CN109325432A - A kind of recognition methods of three dimensional object, equipment and computer-readable storage media - Google Patents
A kind of recognition methods of three dimensional object, equipment and computer-readable storage media Download PDFInfo
- Publication number
- CN109325432A CN109325432A CN201811070706.XA CN201811070706A CN109325432A CN 109325432 A CN109325432 A CN 109325432A CN 201811070706 A CN201811070706 A CN 201811070706A CN 109325432 A CN109325432 A CN 109325432A
- Authority
- CN
- China
- Prior art keywords
- feature point
- fisrt feature
- image
- point
- fisrt
- 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
- G06V20/60—Type of objects
- G06V20/64—Three-dimensional objects
Abstract
The embodiment of the invention discloses a kind of recognition methods of three dimensional object, which comprises based on the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of the fisrt feature point and obtains second feature point set；Wherein, the fisrt feature point is the characteristic point for each image that pre-set image is concentrated；Obtain the parameter of the fisrt feature point；Wherein, the parameter is used to characterize the relevance between the pre-set image collection of the fisrt feature point；Based on the parameter, key feature points corresponding with the fisrt feature point are determined from the three-dimensional feature point of the second feature point set；Second feature point and the key feature points based on image to be processed, identify the image to be processed.The embodiment of the present invention also discloses the identification equipment and computer-readable storage media of a kind of three dimensional object simultaneously.
Description
Technical field
The present invention relates to augmented reality (Augmented Reality, AR) technical field more particularly to a kind of three dimensional objects
Recognition methods, equipment and computer-readable storage media.
Background technique
AR technology be it is a kind of by video camera in the certain time and spatial dimension of real world entity information carry out
Scanning analog simulation and is superimposed again in a computer, virtual information is applied to real world later, is surmounted to reach
The technology of the sensory experience of reality.The identification of three dimensional object is generally comprised using AR technology: feature extraction, characteristic matching and
Camera is calculated relative to the pose of three dimensional object；That is, needing to extract when identifying three dimensional object using AR technology
Characteristic point is matched with preset characteristic point.
But in the related art, all characteristic points and own that needs will extract when identifying to three dimensional object
Default characteristic point is matched, and required matched characteristic point data amount is larger, causes the matching speed of characteristic point slower, and is known
Other inefficiency.
Summary of the invention
In order to solve the above technical problems, an embodiment of the present invention is intended to provide a kind of recognition methods of three dimensional object, equipment and
Computer-readable storage media solves all features that needs will extract when identifying in relative skill to three dimensional object
Point is matched with all default characteristic points, and the larger problem of required matched characteristic point data amount improves characteristic point
Matching speed, and improve the efficiency of image recognition.
The technical scheme of the present invention is realized as follows:
A kind of recognition methods of three dimensional object, which comprises
Based on the fisrt feature point that preset fisrt feature point is concentrated, the corresponding three-dimensional feature of the fisrt feature point is obtained
Point obtains second feature point set；Wherein, the fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of the fisrt feature point；Wherein, the parameter be used to characterize the fisrt feature point with it is described
Relevance between pre-set image collection；
Based on the parameter, determination is corresponding with the fisrt feature point from the three-dimensional feature point of the second feature point set
Key feature points；
Second feature point and the key feature points based on image to be processed, identify the image to be processed.
Optionally, the parameter for obtaining the fisrt feature point, comprising:
Calculating the fisrt feature point, to concentrate the identical fisrt feature point to concentrate in the fisrt feature point shared
Specific gravity；Wherein, first parameter includes specific gravity.
Optionally, it is described based on the parameter, it is determining with described the from the three-dimensional feature point of the second feature point set
The corresponding key feature points of one characteristic point, comprising:
The three-dimensional feature point for obtaining the specific gravity and being greater than preset value is concentrated from the second feature point, is obtained described crucial special
Sign point.
Optionally, it is described based on the parameter, it is determining with described the from the three-dimensional feature point of the second feature point set
The corresponding key feature points of one characteristic point, further includes:
It sorts according to the specific gravity to the three-dimensional feature point, obtains the first sequence；
Based on first sequence, the key feature points are determined from the three-dimensional feature point of the second feature point set.
Optionally, described based on first sequence, from the three-dimensional feature point of the second feature point set described in determination
Key feature points, comprising:
Based on first sequence, is concentrated from the second feature point and obtain the three-dimensional spy of the specific gravity within a preset range
Sign point obtains the key feature points.
Optionally, the second feature point based in image to be processed and the key feature points, identification are described wait locate
Manage image, comprising:
Based on the specific gravity, the second sequence is determined；
According to second sequence, the characteristic value of the second feature point and the characteristic value of the key feature points are carried out
Matching；
Based on matching result, the image to be processed is identified.
Optionally, the identical fisrt feature point of the fisrt feature point concentration that calculates is in the fisrt feature point
Concentrate shared specific gravity, comprising:
Obtain the first quantity of the frame for the image that the pre-set image is concentrated；
Calculating includes the second quantity of the frame of identical fisrt feature point；
The ratio for calculating second quantity Yu first quantity, obtains the specific gravity.
A kind of identification equipment of three dimensional object, the equipment include: processor, reservoir and versabus；Wherein,
The versabus is for realizing the communication connection between processor and reservoir；
The processor is used to execute the recognizer of the three dimensional object stored in the reservoir, to realize following step
It is rapid:
Based on the fisrt feature point that preset fisrt feature point is concentrated, the corresponding three-dimensional feature of the fisrt feature point is obtained
Point obtains second feature point set；Wherein, the fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of the fisrt feature point；Wherein, the parameter is used to characterize the quantity of the fisrt feature point
How much；
Based on the parameter, it is concentrated from the second feature point and determines key feature corresponding with the fisrt feature point
Point；
Second feature point and the key feature points based on image to be processed, identify the image to be processed.
Optionally, the processor is for executing the data processor stored in the reservoir, can also realize with
Lower step:
Calculating the fisrt feature point, to concentrate the identical fisrt feature point to concentrate in the fisrt feature point shared
Specific gravity；Wherein, first parameter includes specific gravity.
A kind of computer-readable storage media, the computer-readable storage media store one or more programs, institute
To state one or more programs can be executed by one or more processors, to realize the recognition methods of three dimensional object described above
Step.
Recognition methods, equipment and the computer-readable storage media of three dimensional object provided by the embodiment of the present invention, base
In the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of fisrt feature point and obtain second feature
Point set, fisrt feature point are the characteristic points for each image that pre-set image is concentrated, and obtain fisrt feature point is used for characterization first
The parameter of the relevance between pre-set image collection of characteristic point, it is true from the three-dimensional feature point of second feature point set based on parameter
Fixed key feature points corresponding with fisrt feature point, in this way, being by all characteristic points of image to be processed carrying out image recognition
It is matched with the key feature points for predefining the part come out, rather than will be image to be processed as in relative skill
All characteristic points are matched with all characteristic points, need to carry out the quantity of matched characteristic point than the reduction in relative skill
, preset to solve all characteristic points for needing to extract when identifying three dimensional object in relative skill with all
Characteristic point is matched, and the larger problem of required matched characteristic point data amount improves the matching speed of characteristic point, and mention
The high efficiency of image recognition.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the recognition methods for three dimensional object that the embodiment of the present invention provides；
Fig. 2 is the flow diagram of the recognition methods for another three dimensional object that the embodiment of the present invention provides；
Fig. 3 is the flow diagram of the recognition methods for another three dimensional object that the embodiment of the present invention provides；
Fig. 4 is a kind of structural schematic diagram of the identification equipment for three dimensional object that the embodiment of the present invention provides.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description.
The embodiment of the present invention provides a kind of recognition methods of three dimensional object, and shown referring to Fig.1, this method includes following step
It is rapid:
Step 101, the fisrt feature point concentrated based on preset fisrt feature point obtain the corresponding three-dimensional of fisrt feature point
Characteristic point obtains second feature point set.
Wherein, fisrt feature point is the characteristic point for each image that pre-set image is concentrated.
It should be noted that the fisrt feature point that step 101 is concentrated based on preset fisrt feature point, obtains fisrt feature
The corresponding three-dimensional feature point of point obtains second feature point set can be realized by the identification equipment of three dimensional object；The three dimensional object
Identification equipment can have the equipment for being able to carry out image recognition of camera；For example, the identification equipment of the three dimensional object can be with
It is the electronic equipment etc. with camera.
Fisrt feature point concentrates all characteristic points for all images concentrated including pre-set image, which is
The set of all characteristic points of all images of characteristic point.It includes the corresponding three-dimensional of each fisrt feature point that second feature point, which is concentrated,
Characteristic point；Certainly, second feature point set can be each fisrt feature point is converted to obtain after three-dimensional feature point it is all
The set of the corresponding three-dimensional feature point of fisrt feature point.
Step 102, the parameter for obtaining fisrt feature point.
Wherein, the parameter of fisrt feature point is used to characterize the relevance between pre-set image collection of fisrt feature point.
In other embodiments of the invention, the parameter that step 102 obtains fisrt feature point can be by the knowledge of three dimensional object
Other equipment is realized；The parameter of fisrt feature point can be for characterizing each of fisrt feature point and pre-set image collection figure
As the parameter of the relationship between corresponding fisrt feature point set.
Step 103, the parameter based on fisrt feature point determine and the first spy from the three-dimensional feature point of second feature point set
The corresponding key feature points of sign point.
Wherein, parameter of the step 103 based on fisrt feature point determines and the from the three-dimensional feature point of second feature point set
The corresponding key feature points of one characteristic point can be realized by the identification equipment of three dimensional object；In the ginseng for getting fisrt feature point
After number, it can be concentrated according to the parameter from obtained second feature point and obtain key feature points.Key feature points are second feature
Point concentrates characteristic point corresponding with the parameter of fisrt feature point in the characteristic point for including；That is, key feature points are second
Partial Feature point in the characteristic point for including in feature point set.
Step 104, second feature point and key feature points based on image to be processed, identify image to be processed.
Wherein, second feature point and key feature points of the step 104 based on image to be processed identifies that image to be processed can be with
It is realized by the identification equipment of three dimensional object；The identification of image to be processed can be by by the second feature of image to be processed
Point is matched to realize with key feature points.
The recognition methods of three dimensional object provided by the embodiment of the present invention, the concentrated based on preset fisrt feature point
One characteristic point obtains the corresponding three-dimensional feature point of fisrt feature point and obtains second feature point set, and fisrt feature point is pre-set image
Concentrate each image characteristic point, obtain fisrt feature point for characterize fisrt feature point between pre-set image collection
The parameter of relevance is determined from the three-dimensional feature point of second feature point set corresponding with fisrt feature point crucial special based on parameter
Point is levied, in this way, being the key that by all characteristic points of image to be processed and to predefine the part come out carrying out image recognition
Characteristic point is matched, rather than carries out all characteristic points of image to be processed and all characteristic points as in relative skill
Matching, needs the quantity for carrying out matched characteristic point than reducing in relative skill, to solve in relative skill to three
It needs to match all characteristic points extracted with all default characteristic points when dimensional object is identified, it is required matched
The larger problem of characteristic point data amount, improves the matching speed of characteristic point, and improves the efficiency of image recognition.
Based on previous embodiment, the embodiment of the present invention provides a kind of recognition methods of three dimensional object, referring to shown in Fig. 2,
Method includes the following steps:
The fisrt feature point that step 201, the identification equipment of three dimensional object are concentrated based on preset fisrt feature point obtains the
The corresponding three-dimensional feature point of one characteristic point obtains second feature point set.
Wherein, fisrt feature point is the characteristic point for each image that pre-set image is concentrated.
Step 202, the identification equipment of three dimensional object calculate fisrt feature point and concentrate identical fisrt feature o'clock in the first spy
Sign point concentrates shared specific gravity.
Concentrate identical fisrt feature point in fisrt feature point set it should be noted that step 202 calculates fisrt feature point
In shared specific gravity can be accomplished by the following way:
A, the first quantity of the frame for the image that pre-set image is concentrated is obtained.
Wherein, the first quantity also refers to the quantity that pre-set image concentrates its frame of each image for including
Total quantity after adding up refers to the quantity for all frames that all images that pre-set image is concentrated include.For example, if pre-
If the total quantity that all frames for all images for including in image set add up is M, then the first quantity is exactly M.
B, calculating includes the second quantity of the frame of identical fisrt feature point.
Wherein, the second quantity also refers to pre-set image and concentrates the first spy that the corresponding image of frame includes in all frames
The quantity of the identical frame of sign point.For example, if pre-set image concentrates the first spy for having the corresponding image of N number of frame to include in all frames
Sign point is identical, then the second quantity is exactly N.It should be noted that M and N is positive integer, and N is less than or equal to M.
C, the ratio for calculating the second quantity and the first quantity, obtains specific gravity.
Wherein, specific gravity is with the second quantity divided by the calculated result obtained after the first quantity.Also, the specific gravity obtained at this time
Just refer to that fisrt feature point concentrates identical fisrt feature point to concentrate shared specific gravity in fisrt feature point.In addition, if the
In one feature point set when fisrt feature point identical including multiple groups, then multiple specific gravity will be obtained.
Step 203, the identification equipment of three dimensional object are concentrated from second feature point obtains the three-dimensional spy that specific gravity is greater than preset value
Point is levied, key feature points are obtained.
Wherein, preset value can be preset value, and the preset value can be according to actual demand and history
Use data setting.
Step 204, the identification equipment of three dimensional object are concentrated based on identical fisrt feature point in fisrt feature point shared
Specific gravity determines the second sequence.
Wherein, it concentrates identical fisrt feature point in the size of wherein shared specific gravity according to fisrt feature point, is pressed
It is ranked up to obtain the second sequence according to sequence from big to small or from small to large.
Step 205, the identification equipment of three dimensional object are according to the second sequence, by the characteristic value and key feature of second feature point
The characteristic value of point is matched.
Wherein, after obtaining the second sequence, according to the sequencing in second sequence, will first be sorted in the second sequence
The characteristic value of one second feature point is matched with the characteristic value of key feature points, later by the second of father's sequence second place
The characteristic value of characteristic point is matched with the characteristic value of key feature points, and so on until all second feature points feature
Value matches completion with the characteristic value of key feature points.
Step 206, the identification equipment of three dimensional object are based on matching result, identify image to be processed.
Wherein, in the matching result of the characteristic value for the characteristic value and key feature points for getting each second feature point
Afterwards, it is identified to obtain image to be processed according to the key feature points of successful match.
The recognition methods of three dimensional object provided by the embodiment of the present invention, the concentrated based on preset fisrt feature point
One characteristic point obtains the corresponding three-dimensional feature point of fisrt feature point and obtains second feature point set, and fisrt feature point is pre-set image
Concentrate each image characteristic point, obtain fisrt feature point for characterize fisrt feature point between pre-set image collection
The parameter of relevance is determined from the three-dimensional feature point of second feature point set corresponding with fisrt feature point crucial special based on parameter
Point is levied, in this way, being the key that by all characteristic points of image to be processed and to predefine the part come out carrying out image recognition
Characteristic point is matched, rather than carries out all characteristic points of image to be processed and all characteristic points as in relative skill
Matching, needs the quantity for carrying out matched characteristic point than reducing in relative skill, to solve in relative skill to three
It needs to match all characteristic points extracted with all default characteristic points when dimensional object is identified, it is required matched
The larger problem of characteristic point data amount, improves the matching speed of characteristic point, and improves the efficiency of image recognition.
Based on previous embodiment, the embodiment of the present invention provides a kind of recognition methods of three dimensional object, referring to shown in Fig. 2,
Method includes the following steps:
The fisrt feature point that step 301, the identification equipment of three dimensional object are concentrated based on preset fisrt feature point obtains the
The corresponding three-dimensional feature point of one characteristic point obtains second feature point set.
Wherein, fisrt feature point is the characteristic point for each image that pre-set image is concentrated.
Step 302, the identification equipment of three dimensional object calculate fisrt feature point and concentrate identical fisrt feature o'clock in the first spy
Sign point concentrates shared specific gravity.
Concentrate identical fisrt feature point in fisrt feature point set it should be noted that step 302 calculates fisrt feature point
In shared specific gravity can be accomplished by the following way:
A, the first quantity of the frame for the image that pre-set image is concentrated is obtained.
B, calculating includes the second quantity of the frame of identical fisrt feature point.
C, the ratio for calculating the second quantity and the first quantity, obtains specific gravity.
Step 303, the identification equipment of three dimensional object sort to three-dimensional feature point according to specific gravity, obtain the first sequence.
Wherein, concentrate identical fisrt feature point wherein after shared specific gravity obtaining fisrt feature point, according to arriving greatly
The three-dimensional feature point of second feature point set is ranked up by sequence small or from small to large；That is, what the first sequence referred to
It is putting in order for the three-dimensional feature point of second feature point set.
Step 304, the identification equipment of three dimensional object are based on the first sequence, from the three-dimensional feature point of second feature point set really
Determine key feature points.
Wherein, step 304 is based on the first sequence, and key feature points are determined from the three-dimensional feature point of second feature point set,
It can be accomplished by the following way:
Based on the first sequence, the three-dimensional feature point of acquisition specific gravity within a preset range is concentrated to obtain key from second feature point
Characteristic point.
Wherein, preset range can be preset, and the preset range can be according to actual demand and go through
History uses data setting.In a kind of feasible implementation, preset range also refers to preset numberical range；It is obtaining
When taking key feature points, the sequence for the three-dimensional feature point concentrated according to second feature point can be, successively from second feature point set
The middle three-dimensional feature point for obtaining specific gravity within the scope of default value.
Step 305, the identification equipment of three dimensional object are concentrated based on identical fisrt feature point in fisrt feature point shared
Specific gravity determines the second sequence.
Step 306, the identification equipment of three dimensional object are according to the second sequence, by the characteristic value and key feature of second feature point
The characteristic value of point is matched.
Step 307, the identification equipment of three dimensional object are based on matching result, identify image to be processed.
It should be noted that in the present embodiment with the explanation of same steps in other embodiments and identical content, Ke Yican
According to the description in other embodiments, details are not described herein again.
The recognition methods of three dimensional object provided by the embodiment of the present invention, the concentrated based on preset fisrt feature point
One characteristic point obtains the corresponding three-dimensional feature point of fisrt feature point and obtains second feature point set, and fisrt feature point is pre-set image
Concentrate each image characteristic point, obtain fisrt feature point for characterize fisrt feature point between pre-set image collection
The parameter of relevance is determined from the three-dimensional feature point of second feature point set corresponding with fisrt feature point crucial special based on parameter
Point is levied, in this way, being the key that by all characteristic points of image to be processed and to predefine the part come out carrying out image recognition
Characteristic point is matched, rather than carries out all characteristic points of image to be processed and all characteristic points as in relative skill
Matching, needs the quantity for carrying out matched characteristic point than reducing in relative skill, to solve in relative skill to three
It needs to match all characteristic points extracted with all default characteristic points when dimensional object is identified, it is required matched
The larger problem of characteristic point data amount, improves the matching speed of characteristic point, and improves the efficiency of image recognition.
Based on previous embodiment, the embodiment of the present invention provides a kind of identification equipment of three dimensional object, which can answer
In a kind of recognition methods of the three dimensional object provided for the corresponding embodiment in Fig. 1~3, referring to shown in Fig. 4, which can be with
It include: processor 41, reservoir 42 and versabus 43；Wherein,
Versabus 43 is for realizing the communication connection between processor 41 and reservoir 42；
Processor 41 is used to execute the recognizer of the three dimensional object stored in reservoir, to perform the steps of
Based on the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of fisrt feature point and obtain
To second feature point set；
Wherein, fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of fisrt feature point；
Wherein, parameter be used for characterize fisrt feature point quantity number；
Based on parameter, is concentrated from second feature point and determine key feature points corresponding with fisrt feature point；
Second feature point and key feature points based on image to be processed, identify image to be processed.
In other embodiments of the invention, processor is used to execute the ginseng of the acquisition fisrt feature point stored in reservoir
Number, may be implemented following steps:
Calculating fisrt feature point concentrates identical fisrt feature point to concentrate shared specific gravity in fisrt feature point；
Wherein, the parameter of fisrt feature point includes specific gravity.
In other embodiments of the invention, processor be used to execute stored in reservoir based on parameter, from the second spy
It levies and determines key feature points corresponding with fisrt feature point in the three-dimensional feature point of point set, following steps may be implemented:
The three-dimensional feature point for obtaining specific gravity and being greater than preset value is concentrated from second feature point, is obtained corresponding with fisrt feature point
Key feature points.
In other embodiments of the invention, processor be used to execute stored in reservoir based on parameter, from the second spy
It levies and determines key feature points corresponding with fisrt feature point in the three-dimensional feature point of point set, following steps may be implemented:
It sorts according to specific gravity to three-dimensional feature point, obtains the first sequence；
Based on the first sequence, determined from the three-dimensional feature point of second feature point set corresponding with fisrt feature point crucial special
Sign point.
In other embodiments of the invention, processor be used to execute stored in reservoir based on the first sequence, from the
Key feature points corresponding with fisrt feature point are determined in the three-dimensional feature point of two feature point sets, and following steps may be implemented:
Based on the first sequence, is concentrated from second feature point and obtain specific gravity three-dimensional feature point within a preset range and obtain and the
The corresponding key feature points of one characteristic point.
In other embodiments of the invention, processor is used to execute the based on image to be processed stored in reservoir
Two characteristic points and key feature points identify image to be processed, following steps may be implemented:
Shared specific gravity is concentrated in fisrt feature point based on identical fisrt feature point, determines the second sequence；
According to the second sequence, the characteristic value of second feature point is matched with the characteristic value of key feature points；
Based on matching result, image to be processed is identified.
In other embodiments of the invention, processor is used to execute the calculating fisrt feature point stored in reservoir and concentrates
Identical fisrt feature point concentrates shared specific gravity in fisrt feature point, and following steps may be implemented:
Obtain the first quantity of the frame for the image that pre-set image is concentrated；
Calculating includes the second quantity of the frame of identical fisrt feature point；
The ratio for calculating the second quantity Yu the first quantity obtains the specific gravity.
It should be noted that in the present embodiment step performed by processor specific implementation process, be referred to Fig. 1~
Realization process in the recognition methods for the three dimensional object that 3 corresponding embodiments provide, details are not described herein again.
The identification equipment of three dimensional object provided by the embodiment of the present invention is by image to be processed carrying out image recognition
All characteristic points matched with the key feature points of part come out are predefined, rather than the general as in relative skill
All characteristic points of image to be processed are matched with all characteristic points, need to carry out the quantity of matched characteristic point than opposite skill
Reducing in art, to solve all characteristic points that needs will extract when identifying in relative skill to three dimensional object
It is matched with all default characteristic points, the larger problem of required matched characteristic point data amount improves of characteristic point
With speed, and improve the efficiency of image recognition.
Based on previous embodiment, the embodiment of the present invention provides a kind of computer readable storage medium, this is computer-readable
Storage medium stores one or more programs, which can be executed by one or more processors, to realize
Following steps:
Based on the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of fisrt feature point and obtain
To second feature point set；
Wherein, fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of fisrt feature point；
Wherein, parameter be used for characterize fisrt feature point quantity number；
Based on parameter, is concentrated from second feature point and determine key feature points corresponding with fisrt feature point；
Second feature point and key feature points based on image to be processed, identify image to be processed.
In other embodiments of the invention, which can be executed by one or more processors acquisition
The parameter of one characteristic point, may be implemented following steps:
Calculating fisrt feature point concentrates identical fisrt feature point to concentrate shared specific gravity in fisrt feature point；
Wherein, the parameter of fisrt feature point includes specific gravity.
In other embodiments of the invention, which can be executed by one or more processors based on ginseng
Number determines key feature points corresponding with fisrt feature point from the three-dimensional feature point of second feature point set, may be implemented following
Step:
The three-dimensional feature point for obtaining specific gravity and being greater than preset value is concentrated from second feature point, is obtained corresponding with fisrt feature point
Key feature points.
In other embodiments of the invention, which can be executed by one or more processors based on ginseng
Number determines key feature points corresponding with fisrt feature point from the three-dimensional feature point of second feature point set, may be implemented following
Step:
It sorts according to specific gravity to three-dimensional feature point, obtains the first sequence；
Based on the first sequence, determined from the three-dimensional feature point of second feature point set corresponding with fisrt feature point crucial special
Sign point.
In other embodiments of the invention, which can be executed by one or more processors based on
One sequence, key feature points corresponding with fisrt feature point are determined from the three-dimensional feature point of second feature point set, may be implemented
Following steps:
Based on the first sequence, is concentrated from second feature point and obtain specific gravity three-dimensional feature point within a preset range and obtain and the
The corresponding key feature points of one characteristic point.
In other embodiments of the invention, the one or more program can be executed by one or more processors based on to
The second feature point and key feature points for handling image, identify image to be processed, following steps may be implemented:
Shared specific gravity is concentrated in fisrt feature point based on identical fisrt feature point, determines the second sequence；
According to the second sequence, the characteristic value of second feature point is matched with the characteristic value of key feature points；
Based on matching result, image to be processed is identified.
In other embodiments of the invention, which can be executed by one or more processors calculating
Identical fisrt feature point concentrates shared specific gravity in fisrt feature point in one feature point set, and following steps may be implemented:
Obtain the first quantity of the frame for the image that pre-set image is concentrated；
Calculating includes the second quantity of the frame of identical fisrt feature point；
The ratio for calculating the second quantity Yu the first quantity obtains the specific gravity.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, the shape of hardware embodiment, software implementation or embodiment combining software and hardware aspects can be used in the present invention
Formula.Moreover, the present invention, which can be used, can use storage in the computer that one or more wherein includes computer usable program code
The form for the computer program product implemented on medium (including but not limited to magnetic disk storage and optical memory etc.).
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention.
Claims (10)
1. a kind of recognition methods of three dimensional object, which comprises
Based on the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of the fisrt feature point and obtain
To second feature point set；Wherein, the fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of the fisrt feature point；Wherein, the parameter is used to characterize presetting with described for the fisrt feature point
Relevance between image set；
Based on the parameter, pass corresponding with the fisrt feature point is determined from the three-dimensional feature point of the second feature point set
Key characteristic point；
Second feature point and the key feature points based on image to be processed, identify the image to be processed.
2. the method according to claim 1, wherein the parameter for obtaining the fisrt feature point, comprising:
Calculating the fisrt feature point concentrates the identical fisrt feature point to concentrate shared specific gravity in the fisrt feature point；
Wherein, the parameter includes specific gravity.
3. according to the method described in claim 2, it is characterized in that, it is described based on the parameter, from the second feature point set
Three-dimensional feature point in determine corresponding with fisrt feature point key feature points, comprising:
The three-dimensional feature point for obtaining the specific gravity and being greater than preset value is concentrated from the second feature point, obtains the key feature
Point.
4. according to the method described in claim 2, it is characterized in that, it is described based on the parameter, from the second feature point set
Three-dimensional feature point in determine corresponding with fisrt feature point key feature points, further includes:
It sorts according to the specific gravity to the three-dimensional feature point, obtains the first sequence；
Based on first sequence, the key feature points are determined from the three-dimensional feature point of the second feature point set.
5. according to the method described in claim 4, it is characterized in that, described be based on first sequence, from the second feature
The key feature points are determined in the three-dimensional feature point of point set, comprising:
Based on first sequence, is concentrated from the second feature point and obtain the three-dimensional feature point of the specific gravity within a preset range
Obtain the key feature points.
6. according to the method described in claim 2, it is characterized in that, the second feature point and institute based in image to be processed
Key feature points are stated, identify the image to be processed, comprising:
Based on the specific gravity, the second sequence is determined；
According to second sequence, by the characteristic value of the second feature point and the progress of the characteristic value of the key feature points
Match；
Based on matching result, the image to be processed is identified.
7. according to the described in any item methods of claim 2-6, which is characterized in that described to calculate the fisrt feature point concentration phase
The same fisrt feature point concentrates shared specific gravity in the fisrt feature point, comprising:
Obtain the first quantity of the frame for the image that the pre-set image is concentrated；
Calculating includes the second quantity of the frame of identical fisrt feature point；
The ratio for calculating second quantity Yu first quantity, obtains the specific gravity.
8. a kind of identification equipment of three dimensional object, which is characterized in that the equipment includes: processor, reservoir and versabus；
Wherein,
The versabus is for realizing the communication connection between processor and reservoir；
The processor is used to execute the recognizer of the three dimensional object stored in the reservoir, to perform the steps of
Based on the fisrt feature point that preset fisrt feature point is concentrated, obtains the corresponding three-dimensional feature point of the fisrt feature point and obtain
To second feature point set；Wherein, the fisrt feature point is the characteristic point for each image that pre-set image is concentrated；
Obtain the parameter of the fisrt feature point；Wherein, the parameter be used for characterize the fisrt feature point quantity number；
Based on the parameter, it is concentrated from the second feature point and determines key feature points corresponding with the fisrt feature point；
Second feature point and the key feature points based on image to be processed, identify the image to be processed.
9. equipment according to claim 8, which is characterized in that the processor is used to execute to store in the reservoir
Data processor can also perform the steps of
Calculating the fisrt feature point concentrates the identical fisrt feature point to concentrate shared specific gravity in the fisrt feature point；
Wherein, first parameter includes specific gravity.
10. a kind of computer-readable storage media, which is characterized in that the computer-readable storage media stores one or more
A program, one or more of programs can be executed by one or more processors, to realize as any in claim 1 to 7
The step of recognition methods of three dimensional object described in.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811070706.XA CN109325432B (en) | 2018-09-13 | 2018-09-13 | Three-dimensional object identification method and equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811070706.XA CN109325432B (en) | 2018-09-13 | 2018-09-13 | Three-dimensional object identification method and equipment and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109325432A true CN109325432A (en) | 2019-02-12 |
| CN109325432B CN109325432B (en) | 2021-11-16 |
Family
ID=65265340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811070706.XA Active CN109325432B (en) | 2018-09-13 | 2018-09-13 | Three-dimensional object identification method and equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325432B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114098536A (en) * | 2021-12-01 | 2022-03-01 | 湖南格兰博智能科技有限责任公司 | Obstacle crossing and trapped detection method for sweeping robot and sweeping robot |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103297811A (en) * | 2012-02-24 | 2013-09-11 | 北京明日时尚信息技术有限公司 | Method for realizing video advertisement in intelligently embedding mode |
| CN105138525A (en) * | 2014-05-30 | 2015-12-09 | 株式会社日立制作所 | Traffic video processing device and method, and retrieval device and method |
| CN105913416A (en) * | 2016-04-06 | 2016-08-31 | 中南大学 | Method for automatically segmenting three-dimensional human face model area |
| CN106022210A (en) * | 2016-05-04 | 2016-10-12 | 成都指码科技有限公司 | Vein profile three-dimensional point cloud matching identity identifying method and device |
| US9652896B1 (en) * | 2015-10-30 | 2017-05-16 | Snap Inc. | Image based tracking in augmented reality systems |
-
2018
- 2018-09-13 CN CN201811070706.XA patent/CN109325432B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103297811A (en) * | 2012-02-24 | 2013-09-11 | 北京明日时尚信息技术有限公司 | Method for realizing video advertisement in intelligently embedding mode |
| CN105138525A (en) * | 2014-05-30 | 2015-12-09 | 株式会社日立制作所 | Traffic video processing device and method, and retrieval device and method |
| US9652896B1 (en) * | 2015-10-30 | 2017-05-16 | Snap Inc. | Image based tracking in augmented reality systems |
| CN105913416A (en) * | 2016-04-06 | 2016-08-31 | 中南大学 | Method for automatically segmenting three-dimensional human face model area |
| CN106022210A (en) * | 2016-05-04 | 2016-10-12 | 成都指码科技有限公司 | Vein profile three-dimensional point cloud matching identity identifying method and device |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114098536A (en) * | 2021-12-01 | 2022-03-01 | 湖南格兰博智能科技有限责任公司 | Obstacle crossing and trapped detection method for sweeping robot and sweeping robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109325432B (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200356818A1 (en) | Logo detection | |
| CN104050449B (en) | A kind of face identification method and device | |
| CN107103613B (en) | A kind of three-dimension gesture Attitude estimation method | |
| CN109034159A (en) | image information extracting method and device | |
| CN109816769A (en) | Scene based on depth camera ground drawing generating method, device and equipment | |
| CN110139159A (en) | Processing method, device and the storage medium of video material | |
| CN109977983B (en) | Method and device for obtaining training image | |
| CN110147721A (en) | A kind of three-dimensional face identification method, model training method and device | |
| CN109117897A (en) | Image processing method, device and readable storage medium storing program for executing based on convolutional neural networks | |
| CN112070782B (en) | Method, device, computer readable medium and electronic equipment for identifying scene contour | |
| CN104517113B (en) | The sorting technique and relevant apparatus of a kind of feature extracting method of image, image | |
| CN107784321B (en) | Method and system for quickly identifying digital picture books and computer readable storage medium | |
| CN111667005B (en) | Human interactive system adopting RGBD visual sensing | |
| CN109766873A (en) | A kind of pedestrian mixing deformable convolution recognition methods again | |
| CN110298296A (en) | Face identification method applied to edge calculations equipment | |
| CN107886110A (en) | Method for detecting human face, device and electronic equipment | |
| CN103927530A (en) | Acquiring method, application method and application system of final classifier | |
| CN103955713B (en) | A kind of icon-based programming method and apparatus | |
| CN108021693A (en) | A kind of image search method and device | |
| CN109325432A (en) | A kind of recognition methods of three dimensional object, equipment and computer-readable storage media | |
| CN111079535B (en) | Human skeleton action recognition method and device and terminal | |
| CN109816709B (en) | Monocular camera-based depth estimation method, device and equipment | |
| CN110782390A (en) | Image correction processing method and device and electronic equipment | |
| CN108989681A (en) | Panorama image generation method and device | |
| CN101105769A (en) | Chip validation pretreatment method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |