CN109840535A - The method and apparatus for realizing classification of landform - Google Patents
The method and apparatus for realizing classification of landform Download PDFInfo
- Publication number
- CN109840535A CN109840535A CN201711228122.6A CN201711228122A CN109840535A CN 109840535 A CN109840535 A CN 109840535A CN 201711228122 A CN201711228122 A CN 201711228122A CN 109840535 A CN109840535 A CN 109840535A
- Authority
- CN
- China
- Prior art keywords
- landform
- character subset
- feature
- classification
- concentrated
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses the method and apparatus for realizing classification of landform, are related to Technology of Terrain Classifying field.One specific embodiment of this method includes: to extract character subset from the laser data of landform to be sorted, according to the character subset train classification models；From and the corresponding image data of the laser data in extract characteristic corresponding with the character subset；According to the characteristic and the disaggregated model, classification of landform is realized.The embodiment can greatly improve the accuracy rate of classification of landform, and classify short time, efficiency and real-time are high.
Description
Technical field
The present invention relates to Technology of Terrain Classifying field more particularly to a kind of method and apparatus for realizing classification of landform.
Background technique
Relief form has the characteristics that diversity, complexity, variability, randomness, and the appearance of relief form can also be with
The external conditions such as season, illumination change.These variations will increase the difficulty of classification of landform.
Existing classification of landform method is all to complete characteristic extraction procedure offline, according to the feature off-line training of extraction point
Class device realizes classification of landform according to the classifier of off-line training.
In realizing process of the present invention, at least there are the following problems in the prior art for inventor's discovery:
1) to realize that the extracted feature of classification of landform is more, cause the classification time long, low efficiency；
2) classifier of off-line training is poor to the adaptability of environment, can not adapt to terrain environment complicated and changeable, knows
Other accuracy rate is low.
Summary of the invention
In view of this, the embodiment of the present invention provides a kind of method and apparatus for realizing classification of landform, it can greatly improve ground
The accuracy rate of shape classification, and classify short time, efficiency and real-time are high.
To achieve the above object, according to an aspect of an embodiment of the present invention, a kind of side for realizing classification of landform is provided
Method, comprising:
Character subset is extracted from the laser data of landform to be sorted, according to the character subset train classification models；
From and the corresponding image data of the laser data in extract characteristic corresponding with the character subset；
According to the characteristic and the disaggregated model, classification of landform is realized.
Optionally, character subset is extracted from the laser data of landform to be sorted includes:
Candidate characteristic set is extracted from the laser data of landform to be sorted, according to preset Feature Selection rule from the time
It selects and screens the character subset in feature set.
Optionally, it is concentrated using random forests algorithm from the candidate feature and screens the character subset.
Optionally, screening the character subset from candidate feature concentration using random forests algorithm includes:
The feature for being unsatisfactory for preset condition is concentrated to reject the candidate feature；Remaining spy is concentrated with the candidate feature
Sign establishes random forest grader；Determine the extensive error of the random forest grader；
If the extensive error of the random gloomy inner disaggregated model is unsatisfactory for default precision conditions, recycles and execute above-mentioned step
Suddenly；Otherwise, remaining each feature is concentrated to form the character subset with the candidate feature.
Optionally, the extensive error of the random forest grader is determined using the data error estimation technique outside bag.
It optionally, include: according to the character subset, using random forest according to the character subset train classification models
Algorithm train classification models.
Another aspect according to an embodiment of the present invention provides a kind of device for realizing classification of landform, comprising:
Characteristic extracting module extracts character subset from the laser data of landform to be sorted, from the laser data pair
Characteristic corresponding with the character subset is extracted in the image data answered；
Model training module, according to the character subset train classification models；
Classification of landform module realizes classification of landform according to the characteristic and the disaggregated model.
Optionally, the characteristic extracting module extracts candidate characteristic set from the laser data of landform to be sorted, according to pre-
If Feature Selection rule concentrated from the candidate feature and screen the character subset.
Optionally, the characteristic extracting module is concentrated from the candidate feature using random forests algorithm and screens the feature
Subset.
Optionally, the characteristic extracting module is concentrated from the candidate feature using random forests algorithm and screens the feature
Subset includes:
The feature for being unsatisfactory for preset condition is concentrated to reject the candidate feature；Remaining spy is concentrated with the candidate feature
Sign establishes random forest grader；Determine the extensive error of the random forest grader；
If the extensive error of the random gloomy inner disaggregated model is unsatisfactory for default precision conditions, recycles and execute above-mentioned step
Suddenly；Otherwise, remaining each feature is concentrated to form the character subset with the candidate feature.
Optionally, the characteristic extracting module determines the random forest grader using the outer data error estimation technique of bag
Extensive error.
Optionally, the model training module includes: according to the feature according to the character subset train classification models
Subset, using random forests algorithm train classification models.
Other side according to an embodiment of the present invention provides a kind of electronic equipment for realizing classification of landform, comprising:
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes the method that the embodiment of the present invention realizes classification of landform.
Still another aspect according to an embodiment of the present invention provides a kind of computer-readable medium, is stored thereon with calculating
Machine program realizes the method that the embodiment of the present invention realizes classification of landform when described program is executed by processor.
One embodiment in foregoing invention have the following advantages that or the utility model has the advantages that
1) character subset train classification models are extracted from the laser data of landform to be sorted by basis, so that training
Disaggregated model can be suitable for the actual conditions of current landform to be sorted, greatly improve the real-time and ring of classification of landform method
Border adaptability, and then improve the accuracy of classification of landform；
2) candidate characteristic set is extracted from the laser data of landform to be sorted, according to preset Feature Selection rule from candidate
Character subset is screened in feature set, so as to use less feature to carry out classification of landform, is improved the speed of classification of landform, is dropped
The classification time of low classification of landform, improve the efficiency of classification of landform；
3) screening character subset is concentrated from candidate feature using random forests algorithm, can filtered out and landform to be sorted
The feature of actual environment matched, to greatly improve landform point while carrying out classification of landform using less feature
The accuracy of class；
4) the extensive error of random forest grader, resulting extensive error are determined using the outer data error estimation technique of bag is
Unbiased esti-mator no longer needs to carry out cross validation or obtains the unbiased esti-mator of extensive error by test set, is simple and efficient；
5) random forests algorithm train classification models are used, training speed is fast, carries out based on its disaggregated model trained
The accuracy of classification of landform is high.
Further effect possessed by above-mentioned non-usual optional way adds hereinafter in conjunction with specific embodiment
With explanation.
Detailed description of the invention
Attached drawing for a better understanding of the present invention, does not constitute an undue limitation on the present invention.Wherein:
Fig. 1 is the schematic diagram of the main flow of the method according to an embodiment of the present invention for realizing classification of landform；
Fig. 2 be according to the present invention alternative embodiment from the laser data of landform to be sorted extract character subset it is main
The schematic diagram of process；
Fig. 3 is the schematic diagram of the key step of the method for the realization classification of landform of alternative embodiment according to the present invention；
Fig. 4 is the schematic diagram of the main modular of the device according to an embodiment of the present invention for realizing classification of landform；
Fig. 5 is that the embodiment of the present invention can be applied to exemplary system architecture figure therein；
Fig. 6 is adapted for the structural representation of the computer system for the terminal device or server of realizing the embodiment of the present invention
Figure.
Specific embodiment
Below in conjunction with attached drawing, an exemplary embodiment of the present invention will be described, including the various of the embodiment of the present invention
Details should think them only exemplary to help understanding.Therefore, those of ordinary skill in the art should recognize
It arrives, it can be with various changes and modifications are made to the embodiments described herein, without departing from scope and spirit of the present invention.Together
Sample, for clarity and conciseness, descriptions of well-known functions and structures are omitted from the following description.
It should be noted that without departing substantially from scope and spirit of the present invention, it is as mentioned in the embodiments of the present invention
Each technical characteristic can in any combination, and the sequence of each step can also exchange.
Fig. 1 is the schematic diagram of the main flow of the method according to an embodiment of the present invention for realizing classification of landform, such as Fig. 1 institute
Show, provide a kind of method for realizing classification of landform, comprising:
Step S101, character subset is extracted from the laser data of landform to be sorted, according to character subset training point
Class model；
Step S102, from and the corresponding image data of the laser data in extract feature corresponding with the character subset
Data；
Step S103, according to the characteristic and the disaggregated model, classification of landform is realized.
In the embodiment of the present invention, estimated by the ground that laser data can obtain landform to be sorted, for example, using this hair
The data point of laser data is divided into ground data point and non-ground data point by the disaggregated model of bright embodiment, and characteristic is defeated
Enter in disaggregated model and classify, so that these data points of laser data are projected to figure corresponding with this frame laser data
As obtaining passable ground region and not passable barrier region in data.The embodiment of the present invention also can be applied to
Classification of landform field of the unmanned vehicle in field environment.
In the embodiment of the present invention, for each landform to be sorted, extracted according to from the laser data of the landform to be sorted
Character subset train classification models.Compared with using offline mode to obtain in the prior art and train classification models, the present embodiment
Classification landform can be treated in real time to classify, real-time is good.Since the character subset for train classification models is from wait divide
It is extracted in the laser data of class landform, therefore highly relevant with the actual environment of landform to be sorted, landform can be greatly improved
The environmental suitability and accuracy of classification method.
Relief form has diversity, complexity, for the ease of fully understanding relief form, when extracting character subset,
It can classify according to default rule to each feature, for example, each feature is divided into color characteristic, textural characteristics, several
What feature etc..It should be understood that whether carrying out classifying and how to carry out to classify to feature having no effect on the embodiment of the present invention
The implementation of technical solution, therefore, the application is not specifically limited in this embodiment.
In some alternative embodiments, candidate characteristic set can be extracted from the laser data of landform to be sorted, according to
Preset Feature Selection rule concentrates screening character subset from candidate feature.Candidate characteristic set mentioned herein, refers in landform
Classification field is common and has the set of the feature centainly influenced on landform classification results.How the embodiment of the present invention is to extracting
The feature of candidate characteristic set and candidate characteristic set is not specifically limited.When carrying out Feature Selection, those skilled in the art can
To select screening technique according to the actual conditions of application scenarios, for example, classification results accuracy is influenced according to each feature
Candidate feature is concentrated the lower feature of importance to reject, is formed with remaining feature by size, i.e., the importance of each feature
Gather the character subset as the embodiment of the present invention.By concentrating screening special from candidate feature according to preset Feature Selection rule
Subset is levied, less feature can be used and carry out classification of landform, the speed of classification of landform is improved, when reducing the classification of classification of landform
Between, improve the efficiency of classification of landform.
In actual application, screening character subset can be concentrated from candidate feature using random forests algorithm.It is random gloomy
Woods algorithm is exactly a kind of algorithm by the thought of integrated study that more trees is integrated, its basic unit is decision tree, and every
Decision tree is all a classifier, is not associated between each decision tree.When test data enters random forest, it is exactly in fact
Each decision tree is allowed to classify, finally taking that class that classification results are most in all decision trees is final result.Therefore
Random forest is the classifier comprising multiple decision trees, and classification of its output is the classification exported by individual decision trees
Mode depending on.
Screening character subset is concentrated from candidate feature using random forests algorithm, can be obtained for the environment of different terrain
Different character subsets is made the disaggregated model obtained based on character subset training can adapt to the environmental change of landform, improves ground
The adaptability of shape classification method.Further, since the feature with the actual environment matched of landform to be sorted can be filtered out, from
And the accuracy of classification of landform can be greatly improved while carrying out classification of landform using less feature.
Fig. 2 be according to the present invention alternative embodiment from the laser data of landform to be sorted extract character subset it is main
The schematic diagram of process, as shown in Fig. 2, the main flow for extracting character subset from the laser data of landform to be sorted includes:
Step S201, candidate characteristic set is extracted from the laser data of landform to be sorted.
Step S202, the feature for being unsatisfactory for preset condition is concentrated to reject candidate feature.
Preset condition can be set according to the actual situation.For example, preset condition can be arranged are as follows: screening every time
When, the spy for the certain amount (for example, 20% of candidate feature concentration feature total quantity) for concentrating importance minimum candidate feature
Sign is rejected.For another example preset condition can be arranged are as follows: every time when screening, concentrate importance to be less than candidate feature a certain pre-
If the feature of value is rejected.The embodiment of the present invention is not specifically limited preset condition.
Step S203, remaining feature is concentrated to establish random forest grader with candidate feature.
Step S204, the extensive error of random forest grader is determined.
Extensive error refers to the gap between the actual classification result of random forest grader and theoretic classification result.It is extensive
Error is smaller, shows that the accuracy of the random forest grader is higher.
Step S205, judge whether the extensive error of random forest grader meets preset precision conditions.
Preset precision conditions are used to measure the accuracy of random forest grader, and the content for presetting precision conditions can root
It is configured according to the needs of practical application scene and the error level of requirement, the embodiment of the present invention is not specifically limited in this embodiment.
For example, preset precision conditions are as follows: the extensive error of random forest grader is not more than preset error threshold, then when random gloomy
When the extensive error of woods classifier is less than or equal to the error threshold, judge that the random forest grader meets preset precision item
Part.
If the extensive error of step S206, random gloomy inner disaggregated model meets default precision conditions, concentrated with candidate feature
Remaining each feature forms character subset；
If the extensive error of random gloomy inner disaggregated model is unsatisfactory for default precision conditions, execution above-mentioned steps are recycled
S202-S205。
In some embodiments, the extensive mistake of random forest grader can be determined using the outer data error estimation technique of bag
Difference.The outer data error estimation technique of bag refers to, for the random forest generated, using its performance of data test outside bag.Assuming that
The outer data count of bag is O, and data outside this O item bag is used to bring the random forest grader generated, random forest into as input
Classifier can provide the classification of the outer data of O item bag.The number for counting random forest grader classification error, is set as X, because of this O
The type of data is known outside item bag, is compared with known classification with the result of random forest grader, then random gloomy
Outer data error=the X/O of the bag of woods classifier.The extensive mistake of random forest grader is determined using the data error estimation technique outside bag
Difference, resulting extensive error is unbiased esti-mator, no longer needs to carry out cross validation or obtains the unbiased of extensive error by test set
Estimation, is simple and efficient.
It may include: to be instructed according to the character subset using random forests algorithm according to character subset train classification models
Practice disaggregated model.Using random forests algorithm train classification models, training speed is fast, is carried out based on its disaggregated model trained
The accuracy of classification of landform is high.
Fig. 3 is the schematic diagram of the key step of the method for the realization classification of landform of alternative embodiment according to the present invention, such as Fig. 3
It is shown, comprising:
Step S301, classify to laser data.It, can be according to default rule to each when extracting character subset
Feature is classified, for example, each feature is divided into color characteristic, textural characteristics, geometrical characteristic etc..
Step S302, candidate characteristic set is extracted from the laser data of landform to be sorted；
Step S303, it is concentrated using random forests algorithm from candidate feature and extracts character subset；
Step S304, according to character subset train classification models；
Step S305, from and the corresponding image data of laser data in extract characteristic corresponding with character subset；
Step S306, according to characteristic and disaggregated model, classification of landform is realized.Specifically, by characteristic input point
Classify in class model, the result of classification of landform is determined according to the output content of disaggregated model.
Fig. 4 is the schematic diagram of the main modular of the device according to an embodiment of the present invention for realizing classification of landform, such as Fig. 4 institute
Show, realizes that the device 400 of classification of landform includes:
Characteristic extracting module 401 extracts character subset from the laser data of landform to be sorted, from the laser data
Characteristic corresponding with the character subset is extracted in corresponding image data；
Model training module 402, according to the character subset train classification models；
Classification of landform module 403 realizes classification of landform according to the characteristic and the disaggregated model.
Optionally, the characteristic extracting module extracts candidate characteristic set from the laser data of landform to be sorted, according to pre-
If Feature Selection rule concentrated from the candidate feature and screen the character subset.
Optionally, the characteristic extracting module is concentrated from the candidate feature using random forests algorithm and screens the feature
Subset.
Optionally, the characteristic extracting module is concentrated from the candidate feature using random forests algorithm and screens the feature
Subset includes:
The feature for being unsatisfactory for preset condition is concentrated to reject the candidate feature；Remaining spy is concentrated with the candidate feature
Sign establishes random forest grader；Determine the extensive error of the random forest grader；
If the extensive error of the random gloomy inner disaggregated model is unsatisfactory for default precision conditions, recycles and execute above-mentioned step
Suddenly；Otherwise, remaining each feature is concentrated to form the character subset with the candidate feature.
Optionally, the characteristic extracting module determines the random forest grader using the outer data error estimation technique of bag
Extensive error.
Optionally, the model training module includes: according to the feature according to the character subset train classification models
Subset, using random forests algorithm train classification models.
Other side according to an embodiment of the present invention provides a kind of electronic equipment for realizing classification of landform, comprising:
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes the method that the embodiment of the present invention realizes classification of landform.
Fig. 5 is shown can be using the method for the realization classification of landform of the embodiment of the present invention or the device of realization classification of landform
Exemplary system architecture 500.
As shown in figure 5, system architecture 500 may include terminal device 501,502,503, network 504 and server 505.
Network 504 between terminal device 501,502,503 and server 505 to provide the medium of communication link.Network 504 can be with
Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be used terminal device 501,502,503 and be interacted by network 504 with server 505, to receive or send out
Send message etc..Various telecommunication customer end applications, such as the application of shopping class, net can be installed on terminal device 501,502,503
(merely illustrative) such as the application of page browsing device, searching class application, instant messaging tools, mailbox client, social platform softwares.
Terminal device 501,502,503 can be the various electronic equipments with display screen and supported web page browsing, packet
Include but be not limited to smart phone, tablet computer, pocket computer on knee and desktop computer etc..
Server 505 can be to provide the server of various services, such as utilize terminal device 501,502,503 to user
The shopping class website browsed provides the back-stage management server (merely illustrative) supported.Back-stage management server can dock
The data such as the information query request received analyze etc. processing, and by processing result (such as target push information, product
Information -- merely illustrative) feed back to terminal device.
It should be noted that realizing that the method for classification of landform is generally held by server 505 provided by the embodiment of the present invention
Row, correspondingly, realizes that the device of classification of landform is generally positioned in server 505.
It should be understood that the number of terminal device, network and server in Fig. 5 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.
Below with reference to Fig. 6, it illustrates the computer systems 600 for the terminal device for being suitable for being used to realize the embodiment of the present invention
Structural schematic diagram.Terminal device shown in Fig. 6 is only an example, function to the embodiment of the present invention and should not use model
Shroud carrys out any restrictions.
As shown in fig. 6, computer system 600 includes central processing unit (CPU) 601, it can be read-only according to being stored in
Program in memory (ROM) 602 is loaded into the program in random access storage device (RAM) 603 from storage section 608
And execute various movements appropriate and processing.In RAM 603, also it is stored with system 600 and operates required various program sum numbers
According to.CPU 601, ROM 602 and RAM 603 are connected with each other by bus 604.Input/output (I/O) interface 605 also connects
To bus 604.
I/O interface 605 is connected to lower component: the importation 606 including keyboard, mouse etc.；It is penetrated including such as cathode
The output par, c 607 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage section 608 including hard disk etc.；
And the communications portion 609 of the network interface card including LAN card, modem etc..Communications portion 609 via such as because
The network of spy's net executes communication process.Driver 610 is also connected to I/O interface 605 as needed.Detachable media 611, such as
Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on as needed on driver 610, in order to read from thereon
Computer program be mounted into storage section 608 as needed.
Particularly, disclosed embodiment, the process described above with reference to flow chart may be implemented as counting according to the present invention
Calculation machine software program.For example, embodiment disclosed by the invention includes a kind of computer program product comprising be carried on computer
Computer program on readable medium, the computer program include the program code for method shown in execution flow chart.?
In such embodiment, which can be downloaded and installed from network by communications portion 609, and/or from can
Medium 611 is dismantled to be mounted.When the computer program is executed by central processing unit (CPU) 601, executes and of the invention be
The above-mentioned function of being limited in system.
It should be noted that computer-readable medium shown in the present invention can be computer-readable signal media or meter
Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not
Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter
The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires
Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage
Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device,
Or above-mentioned any appropriate combination.In the present invention, computer readable storage medium can be it is any include or storage journey
The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this
In invention, computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned
Any appropriate combination.
Flow chart and block diagram in attached drawing are illustrated according to the system of various embodiments of the invention, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of above-mentioned module, program segment or code include one or more
Executable instruction for implementing the specified logical function.It should also be noted that in some implementations as replacements, institute in box
The function of mark can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are practical
On can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it wants
It is noted that the combination of each box in block diagram or flow chart and the box in block diagram or flow chart, can use and execute rule
The dedicated hardware based systems of fixed functions or operations is realized, or can use the group of specialized hardware and computer instruction
It closes to realize.
Being described in module involved in the embodiment of the present invention can be realized by way of software, can also be by hard
The mode of part is realized.Described module also can be set in the processor, for example, can be described as: a kind of processor packet
Include: characteristic extracting module extracts character subset from the laser data of landform to be sorted, from figure corresponding with the laser data
As extracting characteristic corresponding with the character subset in data；Model training module, according to character subset training point
Class model；Classification of landform module realizes classification of landform according to the characteristic and the disaggregated model.Wherein, these modules
Title do not constitute the restriction to the module itself under certain conditions, for example, characteristic extracting module is also described as
" module for extracting the request of character subset is sent to the server-side connected ".
As on the other hand, the present invention also provides a kind of computer-readable medium, which be can be
Included in equipment described in above-described embodiment；It is also possible to individualism, and without in the supplying equipment.Above-mentioned calculating
Machine readable medium carries one or more program, when said one or multiple programs are executed by the equipment, makes
Obtaining the equipment includes:
Character subset is extracted from the laser data of landform to be sorted, according to the character subset train classification models；
From and the corresponding image data of the laser data in extract characteristic corresponding with the character subset；
According to the characteristic and the disaggregated model, classification of landform is realized.
One embodiment in foregoing invention have the following advantages that or the utility model has the advantages that
1) character subset train classification models are extracted from the laser data of landform to be sorted by basis, so that training
Disaggregated model can be suitable for the actual conditions of current landform to be sorted, greatly improve the real-time and ring of classification of landform method
Border adaptability, and then improve the accuracy of classification of landform；
2) candidate characteristic set is extracted from the laser data of landform to be sorted, according to preset Feature Selection rule from candidate
Character subset is screened in feature set, so as to use less feature to carry out classification of landform, is improved the speed of classification of landform, is dropped
The classification time of low classification of landform, improve the efficiency of classification of landform；
3) screening character subset is concentrated from candidate feature using random forests algorithm, can filtered out and landform to be sorted
The feature of actual environment matched, to greatly improve landform point while carrying out classification of landform using less feature
The accuracy of class；
4) the extensive error of random forest grader, resulting extensive error are determined using the outer data error estimation technique of bag is
Unbiased esti-mator no longer needs to carry out cross validation or obtains the unbiased esti-mator of extensive error by test set, is simple and efficient；
5) random forests algorithm train classification models are used, training speed is fast, carries out based on its disaggregated model trained
The accuracy of classification of landform is high.
Above-mentioned specific embodiment, does not constitute a limitation on the scope of protection of the present invention.Those skilled in the art should be bright
It is white, design requirement and other factors are depended on, various modifications, combination, sub-portfolio and substitution can occur.It is any
Made modifications, equivalent substitutions and improvements etc. within the spirit and principles in the present invention, should be included in the scope of the present invention
Within.
Claims (14)
1. a kind of method for realizing classification of landform characterized by comprising
Character subset is extracted from the laser data of landform to be sorted, according to the character subset train classification models；
From and the corresponding image data of the laser data in extract characteristic corresponding with the character subset；
According to the characteristic and the disaggregated model, classification of landform is realized.
2. the method as described in claim 1, which is characterized in that extract character subset packet from the laser data of landform to be sorted
It includes:
Candidate characteristic set is extracted from the laser data of landform to be sorted, it is special from the candidate according to preset Feature Selection rule
The character subset is screened in collection.
3. the method as described in claim 1, which is characterized in that concentrated and screened from the candidate feature using random forests algorithm
The character subset.
4. method as claimed in claim 3, which is characterized in that concentrated and screened from the candidate feature using random forests algorithm
The character subset includes:
The feature for being unsatisfactory for preset condition is concentrated to reject the candidate feature；Remaining feature is concentrated to build with the candidate feature
Vertical random forest grader；Determine the extensive error of the random forest grader；
If the extensive error of the random gloomy inner disaggregated model is unsatisfactory for default precision conditions, execution above-mentioned steps are recycled；It is no
Then, remaining each feature is concentrated to form the character subset with the candidate feature.
5. method as claimed in claim 4, which is characterized in that determine the random forest using the data error estimation technique outside bag
The extensive error of classifier.
6. the method as described in claim 1, which is characterized in that according to the character subset train classification models include: basis
The character subset, using random forests algorithm train classification models.
7. a kind of device for realizing classification of landform characterized by comprising
Characteristic extracting module extracts character subset from the laser data of landform to be sorted, from corresponding with the laser data
Characteristic corresponding with the character subset is extracted in image data；
Model training module, according to the character subset train classification models；
Classification of landform module realizes classification of landform according to the characteristic and the disaggregated model.
8. device as claimed in claim 7, which is characterized in that laser data of the characteristic extracting module from landform to be sorted
Middle extraction candidate characteristic set is concentrated from the candidate feature according to preset Feature Selection rule and screens the character subset.
9. device as claimed in claim 8, which is characterized in that the characteristic extracting module is using random forests algorithm from described
Candidate feature, which is concentrated, screens the character subset.
10. device as claimed in claim 9, which is characterized in that the characteristic extracting module uses random forests algorithm from institute
Stating the candidate feature concentration screening character subset includes:
The feature for being unsatisfactory for preset condition is concentrated to reject the candidate feature；Remaining feature is concentrated to build with the candidate feature
Vertical random forest grader；Determine the extensive error of the random forest grader；
If the extensive error of the random gloomy inner disaggregated model is unsatisfactory for default precision conditions, execution above-mentioned steps are recycled；It is no
Then, remaining each feature is concentrated to form the character subset with the candidate feature.
11. device as claimed in claim 10, which is characterized in that the characteristic extracting module is using the outer data error estimation of bag
Method determines the extensive error of the random forest grader.
12. device as claimed in claim 7, which is characterized in that the model training module is according to character subset training
Disaggregated model includes: according to the character subset, using random forests algorithm train classification models.
13. a kind of electronic equipment for realizing classification of landform characterized by comprising
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as method as claimed in any one of claims 1 to 6.
14. a kind of computer-readable medium, is stored thereon with computer program, which is characterized in that described program is held by processor
Such as method as claimed in any one of claims 1 to 6 is realized when row.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711228122.6A CN109840535B (en) | 2017-11-29 | 2017-11-29 | Method and device for realizing terrain classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711228122.6A CN109840535B (en) | 2017-11-29 | 2017-11-29 | Method and device for realizing terrain classification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109840535A true CN109840535A (en) | 2019-06-04 |
| CN109840535B CN109840535B (en) | 2021-10-01 |
Family
ID=66882132
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711228122.6A Active CN109840535B (en) | 2017-11-29 | 2017-11-29 | Method and device for realizing terrain classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109840535B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113256695A (en) * | 2021-06-23 | 2021-08-13 | 武汉工程大学 | Random forest based terrain prediction model method for potassium sulfate production salt pond |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831646A (en) * | 2012-08-13 | 2012-12-19 | 东南大学 | Scanning laser based large-scale three-dimensional terrain modeling method |
| CN103645480A (en) * | 2013-12-04 | 2014-03-19 | 北京理工大学 | Geographic and geomorphic characteristic construction method based on laser radar and image data fusion |
| CN105931224A (en) * | 2016-04-14 | 2016-09-07 | 浙江大学 | Pathology identification method for routine scan CT image of liver based on random forests |
| CN106338736A (en) * | 2016-08-31 | 2017-01-18 | 东南大学 | Full-3D occupation volume element landform modeling method based on laser radar |
| CN106871932A (en) * | 2017-04-20 | 2017-06-20 | 国家测绘地理信息局卫星测绘应用中心 | The in-orbit sensing calibration method of satellite borne laser based on Pyramidal search terrain match |
-
2017
- 2017-11-29 CN CN201711228122.6A patent/CN109840535B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831646A (en) * | 2012-08-13 | 2012-12-19 | 东南大学 | Scanning laser based large-scale three-dimensional terrain modeling method |
| CN103645480A (en) * | 2013-12-04 | 2014-03-19 | 北京理工大学 | Geographic and geomorphic characteristic construction method based on laser radar and image data fusion |
| CN105931224A (en) * | 2016-04-14 | 2016-09-07 | 浙江大学 | Pathology identification method for routine scan CT image of liver based on random forests |
| CN106338736A (en) * | 2016-08-31 | 2017-01-18 | 东南大学 | Full-3D occupation volume element landform modeling method based on laser radar |
| CN106871932A (en) * | 2017-04-20 | 2017-06-20 | 国家测绘地理信息局卫星测绘应用中心 | The in-orbit sensing calibration method of satellite borne laser based on Pyramidal search terrain match |
Non-Patent Citations (3)
| Title |
|---|
| CRISTIAN DIMA 等: "Enabling Learning From Large Datasets: Applying Active Learning to Mobile Robotics", 《IEEE》 * |
| 周媛 等: "基于激光测距数据组合特征的森林环境地形分类", 《PROCEEDINGS OF THE 32ND CHINESE CONTROL CONFERENCE》 * |
| 李树伦: "野外环境中基于激光和视觉的自监督地形分类研究", 《万方数据》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113256695A (en) * | 2021-06-23 | 2021-08-13 | 武汉工程大学 | Random forest based terrain prediction model method for potassium sulfate production salt pond |
| CN113256695B (en) * | 2021-06-23 | 2021-10-08 | 武汉工程大学 | Random forest based terrain prediction model method for potassium sulfate production salt pond |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109840535B (en) | 2021-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108197532B (en) | The method, apparatus and computer installation of recognition of face | |
| CN101593270B (en) | Method for recognizing hand-painted shapes and device thereof | |
| CN108629823A (en) | The generation method and device of multi-view image | |
| CN106462559A (en) | Arbitrary size content item generation | |
| CN107908666A (en) | A kind of method and apparatus of identification equipment mark | |
| CN109063653A (en) | Image processing method and device | |
| CN108769905A (en) | Method and device for the classification for determining wireless access point | |
| CN107679119A (en) | The method and apparatus for generating brand derivative words | |
| CN105721629A (en) | User identifier matching method and device | |
| CN107644106B (en) | Method, terminal device and storage medium for automatically mining service middleman | |
| CN108985954A (en) | A kind of method and relevant device of incidence relation that establishing each mark | |
| CN108776692A (en) | Method and apparatus for handling information | |
| CN107908615A (en) | A kind of method and apparatus for obtaining search term corresponding goods classification | |
| CN109871791A (en) | Image processing method and device | |
| CN108960110A (en) | Method and apparatus for generating information | |
| CN108886512A (en) | The system and method classified automatically for application program network activity | |
| CN110349158A (en) | A kind of method and apparatus handling point cloud data | |
| CN110400201A (en) | Information displaying method, device, electronic equipment and medium | |
| CN108182457A (en) | For generating the method and apparatus of information | |
| CN109934194A (en) | Picture classification method, edge device, system and storage medium | |
| CN109284367A (en) | Method and apparatus for handling text | |
| CN109993749A (en) | The method and apparatus for extracting target image | |
| CN110427546A (en) | A kind of information displaying method and device | |
| CN113626624B (en) | Resource identification method and related device | |
| CN109885708A (en) | The searching method and device of certificate picture |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |