characterised in that the step 2) include：
2.1) grey level histogram of traversing graph picture, and all troughs therein are labeled first, then according to thick fog overhead traveling crane
The highlighted characteristic of lamp, chooses the threshold value that the maximum trough of gray value is split as car light, the foreground image being partitioned into；
2.2) when carrying out Morphological scale-space to foreground image, an opening operation is carried out to image using 3*3 structure first, eliminated
Noise spot in binary image, then choose the closed operation of progress of identical structural element, the cavity of filling car light part；
2.3) the boundary rectangle frame of car light is extracted using region-growing method, while recording the positional information and length-width ratio of car light.
3. thick fog day vehicle checking method according to claim 2, it is characterised in that the step 2.3) middle record car light
Positional information and length-width ratio it is specific as follows：
A car light chained list List1 is set up, boundary rectangle frame C is stored respectivelyiThe lower left corner and upper right angular coordinate (xLD,i,yLD,i),
(xRU,i,yRU,i), and its length-width ratio is CKi, and the connected region is numbered, the car light quantity in each region is set to 1,
Wherein CKiMeet following formula,
<mrow>
<msub>
<mi>CK</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
<mrow>
<mo>|</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
</mfrac>
</mrow>
In formula：CKiFor car light length-width ratio；(xLD,i,yLD,i) it is rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i) it is rectangle frame Ci
Upper right angular coordinate；
Center point coordinate (the h of minimum enclosed rectangle framei,wi) meet following formula：
hi=(xLD,i+xRU,i)/2,wi=(yLD,i+yRU,i)/2
In formula：hiFor minimum enclosed rectangle frame CiCentral point abscissa；wiFor minimum enclosed rectangle frame CiCentral point ordinate；
(xLD,i,yLD,i) it is rectangle frame CiLower-left angular coordinate；(xRU,i,yRU,i) it is rectangle frame CiUpper right angular coordinate.
4. thick fog according to claim 2 day vehicle checking method, it is characterised in that the step 3) in vehicle detection
Specific method it is as follows：
3.1) calculating of the matching rate of car light pairing：
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>E</mi>
<mo>=</mo>
<mi>&alpha;</mi>
<mo>*</mo>
<msub>
<mi>E</mi>
<mrow>
<mi>a</mi>
<mi>n</mi>
<mi>g</mi>
<mi>l</mi>
<mi>e</mi>
</mrow>
</msub>
<mo>+</mo>
<mi>&beta;</mi>
<mo>*</mo>
<msub>
<mi>E</mi>
<mrow>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>+</mo>
<mi>&gamma;</mi>
<mo>*</mo>
<msub>
<mi>E</mi>
<mrow>
<mi>A</mi>
<mi>r</mi>
<mi>e</mi>
<mi>a</mi>
</mrow>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>&alpha;</mi>
<mo>+</mo>
<mi>&beta;</mi>
<mo>+</mo>
<mi>&gamma;</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
In formula：E is car light matching rate；α is the weight of car light symmetry；EangleFor the inclination journey of two car lights in the horizontal direction
Degree；β is the weight of car light spacing；EdistFor the matching rate of car light spacing；γ is the weight of car light similitude；EAreaFor two car lights
The normalization difference of area；
Each car light has and only another car light is matching, and the car light of matching degree highest two is successful matching；
3.2) after the match is successful, a car light pairing chained list List2, the car light numbering and its coordinate letter of storage successful matching are set up
Breath, and the car light quantity of the car light centering is updated to 2.If car light CiWith CjSuccessful matching, and according to the letter stored in List1
Knowable to breath, CiLower-left angular coordinate be (xLD,i,yLD,i), car light CjUpper right angular coordinate be (xRU,j,yRU,j), then set up one newly
Rectangle frame M, and deposit in List2 its lower left corner and upper right angular coordinate (xLD,i,yLD,i), (xRU,j,yRU,j), M represents one
Car light pair；
3.3) Chinese herbaceous peony illuminator group and fog lamp group associate matching.
5. thick fog according to claim 3 day vehicle checking method, it is characterised in that the step 3.1) in car light match
Feature calculation it is as follows：
The inclined degree E of two car lights in the horizontal directionangleFollowing formula should then be met：
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>a</mi>
<mi>n</mi>
<mi>g</mi>
<mi>l</mi>
<mi>e</mi>
</mrow>
</msub>
<mo>=</mo>
<mo>|</mo>
<mfrac>
<mrow>
<msub>
<mi>w</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>w</mi>
<mn>2</mn>
</msub>
</mrow>
<mrow>
<msub>
<mi>h</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>h</mi>
<mn>2</mn>
</msub>
</mrow>
</mfrac>
<mo>|</mo>
</mrow>
In formula：EangleFor the inclined degree of two car lights in the horizontal direction；(h1,w1) be car light 1 to be matched center point coordinate；
(h2,w2) be car light 2 to be matched center point coordinate；
If dthFor distance between the car light of setting, then between car light distance matching rate EdistMeet following formula：
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>h</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>h</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>w</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>w</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>-</mo>
<msub>
<mi>d</mi>
<mrow>
<mi>t</mi>
<mi>h</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
<msub>
<mi>d</mi>
<mrow>
<mi>t</mi>
<mi>h</mi>
</mrow>
</msub>
</mfrac>
</mrow>
In formula：EdistFor the matching rate of distance between car light；dthFor distance between the car light of setting；
If A1、A2The respectively area of two car lights, EAreaFor the normalization difference of two car light areas, then two car light sizes is similar
Degree is shown below：
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>A</mi>
<mi>r</mi>
<mi>e</mi>
<mi>a</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msub>
<mi>A</mi>
<mn>1</mn>
</msub>
<mo>-</mo>
<msub>
<mi>A</mi>
<mn>2</mn>
</msub>
<mo>|</mo>
</mrow>
<msqrt>
<mrow>
<msup>
<msub>
<mi>A</mi>
<mn>1</mn>
</msub>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<msub>
<mi>A</mi>
<mn>2</mn>
</msub>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mfrac>
</mrow>
In formula：EAreaFor the normalization difference of two car light areas；A1For the area of car light 1 to be matched；A2For car light 2 to be matched
Area.
6. thick fog day vehicle checking method according to claim 3, it is characterised in that the step 3.3) Chinese herbaceous peony illuminator group
With fog lamp group associate matching it is specific as follows：
The mark that the size of the centroid position of two groups of car lights pair, fore-and-aft distance and minimum external matrix frame is matched as association
Standard, its process is as follows：
Its centroid position is calculated according to the coordinate position of the boundary rectangle frame of each car light pair, if car light is to MiLower-left angular coordinate be
(xLD,i,yLD,i), upper right angular coordinate is (xRU,i,yRU,i), then its centroid position (xc,i,yc,i) meet following formula：
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>c</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>y</mi>
<mrow>
<mi>c</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
In formula：(xc,i,yc,i) it is i-th group of car light centroid position coordinate；(xLD,i,yLD,i) for car light to MiLower-left angular coordinate；
(xRU,i,yRU,i) for car light to MiUpper right angular coordinate；MiFor in above-mentioned steps 3.2) defined in car light pair, this interval scale i-th
Individual car light pair；
Compare the size of two rectangle frames first, when the area of two rectangle frames is similar, then certain threshold value is set, compared
The abscissa of its centroid position and the difference of ordinate, when the absolute value of the difference is in threshold range, that is, think this two groups
Car light is to belonging to same vehicle.It is specifically shown in following formula：
In formula：S (i, j) is group car light pair of judged result, i.e., i-th and jth group car light to whether belonging to same vehicle；TxFor two
Rectangle frame centroid position abscissa difference；TyFor two rectangle frame centroid position ordinate differences；(xc,i,yc,i) it is i-th group of car
Characteristic of a navigation light heart position coordinates；(xc,j,yc,j) it is jth group car light centroid position coordinate；
For associating the car light of successful matching to Mi, Mj, its minimum enclosed rectangle frame is extracted, and characterize with this rectangle frame one
Car, vehicle detection result is obtained with this, while set up a vehicle chained list List3, deposits the numbering and vehicle of each car light
Boundary rectangle frame coordinate, and the quantity of car light is updated to 3；
If the coordinate in the rectangle frame lower left corner and the upper right corner is respectively (XLD,k, YLD,k)、(XRU,k, YRU,k)；
It can be seen from priori, left side fog lamp is always in headlamp lower right in the picture, and right side fog lamp is then on a left side for headlamp
Lower section, M is obtained from pairing chained list List2i, MjThe upper left corner and upper right corner coordinate position (xLD,i, yLD,i)、(xLD,j, yLD,j)、
(xRU,i, yRU,i)、(xRU,j, yRU,j), then the coordinate of each vehicle meets following formula in vehicle chained list：
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>X</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>Y</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>min</mi>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mi>min</mi>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>y</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>L</mi>
<mi>D</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>X</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>Y</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mrow>
<mi>max</mi>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mi>max</mi>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>y</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>R</mi>
<mi>U</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
<mo>)</mo>
</mrow>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
In formula：(XLD,k, YLD,k) it is the lower-left angular coordinate for representing vehicle rectangle frame；(XRU,k, YRU,k) it is to represent vehicle rectangle frame
Upper right angular coordinate (xLD,i,yLD,i) for car light to MiLower-left angular coordinate；(xRU,i,yRU,i) for car light to MiUpper right angular coordinate；
MiFor the car light pair defined in above-mentioned steps (3.2), i-th of car light pair of this interval scale.
7. thick fog according to claim 1 day vehicle checking method, it is characterised in that the step 4) in given birth to according to region
Foreground target in regular way search binary image, when place closer to the distance has another foreground target below the target, then
The target and lower section target are car light；If conversely, when the car light is individually present, for reflected light, and by reflected light region
Pixel value is set to 0, realizes the rejecting of reflected light.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710452155.2A CN107292260A (en) | 2017-06-15 | 2017-06-15 | The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710452155.2A CN107292260A (en) | 2017-06-15 | 2017-06-15 | The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107292260A true CN107292260A (en) | 2017-10-24 |
Family
ID=60097781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710452155.2A Pending CN107292260A (en) | 2017-06-15 | 2017-06-15 | The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107292260A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319910A (en) * | 2018-01-30 | 2018-07-24 | 海信集团有限公司 | A kind of vehicle identification method, device and terminal |
| CN108749708A (en) * | 2018-05-11 | 2018-11-06 | 长安大学 | The device and method of vehicle fog lamp is automatically turned under a kind of low visibility |
| CN110796698A (en) * | 2019-11-07 | 2020-02-14 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight removing method and device with maximum area and minimum length-width ratio |
| CN110852252A (en) * | 2019-11-07 | 2020-02-28 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight removing method and device based on minimum distance and maximum length-width ratio |
| CN113191248A (en) * | 2021-04-25 | 2021-07-30 | 国能智慧科技发展(江苏)有限公司 | Vehicle deviation route detection system based on video linkage and intelligent Internet of things |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101727748A (en) * | 2009-11-30 | 2010-06-09 | 北京中星微电子有限公司 | Method, system and equipment for monitoring vehicles based on vehicle taillight detection |
| CN103050008A (en) * | 2013-01-14 | 2013-04-17 | 哈尔滨工程大学 | Method for detecting vehicles in night complex traffic videos |
| CN103177587A (en) * | 2013-03-19 | 2013-06-26 | 天津工业大学 | Machine-vision-based night traffic flow detection method |
| CN103208185A (en) * | 2013-03-19 | 2013-07-17 | 东南大学 | Method and system for nighttime vehicle detection on basis of vehicle light identification |
| CN103325102A (en) * | 2013-06-03 | 2013-09-25 | 武汉理工大学 | Method for processing greasy weather degraded image based on histogram equalization |
| CN103440771A (en) * | 2013-09-06 | 2013-12-11 | 天津工业大学 | Application of fuzzy membership grade and feedback correction in night traffic video vehicle detection |
| CN103984917A (en) * | 2014-04-10 | 2014-08-13 | 杭州电子科技大学 | Multi-feature nighttime vehicle detection method based on machine vision |
| CN104318783A (en) * | 2014-10-16 | 2015-01-28 | 西北师范大学 | Method for analyzing night traffic flow through car lamp detection |
-
2017
- 2017-06-15 CN CN201710452155.2A patent/CN107292260A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101727748A (en) * | 2009-11-30 | 2010-06-09 | 北京中星微电子有限公司 | Method, system and equipment for monitoring vehicles based on vehicle taillight detection |
| CN103050008A (en) * | 2013-01-14 | 2013-04-17 | 哈尔滨工程大学 | Method for detecting vehicles in night complex traffic videos |
| CN103177587A (en) * | 2013-03-19 | 2013-06-26 | 天津工业大学 | Machine-vision-based night traffic flow detection method |
| CN103208185A (en) * | 2013-03-19 | 2013-07-17 | 东南大学 | Method and system for nighttime vehicle detection on basis of vehicle light identification |
| CN103325102A (en) * | 2013-06-03 | 2013-09-25 | 武汉理工大学 | Method for processing greasy weather degraded image based on histogram equalization |
| CN103440771A (en) * | 2013-09-06 | 2013-12-11 | 天津工业大学 | Application of fuzzy membership grade and feedback correction in night traffic video vehicle detection |
| CN103984917A (en) * | 2014-04-10 | 2014-08-13 | 杭州电子科技大学 | Multi-feature nighttime vehicle detection method based on machine vision |
| CN104318783A (en) * | 2014-10-16 | 2015-01-28 | 西北师范大学 | Method for analyzing night traffic flow through car lamp detection |
Non-Patent Citations (4)
| Title |
|---|
| WEIBING WAN 等: "Vehicle detection algorithm based on light pairing and tracking at nighttime", 《JOURNAL OF ELECTRONIC IMAGING》 * |
| 余松: "电子警察系统中车辆跟踪及夜间车辆检测算法研究", 《中国优秀硕士学位论文全文数据库（电子期刊） 工程科技II辑》 * |
| 汤春明 等: "基于最小特征匹配代价的夜间车辆检测和追踪", 《计算机应用与软件》 * |
| 许成闯: "基于视频的车流量检测技术研究与实现", 《中国优秀硕士学位论文全文数据库（电子期刊） 工程科技II辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319910A (en) * | 2018-01-30 | 2018-07-24 | 海信集团有限公司 | A kind of vehicle identification method, device and terminal |
| CN108319910B (en) * | 2018-01-30 | 2021-11-16 | 海信集团有限公司 | Vehicle identification method and device and terminal |
| CN108749708A (en) * | 2018-05-11 | 2018-11-06 | 长安大学 | The device and method of vehicle fog lamp is automatically turned under a kind of low visibility |
| CN110796698A (en) * | 2019-11-07 | 2020-02-14 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight removing method and device with maximum area and minimum length-width ratio |
| CN110852252A (en) * | 2019-11-07 | 2020-02-28 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight removing method and device based on minimum distance and maximum length-width ratio |
| CN110796698B (en) * | 2019-11-07 | 2022-11-29 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight removing method and device with maximum area and minimum length-width ratio |
| CN110852252B (en) * | 2019-11-07 | 2022-12-02 | 厦门市美亚柏科信息股份有限公司 | Vehicle weight-removing method and device based on minimum distance and maximum length-width ratio |
| CN113191248A (en) * | 2021-04-25 | 2021-07-30 | 国能智慧科技发展(江苏)有限公司 | Vehicle deviation route detection system based on video linkage and intelligent Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107292260A (en) | The thick fog day vehicle checking method of pairing is associated with fog lamp based on vehicle head lamp | |
| Possatti et al. | Traffic light recognition using deep learning and prior maps for autonomous cars | |
| CN105550665B (en) | A kind of pilotless automobile based on binocular vision can lead to method for detecting area | |
| Wu et al. | Applying a functional neurofuzzy network to real-time lane detection and front-vehicle distance measurement | |
| CN109190523B (en) | Vehicle detection tracking early warning method based on vision | |
| CN105206109B (en) | A kind of vehicle greasy weather identification early warning system and method based on infrared CCD | |
| CN106934808B (en) | Method for identifying and tracking tail lamp of automobile headlight under visual perception | |
| Janahiraman et al. | Traffic light detection using tensorflow object detection framework | |
| CN107924458A (en) | The system and method for object detection | |
| CN109389046B (en) | All-weather object identification and lane line detection method for automatic driving | |
| CN107038420A (en) | A kind of traffic lights recognizer based on convolutional network | |
| Wu et al. | Adaptive vehicle detector approach for complex environments | |
| CN105930791A (en) | Road traffic sign identification method with multiple-camera integration based on DS evidence theory | |
| CN105844257A (en) | Early warning system based on machine vision driving-in-fog road denoter missing and early warning method | |
| CN109919026B (en) | Surface unmanned ship local path planning method | |
| JP7135665B2 (en) | VEHICLE CONTROL SYSTEM, VEHICLE CONTROL METHOD AND COMPUTER PROGRAM | |
| CN108198417B (en) | A kind of road cruising inspection system based on unmanned plane | |
| CN109902676A (en) | A kind of separated based on dynamic background stops detection algorithm | |
| CN108764108A (en) | A kind of Foregut fermenters method based on Bayesian inference | |
| CN110458050B (en) | Vehicle cut-in detection method and device based on vehicle-mounted video | |
| CN112488083B (en) | Identification method, device and medium of traffic signal lamp based on key point extraction of hetmap | |
| CN107985189A (en) | Towards driver's lane change Deep Early Warning method under scorch environment | |
| CN109635737A (en) | Automobile navigation localization method is assisted based on pavement marker line visual identity | |
| CN108163014A (en) | A kind of engine drivers in locomotive depot Fu Zhu lookout method for early warning and device | |
| CN110738081A (en) | Abnormal road condition detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171024 |
|
| RJ01 | Rejection of invention patent application after publication |