CN105654038A - Vehicle lamp identification method and device - Google Patents
Vehicle lamp identification method and device Download PDFInfo
- Publication number
- CN105654038A CN105654038A CN201510974663.8A CN201510974663A CN105654038A CN 105654038 A CN105654038 A CN 105654038A CN 201510974663 A CN201510974663 A CN 201510974663A CN 105654038 A CN105654038 A CN 105654038A
- Authority
- CN
- China
- Prior art keywords
- lamp
- mark
- sequence
- single lamp
- candidate
- 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.)
- Granted
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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Circuit Arrangement For Electric Light Sources In General (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention provides a vehicle lamp identification method and device. The method comprises the steps that images of the front of a current vehicle are periodically acquired; image identification is performed on each frame of image so that multiple candidate single lamps are acquired; multiple single lamp sequences corresponding to every n adjacent frames of images are calculated through a first decision function so that multiple single lamp identification numbers are acquired; multiple lamp pair identifiers in a lamp pair set corresponding to n frames of images starting from the mth frame of image are classified into multiple lamp pair sequences; and the lamp pair identifiers which exist in a stable tracking lamp pair library and are different from the current frame lamp pair identifiers act as current undetected lamp pair identifiers, and a target lamp pair identifier set is formed by the current frame lamp pair identifiers and the current undetected lamp pair identifiers. Precision of matching identification of the vehicle lamps can be greatly enhanced, and detection precision of front vehicles at night can be enhanced.
Description
Technical field
The embodiment of the present invention relates to technical field of vehicle, particularly relates to a kind of car light recognition methods and device.
Background technology
Along with the development of vehicle technology, vehicle has become people and has gone out or the vehicles indispensable in daily life.
The car light installed on vehicle plays a part very important, not only play the effect of traffic lighting, also having the effect reminding car, prior art identifies the car light of current vehicle front by image recognition technology, by the pairing identification of car light is judged front vehicles.
But when vehicle is when night running, prior art is accurate not to the pairing identification of car light, causes night that the accuracy of detection of front vehicles is lower.
Summary of the invention
The embodiment of the present invention provides a kind of car light recognition methods and device, accurate greatly to improve the pairing identification of car light, it is to increase night is to the accuracy of detection of front vehicles.
An aspect of the embodiment of the present invention is to provide a kind of car light recognition methods, comprising:
Periodically obtain the image of current vehicle front;
Every two field picture carrying out pattern recognition and obtains the single lamp of multiple candidate, the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively;
Knowing with single beacon light that the single lamp of multiple candidate in the first two field picture is corresponding respectively and set up multiple single lamp sequence, a corresponding single lamp sequence known by a single beacon light;
Single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence;
Described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively and obtain multiple single lamp identification number, described multiple single lamp identification number forms the single lamp set of candidate;
The single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, and set is comprised multiple lamp to mark by described lamp;
Mark is set up multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started with the first two field picture is corresponding, and a lamp is to identifying a corresponding lamp to sequence;
Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started by m two field picture is corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2;
When m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse;
To storehouse and it is different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark using being present in described stable tracking lamp, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
Another aspect of the embodiment of the present invention is to provide a kind of car light means of identification, comprising:
Acquisition module, for periodically obtaining the image of current vehicle front;
Picture recognition module, obtains the single lamp of multiple candidate for every two field picture carries out pattern recognition, and the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively;
Single lamp sequence sets up module, knows for single beacon light corresponding respectively with the single lamp of multiple candidate in the first two field picture and sets up multiple single lamp sequence, and a corresponding single lamp sequence known by a single beacon light; Single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence;
Calculating module, obtain multiple single lamp identification number for described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively, described multiple single lamp identification number forms the single lamp set of candidate;
Matching module, the single lamp for being identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, and set is comprised multiple lamp to mark by described lamp;
Sequence is set up module by lamp, and mark is set up multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture for starting with the first two field picture is corresponding, and a lamp is to identifying a corresponding lamp to sequence; Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started by m two field picture is corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2;
Described calculating module is also for when m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse;
Detection module, for to storehouse and being different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark using being present in described stable tracking lamp, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
The car light recognition methods that the embodiment of the present invention provides and device, by periodically obtaining the image of current vehicle front, every two field picture is carried out pattern recognition and obtains the single lamp of multiple candidate, multiple single lamp sequence is set up according to the single lamp of multiple candidate that every two field picture is corresponding, calculated respectively by the first decision function and obtain the single lamp set of candidate, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of candidate carries out pairing and obtains lamp to set, by lamp, mark is referred to multiple lamp to sequence by the multiple lamps in set, sequence is calculated by the 2nd decision function by multiple lamp respectively and obtains stable lamp of following the tracks of to mark, to mark by stable tracking lamp and comprise lamp corresponding to the front n two field picture of present frame set is obtained target lamp to mark set, the pairing identification substantially increasing car light is accurate, improve night to the accuracy of detection of front vehicles.
Accompanying drawing explanation
The car light recognition methods schema that Fig. 1 provides for the embodiment of the present invention;
The image schematic diagram that the car light recognition methods that Fig. 2 provides for the embodiment of the present invention is suitable for;
The structure iron of the car light means of identification that Fig. 3 provides for the embodiment of the present invention;
The structure iron of the car light means of identification that Fig. 4 provides for another embodiment of the present invention.
Embodiment
The car light recognition methods schema that Fig. 1 provides for the embodiment of the present invention; The image schematic diagram that the car light recognition methods that Fig. 2 provides for the embodiment of the present invention is suitable for. The embodiment of the present invention is for when vehicle is when night running, prior art is accurate not to the pairing identification of car light, causes night that the accuracy of detection of front vehicles is lower, it provides car light recognition methods, and the method concrete steps are as follows:
Step S101, the image periodically obtaining current vehicle front;
In embodiments of the present invention, being provided with the vehicle-mounted vidicon demarcated between the windshield glass of vehicle and rear vision mirror, this vehicle-mounted vidicon periodically takes the image of current vehicle front, such as, this vehicle-mounted vidicon takes image from a certain moment, and takes an image every one-period T.
Step S102, single lamp that every two field picture is carried out pattern recognition obtains multiple candidate, the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively;
Described every two field picture is carried out pattern recognition obtain the single lamp of multiple candidate, comprising: every two field picture is carried out pattern recognition and obtains multiple target candidate list lamp; Remove the single lamp of the described multiple candidate of the reflective acquisition of jamming light source and ground in described multiple target candidate list lamp.
According to perspective principle and priori it will be seen that car light can not be had more than disappearance line to occur. And its shape of car light that arbitrary region occurs below fade down line must meet certain requirement. As shown in Figure 2, the upper edge of image 21 is positioned at x-axis, and the left side of image 21 is along being positioned at y-axis, and the upper left corner of image 21 is true origin, and dotted line 22 is disappearance line, if vehicle-mounted vidicon demarcates well, then the distance of disappearance line range image upper edge is known.
The position of the profile according to target candidate list lamp and the relation formula (1) of disappearance line, the jamming light source in multiple target candidate list lamps corresponding to this image can be removed:
Cy(L)<DLy(1)
Wherein, Cy(L) the barycenter ordinate zou position of the profile of target candidate list lamp is represented, DLyFor the ordinate zou position of disappearance line.
Judge further the shape of the profile of target candidate list lamp, the width of the concrete profile judging target candidate list lamp, highly, aspect ratio, area whether meet formula (2)-(5):
Wmin��W(L)��Wmax(2)
Hmin��H(L)��Hmax(3)
HWmin��HW(L)��HWmax(4)
Smin��S(L)��Smax(5)
Wherein, W (L), H (L), HW (L), S (L) are the minimum enclosed rectangle frame of the profile of the target candidate list lamp width in world's system of coordinates respectively, high, depth-width ratio and area; WminAnd WmaxIt is the minimum threshold value and max-thresholds that W (L) is corresponding; HminAnd HmaxIt is the minimum threshold value and max-thresholds that H (L) is corresponding; HWminAnd HWmaxIt is the minimum threshold value and max-thresholds that HW (L) is corresponding; SminAnd SmaxIt is the minimum threshold value and max-thresholds that S (L) is corresponding.
Judging whether target candidate list lamp is that ground is reflective further, according to the position relation between the profile of two target candidate list lamps, first judge two profiles whether on same row, judgment rule is formula (6) or (7):
l(L1)��Cx(L2)��r(L1)(6)
l(L2)��Cx(L1)��r(L2)(7)
Wherein, Cx(L2) and Cx(L1) it is respectively the barycenter X-coordinate of the profile of two target candidate list lamps, l (L1) and r (L1) it is the X-coordinate on the limit, left and right of the minimum enclosed rectangle of the profile of one of them target candidate list lamp, l (L2) and r (L2) it is the X-coordinate on the limit, left and right of the minimum enclosed rectangle of the profile of another target candidate list lamp.
If formula (6) or formula (7) are set up, then whether judgment formula (8)-(10) set up further, if setting up, then L are described2For L1Ground reflective:
W(L1)�ݦ���W(L2)(8)
W(L2)/H(L2)��Thr(9)
S(L1)/S(L2)��Ths(10)
Wherein, the span of �� is [0,1], ThrAnd ThsBe be less than 1 positive constant.
The jamming light source in the plurality of target candidate list lamp and the ground single lamp of the multiple candidate of reflective rear acquisition is removed according to formula (1)-(10).
Step S103, knowing with single beacon light that the single lamp of multiple candidate in the first two field picture is corresponding respectively and set up multiple single lamp sequence, a corresponding single lamp sequence known by a single beacon light;
Such as, the first two field picture comprises the single lamp of 5 candidates, is identified respectively by the single lamp of 5 candidates, and identification number is followed successively by 1,2,3,4,5, sets up single lamp sequence with identification number 1Single lamp sequence is set up with identification number 2Single lamp sequence is set up with identification number 3Single lamp sequence is set up with identification number 4Single lamp sequence is set up with identification number 5
Step S104, single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence;
Reasonable assumption first two field picture is taken in t, takes an image every one-period T, and namely the t+T moment takes the 2nd two field picture, and the t+2T moment takes the 3rd two field picture, analogizes t+ (n-1) the T moment successively and takes the n-th two field picture. From single lamp sequenceWithIn appoint and get a single lamp sequence, such as single lamp sequenceJudge whether the 2nd two field picture exists the single lamp of candidate of the candidate corresponding with identification number 1 single lamp coupling, if the 2nd two field picture exists the single lamp of candidate of the candidate corresponding with identification number 1 single lamp coupling, then in single lamp sequenceMiddle interpolation identification number 1, and the single lamp of candidate of candidate's single lamp coupling corresponding with identification number 1 in single for multiple candidate corresponding for the 2nd two field picture lamp is deleted; If the 2nd two field picture not existing the single lamp of candidate of the candidate corresponding with identification number 1 single lamp coupling, then in single lamp sequenceMiddle interpolation NULL, all single lamp sequences are traveled through according to this method, if after having traveled through all single lamp sequences, 2nd two field picture exists the single lamp of remaining candidate, then set up new single lamp sequence with the identification number of the single lamp of remaining candidate, and a remaining candidate corresponding new single lamp sequence of single lamp, such as, the identification number that there is the single lamp of remaining candidate in 2nd two field picture is 6, then set up single lamp sequenceEvery two field picture with reason, after traversal the 2nd two field picture.
Preferably, the element number in each single lamp sequence maintains 5, when namely there is the 6th element, is deleted by the 1st element from single lamp sequence, such as, after foundation aforesaid method process the 5th two field picture, and single lamp sequenceComprise 1,1, NULL, 1,1, illustrate that the first two field picture, the 2nd two field picture, the 4th two field picture, the 5th two field picture include the single lamp of candidate of identification number 1 correspondence, 3rd two field picture does not comprise the single lamp of candidate of identification number 1 correspondence, when after process the 6th two field picture, if the 6th two field picture comprises the single lamp of candidate of identification number 1 correspondence, then single lamp sequenceIt is updated to 1, NULL, 1,1,1; If the 6th two field picture does not comprise the single lamp of candidate of identification number 1 correspondence, then single lamp sequenceIt is updated to 1, NULL, 1,1, NULL. In all the other single lamp sequences, the renewal of element is with managing in this.
Step S105, described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively and obtain multiple single lamp identification number, described multiple single lamp identification number forms the single lamp set of candidate;
Such as obtain single lamp sequence according to after aforesaid method process the 5th two field picture WithComprise 1,1, NULL, 1,1;Comprise 2, NULL, 2, NULL, 2;Comprise 3,3,3, NULL, 3;Comprise 4, NULL, 4, NULL, NULL;Comprise 5, NULL, 5, NULL, 5;Comprise 6, NULL, 6, NULL; By list lamp sequenceWithInput the first decision function respectivelyIt is single lamp sequence that single lamp is designated id,To each single lamp sequenceThere is an output, for satisfying conditionSingle lamp sequenceExport as id, for not satisfying conditionSingle lamp sequenceExporting as NULL, wherein, �� is constant. Assume single lamp sequenceWithRespectively by the first decision functionObtaining multiple single lamp identification number 1,2,3,4,5,6 after calculating, the plurality of single lamp identification number 1,2,3,4,5,6 forms the single lamp set of candidate corresponding to the 1st frame-5 frame; The single lamp set of candidate corresponding to the 2nd frame-6 frame can be obtained with reason, the single lamp set of candidate that the 3rd frame-7 frame is corresponding, the single lamp set of candidate that the 4th frame-8 frame is corresponding, the single lamp set of candidate that the 5th frame-9 frame is corresponding, the single lamp set of candidate that often adjacent 5 two field pictures are corresponding can be obtained.
Step S106, the single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate carry out pairing and obtain lamp to set, and set is comprised multiple lamp to mark by described lamp;
To multiple single lamp identification number 1 in the single lamp set of the candidate that the 1st frame-5 frame is corresponding, 2, 3, 4, 5, the 6 multiple single lamps identified respectively carry out pairing and obtain first lamp to set, two single beacon lights after pairing are known and are formed a lamp to mark, such as, 1-2, 3-4, 5-6 is respectively the successful lamp of pairing to mark, with reason, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of the candidate that 2nd frame-6 frame is corresponding carries out pairing and obtains the 2nd lamp to set, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of the candidate that 3rd frame-7 frame is corresponding carries out pairing and obtains the 3rd lamp to set, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of the candidate that 4th frame-8 frame is corresponding carries out pairing and obtains the 4th lamp to set, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of the candidate that 5th frame-9 frame is corresponding carries out pairing and obtains the 5th lamp to set.
Mark is set up multiple lamp to sequence by the multiple lamps in set by the lamp that step S107, the n two field picture started with the first two field picture are corresponding, lamp to the corresponding lamp of mark to sequence;
The preferred n of the embodiment of the present invention is 5, lamp corresponding to 5 two field pictures started with the first two field picture to set, the single lamp set of the candidate that namely above-mentioned 1st frame-5 frame is corresponding match after lamp mark 1-2,3-4,5-6 is set up 3 lamps to sequence by the multiple lamps in setWherein,Represent lamp to the lamp being designated 1-2 to sequence,Represent lamp to the lamp being designated 3-4 to sequence,Represent lamp to the lamp being designated 5-6 to sequence.
Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that step S108, the n two field picture started by m two field picture are corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2;
Mark is referred to lamp to sequence by the multiple lamps in set by the 2nd corresponding for the 2nd frame-6 frame lampIn, such as, the 2nd lamp corresponding to the 2nd frame-6 frame to the multiple lamps in set to being designated 3-4,5-6,7-8, thenIn element be 1-2, NULL (due to the 2nd lamp to set in do not have lamp to mark 1-2, soMiddle press-in NULL);In element be 3-4,3-4;In element be 5-6,5-6; Newly-built lamp is to sequenceIn element be 7-8. With reason, mark is referred to lamp to sequence by the multiple lamps in set by the 3rd lamp that the 3rd frame-7 frame is corresponding In, if the 3rd lamp new lamp occurs to mark in set, then with new lamp, mark is set up new lamp to sequence. The 5th lamp gathered, the 5th frame-9 frame is corresponding is managed in this by the 4th lamp that the 4th frame-8 frame is corresponding to the process of set is same.
Step S109, when m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse;
Assume to process the 5th lamp corresponding to the 5th frame-9 frame to, after set, sequence is by all lampsThe element comprised is 1-2,1-2,1-2,1-2,1-2;The element comprised is 3-4,3-4, NULL, NULL, NULL;The element comprised is 5-6,5-6, NULL, 5-6,5-6;The element comprised is 7-8,7-8,7-8,7-8. By lamp to sequenceInput the 2nd decision function respectivelyIt is lamp to the lamp being designated id to sequence,Judge that lamp is to sequenceThe lamp comprised to mark id be whether present frame lamp to mark, specifically will satisfy conditionId as present frame lamp to mark, for not meetingLamp to sequenceExporting as NULL, such as, �� 1=4, then 1-2,5-6,7-8 are all that present frame lamp is to mark (occurring in 5 times being more than or equal to 4 times).
By lamp to sequenceInput the 3rd decision function respectivelyIt is lamp to the lamp being designated id to sequence,Judge that lamp is to sequenceThe lamp comprised, to whether mark id is stablize to follow the tracks of lamp to mark, specifically will satisfy conditionId as stable tracking lamp to mark, for not meetingLamp to sequenceExport as NULL, such as, �� 2=5, then 1-2 stablizes to follow the tracks of lamp to mark (occurring 5 times in 5 times), and by stable tracking lamp to identifying 1-2 stored in stable tracking lamp to storehouse.
When getting the 10th two field picture, the 6th lamp is obtained to set by aforesaid method according to the 6th frame-10 frame, mark is referred to existing lamp to sequence by the multiple lamps in set by the 6th lamp, assume the 6th lamp to the multiple lamps in set to being designated 1-3,5-6,7-8, preferably, the element number that sequence is comprised by each lamp is 5, deletes this lamp to first element in sequence when namely there is the 6th element, then the lamp after upgrading is to sequenceThe element comprised is 1-2,1-2,1-2,1-2, NULL; Lamp after renewal is to sequenceThe element comprised is 3-4, NULL, NULL, NULL, NULL; Lamp after renewal is to sequenceThe element comprised is 5-6, NULL, 5-6,5-6,5-6; Lamp after renewal is to sequenceThe element comprised is 7-8,7-8,7-8,7-8,7-8 and sets up new lamp to sequenceThe element comprised is 1-3.
By lamp to sequenceInput the 2nd decision function respectivelySpecifically will satisfy conditionId as present frame lamp to mark, for not meetingLamp to sequenceExport as NULL, such as, �� 1=1, then 1-2,3-4,5-6,7-8,1-3 are all present frame lamps to mark (occurring in 5 times being more than or equal to 1 time), wherein, know due to a single beacon light and can only belong to a lamp to mark, then 1-2 and 1-3 is that the lamp of contradiction is to mark, therefore selecting the more lamp of occurrence number to being designated final present frame lamp to mark, the number of times that 1-2 occurs is greater than the number of times that 1-3 occurs, then 1-2,3-4,5-6,7-8 are that final present frame lamp is to mark.
By lamp to sequenceInput the 3rd decision function respectivelySpecifically will satisfy conditionId as stable tracking lamp to mark, for not meetingLamp to sequenceExport as NULL, such as, �� 2=5, then 7-8 is for stable lamp of following the tracks of is to mark (occurring 5 times in 5 times), and by stable tracking lamp to identifying 7-8 stored in stable tracking lamp to storehouse, now, stable lamp of following the tracks of is to having stable lamp of following the tracks of in storehouse to identifying 1-2 and 7-8.
When getting the 11st two field picture, the 7th lamp is obtained to set by aforesaid method according to the 7th frame-11 frame, mark is referred to existing lamp to sequence by the multiple lamps in set by the 7th lamp, assume the 7th lamp to the multiple lamps in set to being designated 5-6, preferably, the element number that sequence is comprised by each lamp is 5, deletes this lamp to first element in sequence when namely there is the 6th element, then the lamp after upgrading is to sequenceThe element comprised is 1-2,1-2,1-2, NULL, NULL; Lamp after renewal is to sequenceThe element comprised is NULL, NULL, NULL, NULL, NULL; Lamp after renewal is to sequenceThe element comprised is NULL, 5-6,5-6,5-6,5-6; Lamp after renewal is to sequenceThe element comprised is 7-8,7-8,7-8,7-8, NULL, and the lamp after renewal is to sequenceThe element comprised is 1-3, NULL. Due toElement be NULL entirely, it is preferable that, willDelete, retain lamp to sequence
By lamp to sequenceInput the 2nd decision function respectivelySpecifically will satisfy conditionId as present frame lamp to mark, for not meetingLamp to sequenceExporting as NULL, such as, �� 1=4, then 5-6,7-8 are that present frame lamp is to mark (occurring in 5 times being more than or equal to 4 times).
By lamp to sequenceInput the 3rd decision function respectivelySpecifically will satisfy conditionId as stable tracking lamp to mark, for not meetingLamp to sequenceExporting as NULL, such as, �� 2=5, then without stable lamp of following the tracks of to mark (occurring in 5 times that mark is not existed by the lamp of 5 times), now, stable lamp of following the tracks of is to still having stable lamp of following the tracks of in storehouse to mark 1-2 and 7-8.
Step S110, using be present in described stable tracking lamp to storehouse and be different from described present frame lamp to mark lamp to mark as currently not detecting lamp to mark, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
The image assuming up-to-date acquisition is the 11st two field picture, according to the known 5-6 of above-mentioned steps, 7-8 is that present frame lamp is to mark, stable lamp of following the tracks of is to having stable lamp of following the tracks of in storehouse to mark 1-2 and 7-8, using being present in, described stable tracking lamp to storehouse and is different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark to the embodiment of the present invention, namely 1-2 is not for currently to detect lamp to mark, by present frame lamp to mark 5-6, 7-8 forms target lamp to mark set with the current lamp that do not detect together with mark 1-2, then 1-2, 5-6, 7-8 is target lamp pair, namely the vehicle of correspondence is the target vehicle traced into by this target lamp.
The target vehicle that the embodiment of the present invention obtains according to above-mentioned steps S101-S110 can be applicable in following three scenes: the image 1) periodically being taken current vehicle front by vehicle-mounted vidicon, the obstacle analyzing vehicle front can be known, determine in addition target lamp to after can calculate the distance between current vehicle and target vehicle, for vehicle collision prewarning, 2) tolerance range of the target vehicle detected out by above-mentioned steps S101-S110 is higher, is applicable to current vehicle when night running, analyzes the target vehicle in detection front, it is to increase the security of night running, 3) target vehicle detected out by above-mentioned steps S101-S110, the angle of the headlight of adjustable current vehicle and light intensity, in order to avoid causing dizzy sensation to target vehicle officer. the embodiment of the present invention is by periodically obtaining the image of current vehicle front, every two field picture is carried out pattern recognition and obtains the single lamp of multiple candidate, multiple single lamp sequence is set up according to the single lamp of multiple candidate that every two field picture is corresponding, calculated respectively by the first decision function and obtain the single lamp set of candidate, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of candidate carries out pairing and obtains lamp to set, by lamp, mark is referred to multiple lamp to sequence by the multiple lamps in set, sequence is calculated by the 2nd decision function by multiple lamp respectively and obtains stable lamp of following the tracks of to mark, to mark by stable tracking lamp and comprise lamp corresponding to the front n two field picture of present frame set is obtained target lamp to mark set, the pairing identification substantially increasing car light is accurate, improve night to the accuracy of detection of front vehicles.
On the basis of above-described embodiment, described using be present in described stable tracking lamp to storehouse and be different from described present frame lamp to the lamp of mark to mark as currently not detecting lamp to, after mark, also comprising: judge that the described lamp that currently do not detect is to the whether appearance in described multiple single lamp sequence of two single lamp identification numbers corresponding to mark; If there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL; If described two single lamp identification numbers all do not appear in described multiple single lamp sequence, then mark is deleted mark set by the described lamp that currently do not detect from described target lamp.
1-2 is that undetected lamp is to mark in the above-described embodiments, before judging to comprise present frame, whether 5 two field pictures multiple single lamp sequence that is the 6th frame-10 two field picture is corresponding there are 1 or 2, not occurring if there being one in 1 and 2, then the single beacon light not occurred is known and replace as NULL, such as, there is no appearance 2, then 1-2 is labeled as 1-NULL, now, mark set is comprised lamp to mark 1-NULL by target lamp, 3-4,5-6,7-8;If 1 and 2 all do not occur, then being deleted from stable tracking lamp storehouse by 1-2, now, mark set is only included lamp to mark 3-4,5-6,7-8 by target lamp.
Described method also comprises: judge described stable tracking lamp to each lamp in storehouse to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence; If there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL; If described stable tracking lamp is to occurring new lamp in storehouse to mark, and two single lamp identification numbers that mark is corresponding comprise another single lamp identification number corresponding with NULL by described new lamp, then the lamp comprising NULL replaced mark for described new lamp is to mark.
Before comprising present frame as mentioned above the 5 two field pictures lamp that is the 6th frame-10 two field picture is corresponding to set in multiple lamps to being designated 3-4,5-6,7-8, the stable tracking lamp that storehouse is comprised by stable tracking lamp is to being 1-2 and 7-8, judge that each in storehouse is stablized tracking lamp to whether two single lamp identification numbers that such as 1-2 or 7-8 is corresponding occur in multiple single lamp sequence that the 6th frame-10 two field picture is corresponding by stable lamp of following the tracks of, if there being a single lamp identification number 2 not appear in the plurality of single lamp sequence in 1-2, then 1-2 is labeled as 1-NULL; Follow the tracks of lamp if stable to storehouse occurring new lamp to mark 1-10, then be that new lamp is to identifying 1-10 at a low speed by 1-NULL.
The described single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, comprising: the lamp marking matched acquisition first kind matched by the stable tracking lamp in storehouse with described stable tracking lamp respectively by multiple single lamp identification number in the single lamp set of described candidate is to mark; By in single for described candidate lamp set except single lamp that the single lamp identification number except mark is corresponding is carried out matching the lamp obtaining the 2nd class pairing to mark by the lamp of described first kind pairing, the lamp of described first kind pairing to the lamp of mark and described 2nd class pairing to the mark described lamp of formation to set.
When getting the 10th two field picture, in the single lamp set of the candidate obtained according to the 6th frame-10 frame, multiple single lamp identification number is 3,4,5,6,7,8, follow the tracks of lamp if stable to the stable tracking lamp in storehouse to being designated 1-2,3-4, then in the plurality of single lamp identification number 3,4 can structure in a pair lamp to the lamp of mark 3-4 and first kind pairing to mark; Being judged whether according to formula (11) by corresponding for any two single lamp identification numbers in single lamp identification number 5,6,7,8 remaining in the plurality of single lamp identification number two single lamps can structure lamp pair in a pair:
Low(cx)��cx��High(cx)(11)
Wherein, cx, x=1 ... n is the characteristic item that two single lamps that any two the single lamp identification numbers in single lamp identification number 5,6,7,8 are corresponding are associated, and described characteristic item comprises depth-width ratio, area, position, symmetry and speed etc., Low (cx)��High(cx) it is respectively minimum value and the maximum value of characteristic item.
Assume that the characteristic item that two single lamps of single lamp identification number 5,6 correspondence are associated meets formula (11), the characteristic item that two single lamps of single lamp identification number 5,7 correspondence are associated meets formula (11), then judge that optimum lamp is to mark by formula (12) further:
By the characteristic item c that two single lamps of single lamp identification number 5,6 correspondence are associatedx, x=1 ... n and each cxCorresponding maximum value High (cx) bring formula (12) acquisition F (c intox)|lamppair1, by the characteristic item c that two single lamps of single lamp identification number 5,7 correspondence are associatedx, x=1 ... n and each cxCorresponding maximum value High (cx) bring formula (12) acquisition F (c intox)|lamppair2, compare F (cx)|lamppair1With F (cx)|lamppair2Size, if F (cx)|lamppair1It is less than F (cx)|lamppair2, then single lamp identification number 5,6 can form optimum lamp to mark 5-6.The characteristic item that two single lamps that any two single lamp identification numbers in single lamp identification number 5,6,7,8 are corresponding are associated can be judged according to the method.
The embodiment of the present invention by stable tracking lamp to the renewal in storehouse, detect out the single lamp being blocked, and by the lamp that is blocked to replacing as the new stable tracking lamp pair occurred, further increase accuracy of detection.
The structure iron of the car light means of identification that Fig. 3 provides for the embodiment of the present invention. The car light means of identification that the embodiment of the present invention provides can perform the treatment scheme that car light recognition methods embodiment provides, as shown in Figure 3, car light means of identification 30 comprises acquisition module 31, picture recognition module 32, single lamp sequence set up module 33, calculate module 34, matching module 35, lamp sets up module 36 and detection module 37 to sequence, wherein, acquisition module 31 is for periodically obtaining the image of current vehicle front; For being carried out by every two field picture, pattern recognition obtains the single lamp of multiple candidate to picture recognition module 32, and the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively; Single lamp sequence is set up module 33 and is set up multiple single lamp sequence for single beacon light knowledge corresponding respectively with the single lamp of multiple candidate in the first two field picture, and a corresponding single lamp sequence known by a single beacon light; Single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence; Calculating module 34 and obtain multiple single lamp identification number for described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively, described multiple single lamp identification number forms the single lamp set of candidate; Matching module 35 carries out pairing for the single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate and obtains lamp to set, and set is comprised multiple lamp to mark by described lamp; Sequence is set up module 36, for the lamp that the n two field picture that starts with the first two field picture is corresponding, multiple lamps in set is set up multiple lamp to sequence to mark by lamp, lamp to the corresponding lamp of mark to sequence; Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started by m two field picture is corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2; Calculate module 34 also for when m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse; Detection module 37 is for storehouse and being different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark using being present in described stable tracking lamp, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
The embodiment of the present invention is by periodically obtaining the image of current vehicle front, every two field picture is carried out pattern recognition and obtains the single lamp of multiple candidate, multiple single lamp sequence is set up according to the single lamp of multiple candidate that every two field picture is corresponding, calculated respectively by the first decision function and obtain the single lamp set of candidate, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of candidate carries out pairing and obtains lamp to set, by lamp, mark is referred to multiple lamp to sequence by the multiple lamps in set, sequence is calculated by the 2nd decision function by multiple lamp respectively and obtains stable lamp of following the tracks of to mark, to mark by stable tracking lamp and comprise lamp corresponding to the front n two field picture of present frame set is obtained target lamp to mark set, the pairing identification substantially increasing car light is accurate, improve night to the accuracy of detection of front vehicles.
The structure iron of the car light means of identification that Fig. 4 provides for another embodiment of the present invention. On the basis of above-described embodiment, car light means of identification 30 also comprises and judges module 38, replaces module 39 and removing module 40, wherein, judge that module 38 is for judging the described lamp that currently do not detect is to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence; If a single lamp identification number does not appear in described multiple single lamp sequence for having in described two single lamp identification numbers to replace module 39, then described single lamp identification number is replaced as NULL; If removing module 40 does not all appear in described multiple single lamp sequence for described two single lamp identification numbers, then mark is deleted mark set by the described lamp that currently do not detect from described target lamp.
Judge module 38 also for judging described stable tracking lamp to each lamp in storehouse to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence; If replacing module 39 also for there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL; If described stable tracking lamp is to occurring new lamp in storehouse to mark, and two single lamp identification numbers that mark is corresponding comprise another single lamp identification number corresponding with NULL by described new lamp, then the lamp comprising NULL replaced mark for described new lamp is to mark.
Matching module 35 comprises the first matched sub-block 351 and the 2nd matched sub-block 352, wherein, the first matched sub-block 351 for lamp that the marking matched acquisition first kind is matched by multiple single lamp identification number in the single lamp set of described candidate by stable tracking lamp in storehouse respectively with described stable tracking lamp to mark; 2nd matched sub-block 352 for by single for described candidate lamp set except single lamp that the single lamp identification number except mark is corresponding is carried out matching the lamp obtaining the 2nd class pairing to mark by the lamp of described first kind pairing, the lamp of described first kind pairing to the lamp of mark and described 2nd class pairing to the mark described lamp of formation to set.
Specifically for being carried out by every two field picture, pattern recognition obtains multiple target candidate list lamp to picture recognition module 32; Remove the single lamp of the described multiple candidate of the reflective acquisition of jamming light source and ground in described multiple target candidate list lamp.
The car light means of identification that the embodiment of the present invention provides can specifically for performing the embodiment of the method that above-mentioned Fig. 1 provides, and concrete function repeats no more herein.
The embodiment of the present invention by stable tracking lamp to the renewal in storehouse, detect out the single lamp being blocked, and by the lamp that is blocked to replacing as the new stable tracking lamp pair occurred, further increase accuracy of detection.
In sum, the embodiment of the present invention is by periodically obtaining the image of current vehicle front, every two field picture is carried out pattern recognition and obtains the single lamp of multiple candidate, multiple single lamp sequence is set up according to the single lamp of multiple candidate that every two field picture is corresponding, calculated respectively by the first decision function and obtain the single lamp set of candidate, the multiple single lamp identified respectively by multiple single lamp identification number in the single lamp set of candidate carries out pairing and obtains lamp to set, by lamp, mark is referred to multiple lamp to sequence by the multiple lamps in set, sequence is calculated by the 2nd decision function by multiple lamp respectively and obtains stable lamp of following the tracks of to mark, to mark by stable tracking lamp and comprise lamp corresponding to the front n two field picture of present frame set is obtained target lamp to mark set, the pairing identification substantially increasing car light is accurate, improve night to the accuracy of detection of front vehicles,By to stable tracking lamp to the renewal in storehouse, detect out the single lamp being blocked, and by the lamp that is blocked to replacing as the new stable tracking lamp pair occurred, further increase accuracy of detection.
In provided by the present invention several embodiments, it should be appreciated that, disclosed device and method, it is possible to realize by another way. Such as, device embodiment described above is only schematic, such as, the division of described unit, being only a kind of logic function to divide, actual can have other dividing mode when realizing, such as multiple unit or assembly can in conjunction with or another system can be integrated into, or some features can ignore, or do not perform. Another point, shown or discussed coupling each other or directly coupling or communication connection can be the indirect coupling by some interfaces, device or unit or communication connection, it is possible to be electrical, machinery or other form.
The described unit illustrated as separating component or can may not be and physically separates, and the parts as unit display can be or may not be physical location, namely can be positioned at a place, or can also be distributed on multiple NE. Some or all of unit wherein can be selected according to the actual needs to realize the object of the present embodiment scheme.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, it is also possible to is that the independent physics of each unit exists, it is also possible to two or more unit are in a unit integrated. Above-mentioned integrated unit both can adopt the form of hardware to realize, it is also possible to the form adopting hardware to add software functional unit realizes.
The above-mentioned integrated unit realized with the form of software functional unit, it is possible to be stored in a computer read/write memory medium. Above-mentioned software functional unit is stored in a storage media, comprise some instructions with so that a computer equipment (can be Personal Computer, server, or the network equipment etc.) or treater (processor) perform the part steps of method described in each embodiment of the present invention. And aforesaid storage media comprises: USB flash disk, portable hard drive, read-only storage (Read-OnlyMemory, ROM), random access memory (RandomAccessMemory, RAM), magnetic disc or CD etc. various can be program code stored medium.
Those skilled in the art can be well understood to, for convenience and simplicity of description, only it is illustrated with the division of above-mentioned each function module, in practical application, can complete by different function modules as required and by above-mentioned functions distribution, it is divided into different function modules, to complete all or part of function described above by the internal structure of device. The concrete working process of the device of foregoing description, it is possible to reference to the corresponding process in aforementioned embodiment of the method, do not repeat them here.
Last it is noted that above each embodiment is only in order to illustrate the technical scheme of the present invention, it is not intended to limit; Although with reference to foregoing embodiments to invention has been detailed description, it will be understood by those within the art that: the technical scheme described in foregoing embodiments still can be modified by it, or wherein some or all of technology feature is carried out equivalent replacement; And these amendments or replacement, do not make the scope of the essence disengaging various embodiments of the present invention technical scheme of appropriate technical solution.
Claims (10)
1. a car light recognition methods, it is characterised in that, comprising:
Periodically obtain the image of current vehicle front;
Every two field picture carrying out pattern recognition and obtains the single lamp of multiple candidate, the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively;
Knowing with single beacon light that the single lamp of multiple candidate in the first two field picture is corresponding respectively and set up multiple single lamp sequence, a corresponding single lamp sequence known by a single beacon light;
Single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence;
Described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively and obtain multiple single lamp identification number, described multiple single lamp identification number forms the single lamp set of candidate;
The single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, and set is comprised multiple lamp to mark by described lamp;
Mark is set up multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started with the first two field picture is corresponding, and a lamp is to identifying a corresponding lamp to sequence;
Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started by m two field picture is corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2;
When m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse;
To storehouse and it is different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark using being present in described stable tracking lamp, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
2. method according to claim 1, it is characterised in that, described using be present in described stable tracking lamp to storehouse and be different from described present frame lamp to the lamp of mark to mark as currently not detecting lamp to after identifying, also comprising:
Judge the described lamp that currently do not detect is to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence;
If there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL;
If described two single lamp identification numbers all do not appear in described multiple single lamp sequence, then mark is deleted mark set by the described lamp that currently do not detect from described target lamp.
3. method according to claim 2, it is characterised in that, also comprise:
Judge described stable tracking lamp to each lamp in storehouse to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence;
If there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL;
If described stable tracking lamp is to occurring new lamp in storehouse to mark, and two single lamp identification numbers that mark is corresponding comprise another single lamp identification number corresponding with NULL by described new lamp, then the lamp comprising NULL replaced mark for described new lamp is to mark.
4. method according to claim 3, it is characterised in that, the described single lamp identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, comprising:
The lamp marking matched acquisition first kind matched by stable tracking lamp in storehouse with described stable tracking lamp respectively by multiple single lamp identification number in the single lamp set of described candidate is to mark;
By in single for described candidate lamp set except single lamp that the single lamp identification number except mark is corresponding is carried out matching the lamp obtaining the 2nd class pairing to mark by the lamp of described first kind pairing, the lamp of described first kind pairing to the lamp of mark and described 2nd class pairing to the mark described lamp of formation to set.
5. method according to the arbitrary item of claim 1-4, it is characterised in that, described every two field picture is carried out pattern recognition obtain the single lamp of multiple candidate, comprising:
Every two field picture is carried out pattern recognition and obtains multiple target candidate list lamp;
Remove the single lamp of the described multiple candidate of the reflective acquisition of jamming light source and ground in described multiple target candidate list lamp.
6. a car light means of identification, it is characterised in that, comprising:
Acquisition module, for periodically obtaining the image of current vehicle front;
Picture recognition module, obtains the single lamp of multiple candidate for every two field picture carries out pattern recognition, and the single lamp set of composition candidate known by single beacon light that the single lamp of described multiple candidate is corresponding respectively;
Single lamp sequence sets up module, knows for single beacon light corresponding respectively with the single lamp of multiple candidate in the first two field picture and sets up multiple single lamp sequence, and a corresponding single lamp sequence known by a single beacon light; Single for multiple candidate corresponding for two field picture every after described first two field picture lamp is referred to described multiple single lamp sequence respectively, described single lamp sequence comprises the identification number of the single lamp of multiple same candidate, if the single lamp of described candidate does not belong to the single lamp sequence of any one in described multiple single lamp sequence, then know with single beacon light that the single lamp of described candidate is corresponding and set up new single lamp sequence;
Calculating module, obtain multiple single lamp identification number for described multiple single lamp sequence corresponding for every adjacent n two field picture being calculated by the first decision function respectively, described multiple single lamp identification number forms the single lamp set of candidate;
Matching module, the single lamp for being identified respectively by multiple single lamp identification number in the single lamp set of described candidate carries out pairing and obtains lamp to set, and set is comprised multiple lamp to mark by described lamp;
Sequence is set up module by lamp, and mark is set up multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture for starting with the first two field picture is corresponding, and a lamp is to identifying a corresponding lamp to sequence; Mark is referred to described multiple lamp to sequence by the multiple lamps in set by the lamp that the n two field picture started by m two field picture is corresponding, sequence is comprised the lamp of multiple same lamp pair to mark by described lamp, if mark is comprised new lamp that sequence all do not occur by described multiple lamp to mark by described multiple lamp, then with described new lamp, mark is set up new lamp to sequence, and a new lamp is to identifying a corresponding new lamp to sequence, wherein, m is more than or equal to 2;
Described calculating module is also for when m is more than or equal to n, sequence is calculated by the 2nd decision function by described multiple lamp respectively and obtains present frame lamp to mark or NULL, sequence is obtained stable lamp of following the tracks of to mark or NULL by the 3rd decision function calculating by described multiple lamp respectively, and by described stable tracking lamp to identifying stored in stable tracking lamp storehouse;
Detection module, for to storehouse and being different from described present frame lamp to the lamp of mark to identifying as currently not detecting lamp to mark using being present in described stable tracking lamp, mark is formed target lamp to mark set by mark and the described lamp that currently do not detect by described present frame lamp, described target lamp to the lamp in mark set to lamp corresponding to mark to being target lamp pair.
7. car light means of identification according to claim 6, it is characterised in that, also comprise:
Judge module, for judging the described lamp that currently do not detect is to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence;
Replacing module, if for there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL;
Removing module, if all not appearing in described multiple single lamp sequence for described two single lamp identification numbers, then deletes the described lamp that currently do not detect from described target lamp mark set mark.
8. car light means of identification according to claim 7, it is characterised in that, described judge module also for judging described stable tracking lamp to each lamp in storehouse to whether two single lamp identification numbers that mark is corresponding occur in described multiple single lamp sequence;
If described replacement module is also for there being a single lamp identification number not appear in described multiple single lamp sequence in described two single lamp identification numbers, then described single lamp identification number is replaced as NULL; If described stable tracking lamp is to occurring new lamp in storehouse to mark, and two single lamp identification numbers that mark is corresponding comprise another single lamp identification number corresponding with NULL by described new lamp, then the lamp comprising NULL replaced mark for described new lamp is to mark.
9. car light means of identification according to claim 8, it is characterised in that, described matching module comprises:
First matched sub-block, for the lamp that the marking matched acquisition first kind matched by the stable tracking lamp in storehouse with described stable tracking lamp respectively by multiple single lamp identification number in the single lamp set of described candidate to mark;
2nd matched sub-block, for by single for described candidate lamp set except single lamp that the single lamp identification number except mark is corresponding is carried out matching the lamp obtaining the 2nd class pairing to mark by the lamp of described first kind pairing, the lamp of described first kind pairing to the lamp of mark and described 2nd class pairing to the mark described lamp of formation to set.
10. car light means of identification according to the arbitrary item of claim 6-9, it is characterised in that, specifically for being carried out by every two field picture, pattern recognition obtains multiple target candidate list lamp to described picture recognition module; Remove the single lamp of the described multiple candidate of the reflective acquisition of jamming light source and ground in described multiple target candidate list lamp.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510974663.8A CN105654038B (en) | 2015-12-22 | 2015-12-22 | Car light recognition methods and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510974663.8A CN105654038B (en) | 2015-12-22 | 2015-12-22 | Car light recognition methods and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105654038A true CN105654038A (en) | 2016-06-08 |
| CN105654038B CN105654038B (en) | 2019-03-08 |
Family
ID=56476635
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510974663.8A Active CN105654038B (en) | 2015-12-22 | 2015-12-22 | Car light recognition methods and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105654038B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109612579A (en) * | 2018-12-03 | 2019-04-12 | 青岛讯极科技有限公司 | A kind of system and method detecting dazzle light unlatching and its light type |
| CN110276742A (en) * | 2019-05-07 | 2019-09-24 | 平安科技（深圳）有限公司 | Tail light for train monitoring method, device, terminal and storage medium |
| 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 |
| CN112465910A (en) * | 2020-11-26 | 2021-03-09 | 成都新希望金融信息有限公司 | Target shooting distance obtaining method and device, storage medium and electronic equipment |
| CN117456511A (en) * | 2023-12-25 | 2024-01-26 | 安徽蔚来智驾科技有限公司 | Car light identification method, storage medium and intelligent device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196980A (en) * | 2006-12-25 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for accurately recognizing high speed mobile vehicle mark based on video |
| CN101211408A (en) * | 2006-12-29 | 2008-07-02 | 沈阳东软软件股份有限公司 | Vehicle side image recognition method and apparatus, car lamp error identification detection and driving safety prediction method |
| CN103345766A (en) * | 2013-06-21 | 2013-10-09 | 东软集团股份有限公司 | Method and device for identifying signal light |
-
2015
- 2015-12-22 CN CN201510974663.8A patent/CN105654038B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196980A (en) * | 2006-12-25 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for accurately recognizing high speed mobile vehicle mark based on video |
| CN101211408A (en) * | 2006-12-29 | 2008-07-02 | 沈阳东软软件股份有限公司 | Vehicle side image recognition method and apparatus, car lamp error identification detection and driving safety prediction method |
| CN103345766A (en) * | 2013-06-21 | 2013-10-09 | 东软集团股份有限公司 | Method and device for identifying signal light |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109612579A (en) * | 2018-12-03 | 2019-04-12 | 青岛讯极科技有限公司 | A kind of system and method detecting dazzle light unlatching and its light type |
| CN109612579B (en) * | 2018-12-03 | 2021-02-05 | 青岛讯极科技有限公司 | System and method for detecting opening of automobile high beam and light type of automobile high beam |
| CN110276742A (en) * | 2019-05-07 | 2019-09-24 | 平安科技（深圳）有限公司 | Tail light for train monitoring method, device, terminal and storage medium |
| CN110276742B (en) * | 2019-05-07 | 2023-10-10 | 平安科技（深圳）有限公司 | Train tail lamp monitoring method, device, terminal and storage medium |
| 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 |
| 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 |
| CN112465910A (en) * | 2020-11-26 | 2021-03-09 | 成都新希望金融信息有限公司 | Target shooting distance obtaining method and device, storage medium and electronic equipment |
| CN112465910B (en) * | 2020-11-26 | 2021-12-28 | 成都新希望金融信息有限公司 | Target shooting distance obtaining method and device, storage medium and electronic equipment |
| CN117456511A (en) * | 2023-12-25 | 2024-01-26 | 安徽蔚来智驾科技有限公司 | Car light identification method, storage medium and intelligent device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105654038B (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chang et al. | Argoverse: 3d tracking and forecasting with rich maps | |
| CN105654038A (en) | Vehicle lamp identification method and device | |
| CN111243277B (en) | Commuting vehicle space-time trajectory reconstruction method and system based on license plate recognition data | |
| CN110599570B (en) | Map generation method and system based on crowdsourcing data open scene | |
| CN109242024B (en) | Vehicle behavior similarity calculation method based on checkpoint data | |
| CN111680377B (en) | Traffic situation simulation method, system and electronic equipment | |
| CN104361350A (en) | Traffic sign identification system | |
| CN108460968A (en) | A kind of method and device obtaining traffic information based on car networking | |
| JP2020191081A (en) | Method and device for detecting parking violation, electronic apparatus, and computer readable medium | |
| CN107305736B (en) | License plate recognition intelligent error checking method and system based on license plate rule and space-time reachability | |
| CN103077617A (en) | Pedestrian crosswalk intelligent traffic light monitoring system and method based on computer vision | |
| CN102682304A (en) | Multi-feature integrated passer-by detection method and device | |
| CN105138525A (en) | Traffic video processing device and method, and retrieval device and method | |
| CN107862072B (en) | Method for analyzing vehicle urban-entering fake plate crime based on big data technology | |
| CN104318206A (en) | Barrier detection method and apparatus | |
| CN103050008B (en) | Method for detecting vehicles in night complex traffic videos | |
| CN105608417A (en) | Traffic signal lamp detection method and device | |
| CN109919144A (en) | Drivable region detection method, device, computer storage medium and drive test visual apparatus | |
| CN105718923A (en) | Method for vehicle detection and counting at night based on inverse projection drawings | |
| CN105869421A (en) | Method and device for processing abnormal road condition information and vehicle | |
| CN110705484B (en) | Method for recognizing continuous lane change illegal behaviors by utilizing driving track | |
| CN111735639A (en) | Automatic driving scene minimum set generation method for intelligent networked automobile demonstration area | |
| CN111814751A (en) | Vehicle attribute analysis method and system based on deep learning target detection and image recognition | |
| CN112329553A (en) | Lane line marking method and device | |
| CN115797910A (en) | Long-tail target identification method, computer-readable storage medium and vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |