CN108629238A - A kind of method and apparatus of identification Chinese character label - Google Patents
A kind of method and apparatus of identification Chinese character label Download PDFInfo
- Publication number
- CN108629238A CN108629238A CN201710171393.6A CN201710171393A CN108629238A CN 108629238 A CN108629238 A CN 108629238A CN 201710171393 A CN201710171393 A CN 201710171393A CN 108629238 A CN108629238 A CN 108629238A
- Authority
- CN
- China
- Prior art keywords
- rectangle
- point
- distance
- cluster
- identified
- 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
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
Abstract
This application provides a kind of methods of identification Chinese character label and identification device, wherein recognition methods to include:It to the point cloud data in a track of a road, is clustered according to distance between points, obtains at least one cluster;For each cluster, the rectangle for covering all the points in the cluster is obtained;According to the spatial relationship of the rectangle of each cluster, the rectangle for belonging to the same Chinese character label is obtained.The application is by clustering the point cloud data in track according to distance between points, obtain at least one cluster, and the rectangle of all the points in each point cluster is covered by acquisition and calculates the spatial relationship between each rectangle, to obtain belonging to the rectangle of same Chinese character label, it is corresponding with same Chinese character label to effectively determine the point covered in which rectangle, to which in the case where determining multiple points, effectively the Chinese character label corresponding to these points is identified.
Description
Technical field
This application involves Point Cloud Processing fields, specifically, being related to a kind of method and dress of identification Chinese character label
It sets.
Background technology
In digital map navigation, automatic Pilot is a critical technology, and the key of automatic Pilot is being capable of high-precision
Road environment around ground identification especially identifies the Chinese character label (the Chinese character label as indicated runway) on road, from
And keep automatic Pilot safe and reliable.The technology of generally use laser scanning acquires road laser point cloud in the prior art, to
Identify the corresponding Chinese character label of acquired road laser point cloud.The accuracy that Chinese character label is identified generally for raising,
Before identifying that Chinese character marks, it is thus necessary to determine that the boundary of Chinese character label or frame, in the case where frame is determined, then into
Road laser point cloud in frame is identified as corresponding Chinese character and marked by row identification operation.But it is special due to Hanzi structure
Property, different from figure, there are certain distances between each stroke of Chinese character, and there is presently no can preferably determine Chinese character label
The technology of frame.
Invention content
In view of the above problems, the application provides a kind of method and apparatus of identification Chinese character label, can accurately determine the Chinese
The frame of word mark.
The application the technical solution adopted is that:
According to the method that a kind of identification Chinese character of one embodiment provided by the present application marks, this method includes:To one
Point cloud data in one track of road, is clustered according to distance between points, obtains at least one cluster;Needle
To each cluster, the rectangle for covering all the points in the cluster is obtained;According to the spatial relationship of the rectangle of each cluster, belonged to
In the rectangle of same Chinese character label.
According to the recognition methods that a kind of Chinese character of another embodiment provided by the present application marks, the method is further
Including:To the point cloud data in a track of a road, is clustered, obtained at least according to distance between points
One cluster;For each cluster, the rectangle for covering all the points in the cluster is obtained;According to the space of the rectangle of each cluster
Relationship obtains the rectangle for belonging to the same label;The rectangle for belonging to the same label is merged, the label pair is obtained
The target rectangle answered, the target rectangle cover all rectangles for belonging to the same label;According to target rectangle
Size and its position in track, judge target rectangle whether belong to Chinese character label, if so, by the target rectangle mark
It is denoted as the rectangle of Chinese character label.
According to the device that a kind of identification Chinese character of one embodiment provided by the present application marks, which includes:Cluster is single
Member is clustered according to distance between points for the point cloud data in a track of a road, obtain to
A few cluster;First acquisition unit, for for each point cluster, obtaining the rectangle for covering all the points in the cluster;Second obtains
Unit is taken, for the spatial relationship according to each rectangle for putting cluster, obtains the rectangle for belonging to the same Chinese character label.
According to the identification device that a kind of Chinese character of another embodiment provided by the present application marks, the identification device packet
It includes:For the point cloud data in a track to a road, clustered according to distance between points, obtain to
The unit of a few cluster;For being directed to each point cluster, the unit for covering the rectangle of all the points in the cluster is obtained;For root
According to the spatial relationship of the rectangle of each cluster, the unit for belonging to the rectangle of the same label is obtained;For the same mark will to be belonged to
The rectangle of note merges, and obtains the corresponding target rectangle of the label, and the target rectangle, which covers, described belongs to same
The unit of all rectangles of one label;For the size according to target rectangle and its position in track, target square is judged
Whether shape belongs to Chinese character label, if so, by the target rectangle labeled as the unit of the rectangle of Chinese character label.
Compared with prior art, the application has the following advantages:
The application compared to the prior art, by the point cloud data in track according between points distance carry out
Cluster obtains at least one cluster, and covers the rectangle of all the points in each point cluster by acquisition and calculate the sky between each rectangle
Between relationship, the purpose for getting the rectangle for belonging to same Chinese character label is realized, to realize that quick identify of Chinese character label provides
Accurate basic data.
Description of the drawings
Fig. 1 is a kind of flow chart of the embodiment of the method for identification Chinese character label provided by the present application;
Fig. 2 is a kind of flow chart of the embodiment of the method for identification Chinese character label of another embodiment provided by the present application;
Fig. 3 is the flow chart further described to the step S101 in the application Fig. 1;
Fig. 4 is the flow chart further described to the step S102 in the application Fig. 1;
Fig. 5 is the legend illustrated to the processing procedure of the application step S102;
Fig. 6 is the flow chart further described to the step S103 in the application Fig. 1;
Fig. 7 is the schematic diagram of the rectangle of the covering point cluster of the application one embodiment;
Fig. 8 is a kind of flow chart of the embodiment of the method for identification Chinese character label of another embodiment provided by the present application;
Fig. 9 is the schematic diagram that the Chinese character of the application one embodiment marks a corresponding target rectangle;
Figure 10 is a kind of schematic diagram of the device embodiment of identification Chinese character label provided by the present application;
Figure 11 is the schematic diagram of the device embodiment of another identification Chinese character label provided by the present application.
Specific implementation mode
Many details are elaborated in the following description in order to fully understand the application.But the application can
Much to implement different from other manner described here, those skilled in the art can be without prejudice to the application intension
In the case of do similar popularization, therefore the application is not limited by following public specific implementation.
This application provides a kind of method and apparatus of identification Chinese character label, in turn below in conjunction with attached drawing to the application's
Embodiment is described in detail.
One road, especially common super expressway, usually may include multiple tracks, and each track is general
Limited by two adjacent lane lines, usually had in each track some label (including arrow mark, Chinese character label
Deng).These labels are for human eye, it is easy to identify, but to be made as using for driving procedure or navigation procedure
Data, need the technology by laser scanning first to acquire road laser point cloud (or " point cloud data "), then, be based on road
Road laser point cloud identifies the acquired corresponding label of road laser point cloud.
A kind of Chinese character mark recognition method that the application proposes, it is therefore an objective to road laser point cloud is based on, to Chinese character label
Frame is accurately identified, subsequently to identify that the particular content of Chinese character label provides basic data.Wherein, point cloud data is logical
Often including the position of the point of magnanimity, reflectance value etc. data on acquired road.
As previously mentioned, some labels are usually had in each track, including arrow mark, Chinese character label etc., actually answering
In, collected point cloud data can be pre-processed, for example, the noise data in point cloud data is filtered, with
And the point cloud data of (such as arrow mark) is marked to preserve respectively the point cloud data of the Chinese character label in a track and non-Chinese character.
If the point cloud data of Chinese character label and the point cloud data of non-Chinese character label preserve respectively, the application may be used and carry
The method that Chinese character label is identified shown in Fig. 1 of confession.If Chinese character marks and the point cloud data of non-Chinese character label still preserves
Together, then method shown in Fig. 2 can be used.
Referring to FIG. 1, it is a kind of flow chart of the method for identification Chinese character label provided by the embodiments of the present application, it is described
The method of identification Chinese character label includes the following steps:
Step S101:To the point cloud data in a track of a road, gathered according to distance between points
Class obtains at least one cluster.
Wherein, the point cloud data in step S101 is the point cloud data for meeting Chinese character marker characteristic.
Step S102:For each cluster, the rectangle for covering all the points in the cluster is obtained.
Step S103:According to the spatial relationship of the rectangle of each cluster, the rectangle for belonging to the same Chinese character label is obtained.
Scheme shown in Fig. 2, the method includes:
Step S201:To the point cloud data in a track of a road, gathered according to distance between points
Class obtains at least one cluster.
Step S202:For each cluster, the rectangle for covering all the points in the cluster is obtained.
Step S203:According to the spatial relationship of the rectangle of each cluster, the rectangle for belonging to the same label is obtained.
Step 204:The rectangle for belonging to the same label is merged, the corresponding target square of the label is obtained
Shape, the target rectangle cover the corresponding all rectangles of the same label;
Step 205:According to the size of target rectangle and its position in track, judge whether target rectangle belongs to the Chinese
Word mark, if so, the target rectangle to be labeled as to the rectangle of Chinese character label.
In general, the target rectangle for belonging to Chinese character label is located in track between two adjacent lane lines and target rectangle
Center to two adjacent lane lines distance it is also of substantially equal, in addition, Chinese character label be usually dimensionally larger than other marks
Note, that is, belong to Chinese character label rectangle on length, width or/and area also greater than other label target rectangle, therefore,
By calculating length, width or/and the area of target rectangle and the numerical value of calculating and preset respective threshold can be compared
Compared with if it is greater than preset respective threshold, then position of the target rectangle in track further being obtained, if the position of the acquisition
It sets and meets predeterminated position condition, then the target rectangle is labeled as to the rectangle of Chinese character label.
Above, step S101, step S102, step S103 respectively with step S201, step S202, step S203 phases
It is same or essentially identical, in addition, the step S104 and above-mentioned steps S204 that are described below are essentially identical, therefore, to step S201
Understanding to step S204 can refer to step S101 to step S104, no longer separately state herein.
First, step S101 or step S201 are described.
According to one embodiment of the application, cluster is put in order to obtain, according in track of distance pair between points
Point cloud data clustered, to which the point cloud data in a track is divided into multiple clusters.Specifically, figure is please referred to
3, the point cloud data in a track to a road is clustered according to distance between points, is obtained at least
The step S101 of one cluster may include:
Step S301 arbitrarily chooses a point as to be identified from the point cloud data in the same track of a road
Point.
For example, point cloud data includes the corresponding positional number of multiple points from number 1 to N (N is more than 1 positive integer)
According to, it is assumed that it chooses number 1 and is used as point to be identified.
Step S302 obtains the distance of left point in the point to the point cloud data to be identified.
For example, after above-mentioned steps S301 chooses the point of number 1 as point to be identified, according to the positional number of the point of number 1
According to the position data with left point in point cloud data, obtain the number 1 point arrive respectively number 2,3 ... the point of N-1, N away from
From.
Step S303 obtains the number of the point at a distance from point to be identified less than preset first distance threshold.
Assuming that the first distance threshold is M, and pass through statistics, obtains small at a distance from point to be identified in above-mentioned steps S302
In M point number be g.
Whether the number of step S304, the point that judgment step S303 is obtained are more than preset number threshold value, if the number
More than preset number threshold value, then enters step S305 and be less than pre-determined distance threshold value by the point to be identified and to its distance
Point is divided into same cluster, and otherwise, return to step S301 arbitrarily chooses a step of point is as to be identified.
Assuming that number threshold value is G, if the number g that above-mentioned steps S303 is obtained is more than number threshold value G, by number 1
The point of point to be identified and distance numbers 1 less than M is divided into same cluster;Otherwise, continue return to step S301 and arbitrarily choose one
The step of a point is as to be identified.
Step S306 deletes the point for being divided into a cluster from the point cloud data.
Step S307 judges whether also have remaining point in the point cloud data, if after deleting, in the point cloud data
Also remaining point, then return to step S301 is arbitrary chooses a step of point is as to be identified, if after deleting, the point
Cloud data are sky, then terminate flow, until the point cloud data is sky.
If number 1 and the point to the distance of number 1 less than M have been divided into same cluster, from the point cloud data
The middle point for deleting number 1 and be less than M to the distance of number 1, if also having remaining point (to arrive number 1 in point cloud data at this time
The number g of point of the distance less than M be less than number sum N), then return to step S301 is executed in the arbitrary remaining point of selection
One point repeats above-mentioned steps S301 to step S305 as point to be identified, until all in all point cloud datas
Point is all divided into corresponding cluster, terminates flow.
The purpose of step S101 is that the point for belonging to the same label condenses together, and by observation, inventor has found,
Belong to same label point can agglomerating appearance, therefore, in practical applications, preset first distance threshold should meet
The distance for belonging to the same label can be filtered out, the occurrence of the first distance threshold can according to circumstances be set by technical staff
It is fixed.Meanwhile in the point of agglomerating appearance, there are one the point for being in center, its peripheries to have gathered a certain number of points for meeting, because
This, the purpose of step S301-S306 is the point found out in center and its point around gathered.
Next, obtaining the rectangle for covering all the points in the cluster for each point cluster to step S102 or step S202
It is described.
Specifically, it please refers to Fig.4 and Fig. 5, the step S102 or step S202 may include:
Step S401, chooses a lane line in the track, its starting point is denoted as point A.
For example, choosing a lane line on right side in Fig. 5, and the starting point of the lane line is denoted as point A, to mark at this
The small circle of lane line indicates.
Step S402 does vertical line by each being put to the lane line in cluster, obtains intersection point of each point on the lane line
Point.
For example, the lane line of a point to the right in Fig. 5 does vertical line, this hanging down on the lane line is obtained
Foot point C.
Step S403 calculates the distance of each intersection point point-to-point A, and it is corresponding to obtain wherein minimum and maximum distance
Vertical line equation L1 and L2.
By calculating the distance of all intersection point point-to-point A in Fig. 5, obtain shown in fig. 5 apart from minimum and maximum, vertical
Directly in vertical line the equation L1 and L2 of lane line.
Step S404, an optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will join
It examines vertical line equation and the intersection point of the lane line is denoted as point B.
Such as the vertical line equation L1 chosen in Fig. 5 is used as with reference to vertical line equation, and the intersection point of L1 and the lane line are denoted as
Point B.
Step S405 is obtained each to put in described cluster and is arrived the reference intersection point point with reference to vertical line equation.
For example, getting a point in Fig. 5 to the reference intersection point point D with reference to vertical line equation L1.
Step S406 calculates the distance for each referring to intersection point point-to-point B, and obtains wherein minimum and maximum distance point
Not corresponding vertical line equation L4 and L3.
For example, calculating with reference to intersection point point D to the distance of point B, the i.e. length of line segment DB.Intersection point is each referred to by calculating
The distance of point-to-point B is obtained apart from minimum and maximum vertical line equation L4 and L3.
Step S407 obtains the intersection point of four vertical line equations L1, L2, L3 and L4, and the intersection point, which is constituted, to be covered in the cluster
The rectangle of all the points.
As shown in figure 5, the intersection point of vertical line equation L1, L2, L3 and L4 include V1, V2, V3 and V4, this four point compositions are covered
Cover the rectangle of all the points in the cluster.
The purpose of step S102 is to be formed by a cluster for the point to condense together, obtains covering each point cluster
Rectangle, since each cluster is usually with a stroke, a radical in label etc. there are correspondence, obtain
The rectangle of covering point cluster, is to obtain profile a part of in overlay marks.
Next, to step S103 according to the spatial relationship of the rectangle of each point cluster, obtain belonging to the same Chinese character label
Rectangle be described.And step S203 is obtained belonging to the same label according to the spatial relationship of the rectangle of each point cluster
Rectangle, the hereafter description to step S103 can be referred to substantially.
Wherein, the spatial relationship of the rectangle include but not limited to the intersection of rectangle, overlapping or mutually from etc..
Specifically, Fig. 6 and Fig. 7 are please referred to, the step S103 may include:
Step S501 arbitrarily chooses a rectangle as rectangle to be identified from the corresponding rectangle of all the points cluster.
As shown in fig. 7, showing 4 rectangles in Fig. 7, each rectangle covers all the points in corresponding points cluster, can be arbitrary
One of rectangle is chosen as rectangle to be identified, such as chooses rectangle 1 and is used as rectangle to be identified.
Step S502 obtains rectangle to be identified at a distance from the rectangle in the rectangle in addition to rectangle to be identified.
Refer to a distance minimum in the distance between all vertex of two rectangles at a distance from the rectangle and rectangle,
That is, calculating the distance to all vertex of another rectangle with each vertex of a rectangle, 16 can be found out out altogether
Distance, a minimum distance is then the distance between the two rectangles in this 16 distances.Still referring to FIG. 7, for example getting
Rectangle 1 is denoted as w12 at a distance from rectangle 2;Rectangle 1 is got at a distance from rectangle 3, is denoted as w13;Get rectangle 1 and square
The distance of shape 4, is denoted as w14.
Step S503 judges whether there is the distance less than preset second distance threshold value, if so, then by described apart from right
The rectangle answered is determined as target rectangle.
Assuming that in the above-mentioned distance got, w12 and w13 are less than preset second distance threshold value, then by rectangle 2 and square
Shape 3 is determined as target rectangle.
Step S504 obtains target rectangle and removes rectangle to be identified and institute in the rectangle for each target rectangle
There is the distance of the rectangle other than target rectangle.
Step S505 judges whether there is the distance less than preset second distance threshold value, if without be less than preset second away from
With a distance from threshold value, then S506 is entered step;If there is the distance less than preset second distance threshold value, S507 is entered step.
The rectangle to be identified and target rectangle are marked labeled as the same Chinese character is belonged to, and execute step by step S506
Rapid S508 judges whether also have not labeled rectangle in the rectangle, is waited for if so, then executing arbitrary one rectangle of selection and being used as
The step S501 of identification rectangle terminates flow if nothing.
Step S507 is also determined as target rectangle apart from corresponding rectangle by described, executes and be directed to each target square
Shape obtains the step of target rectangle is at a distance from the rectangle in the rectangle in addition to rectangle to be identified and all target rectangles
S504。
For example, specifically to above-mentioned rectangle 2, the distance between rectangle 2 and remaining rectangle 4 are obtained, w24 is denoted as, sentences
Whether disconnected w24 is less than preset second distance threshold value.Similarly, for above-mentioned rectangle 3, obtain rectangle 3 and remaining rectangle 4 it
Between distance, be denoted as w34, and judge whether w34 is less than preset second distance threshold value, if w24 and w34 be all not less than it is pre-
If second distance threshold value, then enter step S505;If w24 and w34 at least one be less than preset second distance threshold
Value, then enter step S506.For the previous case, w24 and w34 are not less than preset second distance threshold value, then enter
To step S505, rectangle 1 to be identified and rectangle 2 and rectangle 3 are marked labeled as the same Chinese character is belonged to, and judge rectangle
In whether also have not labeled rectangle, if so, then execute the step of one rectangle of arbitrary selection is as rectangle to be identified, if
Nothing then terminates flow, due to there is rectangle 4 without labeled, a rectangle is arbitrarily chosen as rectangle to be identified
Step, and continue to execute step S501 to corresponding steps later, until rectangle 4 is labeled namely all rectangles are labeled.
For latter situation, it is assumed that w24 and w34 is both less than preset second distance threshold value, then enters step S506, by rectangle 4
Also it is determined as target rectangle, and returns to step S504 and execute for each target rectangle, obtains target rectangle and the rectangle
In the distance of rectangle in addition to rectangle to be identified and all target rectangles terminate stream until all rectangles have all been labeled
Journey.
The purpose of step S103 is to obtain belonging to the same Chinese character label according to the spatial relationship of the rectangle of each point cluster
Rectangle.For the rectangle of each cluster, if distance is closer between rectangle and rectangle, these usual rectangles belong to same
The possibility of a label or Chinese character label is bigger, this is because for belonging to the stroke of same label itself with respect to other strokes,
Distance is usually close very much.
Optionally, Fig. 7 and Fig. 8 and Fig. 9 are please referred to, Fig. 1 the methods further comprise:
Step S104 merges the rectangle for belonging to the same Chinese character label, obtains the Chinese character label corresponding one
A target rectangle, the target rectangle cover all rectangles for belonging to the same Chinese character label.
For example, four rectangles for belonging to the same Chinese character label in Fig. 7 are merged, to obtain square shown in Fig. 9
Shape, i.e., the described Chinese character mark a corresponding target rectangle, and it is corresponding which covers the same Chinese character label
All rectangles.
The purpose of step S104 is to merge the rectangle for belonging to the same Chinese character label, obtains the Chinese character label
A corresponding target rectangle, in the case where determining the rectangle comprising Chinese character label, can more accurately identify side
Chinese character label in frame.
In addition, following sub-steps included by step S201 to step S203 for above-mentioned scheme shown in Fig. 2, it can
To refer to the description to the sub-step included by step S101 to step S103 above, hereafter repeat no more.
Optionally, the point cloud data in a track to a road is carried out according to distance between points
Cluster, the step S201 for obtaining at least one cluster are specifically included:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance by the point to be identified and to its distance
The point of threshold value is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having residue in the point cloud data
Point, then execute the step of one point of arbitrary selection is as to be identified, until the point cloud data be sky.
Optionally, described for each point cluster, the step S202 that acquisition covers the rectangle of all the points in the cluster is specifically wrapped
It includes:
A lane line for choosing the track, point A is denoted as by its starting point;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation of wherein minimum and maximum distance
L1 and L2;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line side
Journey and the intersection point of the lane line are denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and it is corresponding to obtain wherein minimum and maximum distance
Vertical line equation L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point, which is constituted, covers all the points in the cluster
Rectangle.
Optionally, the basis each puts the spatial relationship of the rectangle of cluster, obtains the rectangle step for belonging to the same label
S203 is specifically included:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, then by described apart from corresponding rectangle
It is determined as target rectangle;
It is directed to each target rectangle, obtain target rectangle and removes rectangle to be identified and all target squares in the rectangle
The distance of rectangle other than shape judges whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, the rectangle to be identified and target rectangle are labeled as
Belong to the same label, judges whether also have not labeled rectangle in the rectangle, a rectangle is chosen if so, then executing
The step of as rectangle to be identified, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target apart from corresponding rectangle by described
Rectangle executes and is directed to each target rectangle, obtains target rectangle and removes rectangle to be identified and all target squares in the rectangle
Rectangle other than shape apart from the step of.In the above-described embodiment, a kind of method of identification Chinese character label is provided, therewith
Corresponding, the application also provides a kind of device of identification Chinese character label.Referring to FIG. 10, it is one kind provided by the invention
Identify the schematic diagram of the device embodiment of Chinese character label.Since device embodiment is substantially similar to embodiment of the method, so description
Must be fairly simple, the relevent part can refer to the partial explaination of embodiments of method.Device embodiment described below is only
Schematically.
As described above, if the point cloud data of Chinese character label and the point cloud data of non-Chinese character label preserve respectively, ask
With reference to figure 10, a kind of device of identification Chinese character label provided in this embodiment, including:
Cluster cell 101, for the point cloud data in a track of a road, according between points away from
From being clustered, at least one cluster is obtained;
First acquisition unit 102, for for each point cluster, obtaining the rectangle for covering all the points in the cluster;
Second acquisition unit 103 obtains belonging to the same Chinese character for the spatial relationship according to each rectangle for putting cluster
The rectangle of label.
In one embodiment of the application, the device of the application further comprises:
Rectangle combining unit obtains the Chinese character label for merging the rectangle for belonging to the same Chinese character label
A corresponding target rectangle, the target rectangle cover all rectangles for belonging to the same Chinese character label.
In one embodiment of the application, cluster cell 101 is specifically used for:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance by the point to be identified and to its distance
The point of threshold value is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having residue in the point cloud data
Point, then execute the step of one point of arbitrary selection is as to be identified, until the point cloud data be sky.
In one embodiment of the application, second acquisition unit 103 is specifically used for:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, then by described apart from corresponding rectangle
It is determined as target rectangle;
It is directed to each target rectangle, obtain target rectangle and removes rectangle to be identified and all target squares in the rectangle
The distance of rectangle other than shape judges whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, the rectangle to be identified and target rectangle are labeled as
Belong to the same Chinese character label, judges whether also have not labeled rectangle in the rectangle, if so, then executing selection one
The step of rectangle is as rectangle to be identified terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target apart from corresponding rectangle by described
Rectangle executes and is directed to each target rectangle, obtains target rectangle and removes rectangle to be identified and all target squares in the rectangle
Rectangle other than shape apart from the step of.
In one embodiment of the application, first acquisition unit 102 is specifically used for:
A lane line for choosing the track, point A is denoted as by its starting point;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation of wherein minimum and maximum distance
L1 and L2;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line side
Journey and the intersection point of the lane line are denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and it is corresponding to obtain wherein minimum and maximum distance
Vertical line equation L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point, which is constituted, covers all the points in the cluster
Rectangle.
According to another embodiment of the application, the device of another identification Chinese character label is also provided.1 is please referred to Fig.1,
As described above, if the point cloud data of Chinese character label and non-Chinese character label is still saved together, described device includes:
For the point cloud data in a track to a road, is clustered, obtained according to distance between points
To the unit 201 of at least one cluster;
For being directed to each point cluster, the unit 202 for covering the rectangle of all the points in the cluster is obtained;
For the spatial relationship according to each rectangle for putting cluster, the unit 203 for belonging to the rectangle of the same label is obtained;
For merging the rectangle for belonging to the same label, the corresponding target rectangle of the label, institute are obtained
State the unit 204 that target rectangle covers all rectangles for belonging to the same label;
For the size according to target rectangle and its position in track, judge whether target rectangle belongs to Chinese character mark
Note, if so, by the target rectangle labeled as the unit 205 of the rectangle of Chinese character label.
Optionally, described to be used for the point cloud data in a track of a road, according to distance between points
It is clustered, the unit 201 for obtaining at least one cluster is specifically used for:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance by the point to be identified and to its distance
The point of threshold value is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having residue in the point cloud data
Point, then execute the step of one point of arbitrary selection is as to be identified, until the point cloud data be sky.
Optionally, described for for each point cluster, obtaining the unit 202 for covering the rectangle of all the points in the cluster and having
Body is used for:
A lane line for choosing the track, point A is denoted as by its starting point;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation of wherein minimum and maximum distance
L1 and L2;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line side
Journey and the intersection point of the lane line are denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and it is corresponding to obtain wherein minimum and maximum distance
Vertical line equation L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point, which is constituted, covers all the points in the cluster
Rectangle.
Optionally, the spatial relationship for according to each rectangle for putting cluster, obtains the rectangle for belonging to the same label
Unit 203 be specifically used for:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, then by described apart from corresponding rectangle
It is determined as target rectangle;
It is directed to each target rectangle, obtain target rectangle and removes rectangle to be identified and all target squares in the rectangle
The distance of rectangle other than shape judges whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, the rectangle to be identified and target rectangle are labeled as
Belong to the same label, judges whether also have not labeled rectangle in the rectangle, a rectangle is chosen if so, then executing
The step of as rectangle to be identified, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target apart from corresponding rectangle by described
Rectangle executes and is directed to each target rectangle, obtains target rectangle and removes rectangle to be identified and all target squares in the rectangle
Rectangle other than shape apart from the step of.
More than, for the embodiment of the device of identification Chinese character label provided by the invention.
Although the application is disclosed as above with preferred embodiment, it is not for limiting the application, any this field skill
Art personnel are not departing from spirit and scope, can make possible variation and modification, therefore the guarantor of the application
Shield range should be subject to the range that the application claim defined.
In a typical configuration, computing device include one or more processors (CPU), input/output interface,
Network interface and memory.
Memory may include computer-readable medium in volatile memory, random access memory (RAM) and/
Or the forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
1, computer-readable medium can be by any including permanent and non-permanent, removable and non-removable media
Method or technique realizes information storage.Information can be computer-readable instruction, data structure, the module of program or other
Data.The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM
(SRAM), dynamic random access memory (DRAM), other kinds of random access memory (RAM), read-only memory
(ROM), electrically erasable programmable read-only memory (EEPROM), fast flash memory bank or other memory techniques, CD-ROM
Read memory (CD-ROM), digital versatile disc (DVD) or other optical storages, magnetic tape cassette, tape magnetic disk storage
Or other magnetic storage apparatus or any other non-transmission medium, it can be used for storing and can be accessed by a computing device information.It presses
It is defined according to herein, computer-readable medium does not include non-temporary computer readable media (transitory media), is such as adjusted
The data-signal and carrier wave of system.
2, it will be understood by those skilled in the art that embodiments herein can be provided as method, system or computer program production
Product.Therefore, complete hardware embodiment, complete software embodiment or implementation combining software and hardware aspects can be used in the application
The form of example.Moreover, can be used can in the computer that one or more wherein includes computer usable program code by the application
With the computer program product implemented on storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Form.
Claims (18)
1. a kind of method of identification Chinese character label, which is characterized in that including:
To the point cloud data in a track of a road, is clustered according to distance between points, obtain at least one
A cluster;
For each cluster, the rectangle for covering all the points in the cluster is obtained;
According to the spatial relationship of the rectangle of each cluster, the rectangle for belonging to the same Chinese character label is obtained.
2. according to the method described in claim 1, it is characterized in that, the method further includes:
The rectangle for belonging to the same Chinese character label is merged, the Chinese character is obtained and marks a corresponding target rectangle, institute
It states target rectangle and covers all rectangles for belonging to the same Chinese character label.
3. method according to claim 1 or 2, which is characterized in that the point cloud in the same track to a road
Data are clustered according to distance between points, are obtained at least one cluster and are specifically included:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance threshold value by the point to be identified and to its distance
Point is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having remaining point in the point cloud data,
The step of one point of arbitrary selection is as to be identified is then executed, until the point cloud data is sky.
4. according to the method described in claim 3, it is characterized in that, the basis each puts the spatial relationship of the rectangle of cluster, obtain
It is specifically included to the rectangle for belonging to the same Chinese character label:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, being then determined as described apart from corresponding rectangle
Target rectangle;
For each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle distance, judge whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, it is labeled as the rectangle to be identified and target rectangle to belong to same
One Chinese character label judges whether also have not labeled rectangle in the rectangle, a rectangle conduct is chosen if so, then executing
The step of rectangle to be identified, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target rectangle apart from corresponding rectangle by described,
It executes and is directed to each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle apart from the step of.
5. according to the method described in claim 3, it is characterized in that, described for each point cluster, acquisition covers institute in the cluster
Rectangle a little specifically includes:
Its starting point is denoted as point A by a lane line for choosing the track;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation L1 and L2 of wherein minimum and maximum distance;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line equation and institute
The intersection point for stating lane line is denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and obtains the corresponding vertical line side of wherein minimum and maximum distance
Journey L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point constitutes the rectangle for covering all the points in the cluster.
6. a kind of recognition methods of Chinese character label, which is characterized in that the method further includes:
To the point cloud data in a track of a road, is clustered according to distance between points, obtain at least one
A cluster;
For each cluster, the rectangle for covering all the points in the cluster is obtained;
According to the spatial relationship of the rectangle of each cluster, the rectangle for belonging to the same label is obtained;
The rectangle for belonging to the same label is merged, the corresponding target rectangle of the label, the target square are obtained
Shape covers all rectangles for belonging to the same label;
According to the size of target rectangle and its position in track, judge whether target rectangle belongs to Chinese character label, if so,
The target rectangle is labeled as to the rectangle of Chinese character label.
7. according to the method described in claim 6, it is characterized in that, point cloud number in a track to a road
According to being clustered according to distance between points, obtain at least one cluster and specifically include:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance threshold value by the point to be identified and to its distance
Point is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having remaining point in the point cloud data,
The step of one point of arbitrary selection is as to be identified is then executed, until the point cloud data is sky.
8. the method according to the description of claim 7 is characterized in that described for each point cluster, acquisition covers institute in the cluster
Rectangle a little specifically includes:
Its starting point is denoted as point A by a lane line for choosing the track;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation L1 and L2 of wherein minimum and maximum distance;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line equation and institute
The intersection point for stating lane line is denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and obtains the corresponding vertical line side of wherein minimum and maximum distance
Journey L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point constitutes the rectangle for covering all the points in the cluster.
9. the method according to the description of claim 7 is characterized in that the basis each puts the spatial relationship of the rectangle of cluster, obtain
It is specifically included to the rectangle for belonging to the same label:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, being then determined as described apart from corresponding rectangle
Target rectangle;
For each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle distance, judge whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, it is labeled as the rectangle to be identified and target rectangle to belong to same
One label judges whether also have not labeled rectangle in the rectangle, waits knowing if so, then executing one rectangle of selection and being used as
The step of other rectangle, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target rectangle apart from corresponding rectangle by described,
It executes and is directed to each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle apart from the step of.
10. a kind of device of identification Chinese character label, which is characterized in that including:
Cluster cell, for the point cloud data in a track of a road, being gathered according to distance between points
Class obtains at least one cluster;
First acquisition unit, for for each point cluster, obtaining the rectangle for covering all the points in the cluster;
Second acquisition unit obtains the square for belonging to the same Chinese character label for the spatial relationship according to each rectangle for putting cluster
Shape.
11. device according to claim 10, which is characterized in that described device further comprises:
Rectangle combining unit obtains the Chinese character label and corresponds to for merging the rectangle for belonging to the same Chinese character label
A target rectangle, the target rectangle covers all rectangles for belonging to the same Chinese character label.
12. the device according to claim 10 or 11, which is characterized in that the cluster cell is specifically used for:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance threshold value by the point to be identified and to its distance
Point is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having remaining point in the point cloud data,
The step of one point of arbitrary selection is as to be identified is then executed, until the point cloud data is sky.
13. device according to claim 12, which is characterized in that the second acquisition unit is specifically used for:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, being then determined as described apart from corresponding rectangle
Target rectangle;
For each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle distance, judge whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, it is labeled as the rectangle to be identified and target rectangle to belong to same
One Chinese character label judges whether also have not labeled rectangle in the rectangle, a rectangle conduct is chosen if so, then executing
The step of rectangle to be identified, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target rectangle apart from corresponding rectangle by described,
It executes and is directed to each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle apart from the step of.
14. device according to claim 12, which is characterized in that the first acquisition unit is specifically used for:
Its starting point is denoted as point A by a lane line for choosing the track;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation L1 and L2 of wherein minimum and maximum distance;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line equation and institute
The intersection point for stating lane line is denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and obtains the corresponding vertical line side of wherein minimum and maximum distance
Journey L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point constitutes the rectangle for covering all the points in the cluster.
15. a kind of identification device of Chinese character label, which is characterized in that the identification device includes:
For the point cloud data in a track to a road, clustered according to distance between points, obtain to
The unit of a few cluster;
For being directed to each point cluster, the unit for covering the rectangle of all the points in the cluster is obtained;
For the spatial relationship according to each rectangle for putting cluster, the unit for belonging to the rectangle of the same label is obtained;
For merging the rectangle for belonging to the same label, the corresponding target rectangle of the label, the mesh are obtained
Mark rectangle covers the unit of all rectangles for belonging to the same label;
For the size according to target rectangle and its position in track, judge whether target rectangle belongs to Chinese character label, if
It is, then by the target rectangle labeled as the unit of the rectangle of Chinese character label.
16. identification device according to claim 15, which is characterized in that described for in a track of a road
Point cloud data, clustered according to distance between points, the unit for obtaining at least one cluster is specifically used for:
From the point cloud data in the same track of a road, a point is arbitrarily chosen as point to be identified;
Obtain the distance of left point in the point to the point cloud data to be identified;
Obtain the number of the point at a distance from point to be identified less than preset first distance threshold;
If the number is more than preset number threshold value, it is less than pre-determined distance threshold value by the point to be identified and to its distance
Point is divided into same cluster;
From the point cloud data, the point for being divided into a cluster is deleted, if after deleting, also having remaining point in the point cloud data,
The step of one point of arbitrary selection is as to be identified is then executed, until the point cloud data is sky.
17. identification device according to claim 16, which is characterized in that described for for each point cluster, obtaining covering
The unit of the rectangle of all the points is specifically used in the cluster:
Its starting point is denoted as point A by a lane line for choosing the track;
Vertical line is done by each being put to the lane line in cluster, obtains intersection point point of each point on the lane line;
The distance of each intersection point point-to-point A is calculated, and obtains the corresponding vertical line equation L1 and L2 of wherein minimum and maximum distance;
An optional vertical line equation, which is used as, from vertical line the equation L1 and L2 refers to vertical line equation, will refer to vertical line equation and institute
The intersection point for stating lane line is denoted as point B;
It obtains each to put in described cluster and arrives the reference intersection point point with reference to vertical line equation;
The distance for each referring to intersection point point-to-point B is calculated, and obtains the corresponding vertical line side of wherein minimum and maximum distance
Journey L3 and L4;
The intersection point of four vertical line equations L1, L2, L3 and L4 are obtained, the intersection point constitutes the rectangle for covering all the points in the cluster.
18. identification device according to claim 16, which is characterized in that the sky for according to each rectangle for putting cluster
Between relationship, the unit for obtaining belonging to the rectangle of the same label is specifically used for:
From the corresponding rectangle of all the points cluster, a rectangle is arbitrarily chosen as rectangle to be identified;
Rectangle to be identified is obtained at a distance from the rectangle in the rectangle in addition to rectangle to be identified;
The distance less than preset second distance threshold value is judged whether there is, if so, being then determined as described apart from corresponding rectangle
Target rectangle;
For each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle distance, judge whether there is the distance less than preset second distance threshold value;
If without the distance for being less than preset second distance threshold value, it is labeled as the rectangle to be identified and target rectangle to belong to same
One label judges whether also have not labeled rectangle in the rectangle, waits knowing if so, then executing one rectangle of selection and being used as
The step of other rectangle, terminates flow if nothing;
If there is the distance less than preset second distance threshold value, also it is determined as target rectangle apart from corresponding rectangle by described,
It executes and is directed to each target rectangle, obtain in target rectangle and the rectangle in addition to rectangle to be identified and all target rectangles
Rectangle apart from the step of.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710171393.6A CN108629238B (en) | 2017-03-21 | 2017-03-21 | Method and device for identifying Chinese character mark |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710171393.6A CN108629238B (en) | 2017-03-21 | 2017-03-21 | Method and device for identifying Chinese character mark |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108629238A true CN108629238A (en) | 2018-10-09 |
CN108629238B CN108629238B (en) | 2020-07-10 |
Family
ID=63706425
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710171393.6A Active CN108629238B (en) | 2017-03-21 | 2017-03-21 | Method and device for identifying Chinese character mark |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108629238B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111045025A (en) * | 2018-10-15 | 2020-04-21 | 图森有限公司 | Vehicle tracking method and system based on light detection and distance measurement |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103186787A (en) * | 2011-12-31 | 2013-07-03 | 廖志武 | Low-quality Chinese character primary skeleton extraction algorithm based on point cloud model |
CN104951741A (en) * | 2014-03-31 | 2015-09-30 | 阿里巴巴集团控股有限公司 | Character recognition method and device thereof |
CN105740860A (en) * | 2016-01-28 | 2016-07-06 | 河南大学 | Automatic detection method for Chinese character area of shop sign in natural scene |
CN106407947A (en) * | 2016-09-29 | 2017-02-15 | 百度在线网络技术(北京)有限公司 | Target object recognition method and device applied to unmanned vehicle |
-
2017
- 2017-03-21 CN CN201710171393.6A patent/CN108629238B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103186787A (en) * | 2011-12-31 | 2013-07-03 | 廖志武 | Low-quality Chinese character primary skeleton extraction algorithm based on point cloud model |
CN104951741A (en) * | 2014-03-31 | 2015-09-30 | 阿里巴巴集团控股有限公司 | Character recognition method and device thereof |
CN105740860A (en) * | 2016-01-28 | 2016-07-06 | 河南大学 | Automatic detection method for Chinese character area of shop sign in natural scene |
CN106407947A (en) * | 2016-09-29 | 2017-02-15 | 百度在线网络技术(北京)有限公司 | Target object recognition method and device applied to unmanned vehicle |
Non-Patent Citations (1)
Title |
---|
G. X. XU 等: "Clustering Research across Tibetan and Chinese Texts", 《JOURNAL OF DIGITAL INFORMATION MANAGEMENT》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111045025A (en) * | 2018-10-15 | 2020-04-21 | 图森有限公司 | Vehicle tracking method and system based on light detection and distance measurement |
Also Published As
Publication number | Publication date |
---|---|
CN108629238B (en) | 2020-07-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111666921B (en) | Vehicle control method, apparatus, computer device, and computer-readable storage medium | |
CN110634291A (en) | High-precision map topology automatic construction method and system based on crowdsourcing data | |
WO2017041396A1 (en) | Driving lane data processing method, device, storage medium and apparatus | |
US9436682B2 (en) | Techniques for machine language translation of text from an image based on non-textual context information from the image | |
JP2019135544A (en) | Method and apparatus for identifying intersection in electronic map | |
US10984588B2 (en) | Obstacle distribution simulation method and device based on multiple models, and storage medium | |
CN104067314A (en) | Human-shaped image segmentation method | |
CN109871029A (en) | A kind of plant protection drone flight course planning optimization method based on image processing techniques | |
JP6850325B2 (en) | Obstacle distribution simulation methods, devices, terminals, storage media, and programs based on probability plots | |
CN105574527A (en) | Quick object detection method based on local feature learning | |
CN112699834B (en) | Traffic identification detection method, device, computer equipment and storage medium | |
CN107392252A (en) | Computer deep learning characteristics of image and the method for quantifying perceptibility | |
CN108182218A (en) | A kind of video character recognition method, system and electronic equipment based on GIS-Geographic Information System | |
CN110399760A (en) | A kind of batch two dimensional code localization method, device, electronic equipment and storage medium | |
Tran et al. | Pp-linknet: Improving semantic segmentation of high resolution satellite imagery with multi-stage training | |
CN114937293B (en) | GIS-based agricultural service management method and system | |
CN114519819B (en) | Remote sensing image target detection method based on global context awareness | |
Liu et al. | A new multi-channel deep convolutional neural network for semantic segmentation of remote sensing image | |
CN109657540B (en) | Withered tree positioning method and system | |
CN108629238A (en) | A kind of method and apparatus of identification Chinese character label | |
WO2023179382A1 (en) | Method and apparatus for marking obstacle in robot map | |
CN111881752A (en) | Guardrail detection and classification method and device, electronic equipment and storage medium | |
CN107403151A (en) | Method, apparatus, equipment and the computer-readable recording medium positioned by indoor ceiling | |
CN112529116A (en) | Scene element fusion processing method, device and equipment and computer storage medium | |
CN111738051B (en) | Point cloud processing method and device, computer equipment and storage medium |
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 | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20200508 Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province Applicant after: Alibaba (China) Co.,Ltd. Address before: 102200, No. 18, No., Changsheng Road, Changping District science and Technology Park, Beijing, China. 1-5 Applicant before: AUTONAVI SOFTWARE Co.,Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |