CN106808475B - A kind of intelligent robot that visual performance is excellent - Google Patents
A kind of intelligent robot that visual performance is excellent Download PDFInfo
- Publication number
- CN106808475B CN106808475B CN201710184827.6A CN201710184827A CN106808475B CN 106808475 B CN106808475 B CN 106808475B CN 201710184827 A CN201710184827 A CN 201710184827A CN 106808475 B CN106808475 B CN 106808475B
- Authority
- CN
- China
- Prior art keywords
- color
- image
- block
- sub
- indicate
- 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.)
- Active
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
- B25J9/1697—Vision controlled systems
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J11/00—Manipulators not otherwise provided for
- B25J11/008—Manipulators for service tasks
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Image Analysis (AREA)
- Manipulator (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of intelligent robots that visual performance is excellent, including power module, sighting device, control system and robot body, the power module is used to power to the sighting device and control system, the sighting device is for obtaining target image, and recognition result is exported, the control system, which is used to control robot body according to the recognition result, makes corresponding actions.The invention has the benefit that realizing the intelligence and effectively control of robot.
Description
Technical field
The present invention relates to robotic technology fields, and in particular to a kind of intelligent robot that visual performance is excellent.
Background technique
Intelligent robot is as a kind of technology comprising quite a lot of scientific knowledge, almost along with produced by artificial intelligence
's.And intelligent robot becomes more and more important in today's society, more and more fields and post require intelligent robot
It participates in, this makes the research of intelligent robot also more and more frequent.In the near future, continuous with intelligent robot technology
Development and maturation, with the unremitting effort of numerous scientific research personnel, intelligent robot will come into huge numbers of families, preferably service people
Life, make people's lives more comfortable and health.
However, existing robot is mostly mankind's service just for specific occasion, and robot lacks sighting device,
Intelligence degree is low, poor controllability, is to provide service for specific occasion for the mankind mostly, the function of providing is relatively single
One.
Summary of the invention
In view of the above-mentioned problems, the present invention is intended to provide a kind of intelligent robot that visual performance is excellent.
The purpose of the present invention is realized using following technical scheme:
Provide a kind of intelligent robot that visual performance is excellent, including power module, sighting device, control system and machine
Device human body, the power module are used to power to the sighting device and control system, and the sighting device is for obtaining mesh
Logo image, and recognition result is exported, the control system is used to control robot body according to the recognition result and make accordingly
Movement.
The invention has the benefit that realizing the intelligence and effectively control of robot.
Detailed description of the invention
The present invention will be further described with reference to the accompanying drawings, but the embodiment in attached drawing is not constituted to any limit of the invention
System, for those of ordinary skill in the art, without creative efforts, can also obtain according to the following drawings
Other attached drawings.
Fig. 1 is structure connection diagram of the invention.
Appended drawing reference:
Power module 1, sighting device 2, control system 3, robot body 4, storage device 5.
Specific embodiment
The invention will be further described with the following Examples.
Referring to Fig. 1, a kind of excellent intelligent robot of visual performance of the present embodiment, including power module 1, sighting device
2, control system 3 and robot body 4, the power module 1 is used to power to the sighting device 2 and control system 3, described
Sighting device 2 exports recognition result for obtaining target image, and the control system 3 is used for according to the recognition result control
Robot body 4 processed makes corresponding actions.
The present embodiment realizes the intelligence and effectively control of robot.
It preferably, further include storage device 5, the target image obtained for storing the sighting device.
This preferred embodiment realizes the storage of target image.
Preferably, the power module 1 is battery.
This preferred embodiment robot is not necessarily to wiring, and movement is more convenient, improves customer experience.
Preferably, the sighting device 2 is for identifying target, including image capture module, single treatment module,
Secondary treatment module and visual identity module, described image acquisition module is for obtaining target image, the single treatment module
For extracting the color characteristic of target image, the secondary treatment module obtains color histogram, institute according to the color characteristic
Visual identity module is stated for carrying out tax power to the color histogram, and according to the entitled color histogram to the mesh
Logo image is identified.
The present embodiment intelligent robot can accurately identify the target in image.
Preferably, the single treatment module includes the first conversion unit and the second cutting unit, and first conversion is single
Member is for being transformed into CIELab color space, the conversion formula from RGB color for image are as follows:
In above-mentioned formula, EH, EM, CS are respectively RGB face
Red, green in the colour space, blue color component value, L are the brightness in CIELab color space, and a is in CIELab color space
Green to red relative colorimetric, b is relative colorimetric of the blue in CIELab color space to yellow, wherein functionSecond cutting unit is for dividing an image into equal-sized rectangle
Sub-block, the image I for dividing sub-block are indicated are as follows:In above-mentioned formula, UiIt indicates
Any sub-block of image, ZC indicate the image segmentation factor, and ZC ∈ [2,5] and ZC are integer, and i is according to from left to right, from the top down
Sequence successively value be 1 arrive ZC2。
Target image is transformed by this preferred embodiment intelligent robot by single treatment module more meets human vision
The CIELab color space of feature can more accurately reflect the vision difference degree between different color, by figure
As being divided and being set the image segmentation factor, image recognition accuracy and recognition efficiency can be taken into account, intelligence is further improved
The service level of energy robot.
Preferably, the secondary treatment module, specifically: step 1: being divided to CIELab color space, using such as
Lower division methods: when L * component is greater than threshold value T1When or be less than threshold value T2When, a component and b component are not considered further that, obtain 2 face
Color section, when L * component is between threshold value T1And T2Between when, a component and b component are divided into four sections respectively, obtain 16 face
Color section, so that CIELab color space has been divided into 18 color intervals；Wherein, T1∈ [90,100], T2∈[0,10]；
Step 2: defining subordinating degree function σj,k=1；Step 3: the color histogram of image is sought, the color histogram of image subblock
It may be expressed as: MX (Ui)={ z1,z2,…,z18, in above-mentioned formula, MX (Ui) indicate image subblock color histogram, zj(j
=1,2 ..., 18) the pixel distribution situation in any color interval is indicated,σj,kRepresent k-th of pixel
Belong to the degree of membership of j-th of color interval, NiIndicate the number of pixels that sub-block includes；The color histogram of image may be expressed as:In above-mentioned formula, δiIndicate the inverse of sub-block locations weight, whereinThe color histogram of MX (I) expression image subblock.
This preferred embodiment intelligent robot is believed by the spatial distribution that secondary treatment module has incorporated pixel color feature
Sub-block locations weight is ceased and be arranged, histogram that is more accurate and meeting Human Visual System is obtained, further improves view
Feel the expressive faculty of feature.
Preferably, the visual identity module, including the first computing unit, the second computing unit and image comparison unit,
First computing unit is used to calculate the color difference between pixel, calculates central pixel point pAWith 3 × 3 neighborhoodsInterior
Anticipate neighbor pixel pBColor difference RU:Above-mentioned formula
In son, RU (pA,pB) indicate pixel pAAnd pBBetween color difference, μ is normalization factor；Second computing unit is for calculating
The color weight of each sub-block；Described image comparison unit is used to realize image recognition according to image similarity comparison；It is described
The color weight of each sub-block is calculated, specifically includes the following steps: the first step, calculates the color complexity of each pixel, meter
Center pixel is calculated relative to 3 × 3 neighborhoodsThe color change of other interior 8 adjacent pixels, obtains central pixel point pAColor it is multiple
Miscellaneous degree FA:In above-mentioned formula, FAIndicate pixel pAColor complexity；Second step calculates every
The color weight of a sub-block obtains the color weight Q of sub-block by calculating each pixel color weight in any sub-blocki:In above-mentioned formula, UiIndicate any sub-block of image, ZXiIndicate the color weight of sub-block, Ni
Indicate the number of pixels that sub-block includes, γ indicates the color complexity standard deviation of all pixels point in sub-block, FAAnd FkIt is sub-block
In pixel；It is described that image recognition is realized according to image similarity comparison, specifically, according to the color weight of sub-block and directly
Side's figure defines two images I1And I2Similarity MH:In above-mentioned formula, MH (I1,I2) indicate two width
Image I1And I2Similarity,WithRespectively indicate image I1And I2J-th of color area of i-th of sub-block
Between pixel distribution situation, calculate images to be recognized and sample image similarity, choose the high sample image of similarity as knowing
Other result.
Color complexity is described in the visual identity module of this preferred embodiment intelligent robot, reflects vision system
Togetherness knows the sensitivity characteristic of different colours variation, according to the color weight of sub-block and histogram calculation recognisable image and sample image
Between similarity, improve intelligent robot to the identification precision of image.
Target is identified using intelligent robot of the present invention, it is when the image segmentation factor takes different value, identification is quasi-
True rate and recognition time are as evaluation criterion, and compared with ordinary robot, generation is had the beneficial effect that shown in table:
| ZC | Recognition accuracy improves | Recognition time shortens |
| 2 | 20% | 31% |
| 3 | 25% | 25% |
| 4 | 30% | 20% |
| 5 | 32% | 18% |
| 6 | 36% | 12% |
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected
The limitation of range is protected, although explaining in detail referring to preferred embodiment to the present invention, those skilled in the art are answered
Work as understanding, it can be with modification or equivalent replacement of the technical solution of the present invention are made, without departing from the reality of technical solution of the present invention
Matter and range.
Claims (3)
1. a kind of intelligent robot that visual performance is excellent, characterized in that including power module, sighting device, control system and
Robot body, the power module are used to power to the sighting device and control system, and the sighting device is for obtaining
Target image, and recognition result is exported, the control system, which is used to control robot body according to the recognition result, makes phase
It should act；It further include storage device, the target image obtained for storing the sighting device；
The power module is battery；
The sighting device is used to identify target, including image capture module, single treatment module, secondary treatment module
With visual identity module, described image acquisition module is for obtaining target image, and the single treatment module is for extracting target
The color characteristic of image, the secondary treatment module obtain color histogram, the visual identity mould according to the color characteristic
Block knows the target image according to the entitled color histogram for carrying out tax power to the color histogram
Not；
The single treatment module includes the first conversion unit and the second cutting unit, and first conversion unit is used for image
CIELab color space, the conversion formula are transformed into from RGB color are as follows: In above-mentioned formula, EH, EM, CS are respectively red, green, blue color component value in RGB color,
L is the brightness in CIELab color space, and a is relative colorimetric of the green to red in CIELab color space, b CIELab
The relative colorimetric of blue in color space to yellow, wherein functionIt is described
For dividing an image into equal-sized rectangular sub blocks, the image I for dividing sub-block is indicated second cutting unit are as follows: In above-mentioned formula, UiIndicate image any sub-block, ZC indicate image segmentation because
Son, ZC ∈ [2,5] and ZC are integer, and i is according to from left to right, and successively value is 1 to ZC to sequence from the top down2。
2. a kind of excellent intelligent robot of visual performance according to claim 1, characterized in that the secondary treatment mould
Block, specifically: step 1: being divided to CIELab color space, using following division methods: when L * component is greater than threshold value T1
When or be less than threshold value T2When, a component and b component are not considered further that, obtains 2 color intervals, when L * component is between threshold value T1And T2
Between when, a component and b component are divided into four sections respectively, obtain 16 color intervals, thus by CIELab color space
It has been divided into 18 color intervals；Wherein, T1∈ [90,100], T2∈ [0,10]；Step 2: defining subordinating degree function σJ, k=1；
Step 3: seeking the color histogram of image, the color histogram of image subblock may be expressed as: MX (Ui)={ z1, z2...,
z18, in above-mentioned formula, MX (Ui) indicate image subblock color histogram, zj(j=1,2 ..., 18) indicates any color area
Between on pixel distribution situation,σJ, kRepresent the degree of membership that k-th of pixel belongs to j-th of color interval, Ni
Indicate the number of pixels that sub-block includes；The color histogram of image may be expressed as:
In above-mentioned formula, δiIndicate the inverse of sub-block locations weight, whereinMX (I) indicates that the color of image subblock is straight
Fang Tu.
3. a kind of excellent intelligent robot of visual performance according to claim 2, characterized in that the visual identity mould
Block, including the first computing unit, the second computing unit and image comparison unit, first computing unit is for calculating pixel
Between color difference, calculate central pixel point pAWith 3 × 3 neighborhoodsInterior arbitrary neighborhood pixel pBColor difference RU: In above-mentioned formula, RU (pA, pB) indicate pixel pA
And pBBetween color difference, μ is normalization factor；Second computing unit is used to calculate the color weight of each sub-block；The figure
As comparison unit is used to realize image recognition according to image similarity comparison；The color weight for calculating each sub-block, tool
Body calculates the color complexity of each pixel the following steps are included: the first step, calculates center pixel relative to 3 × 3 neighborhoods
The color change of other interior 8 adjacent pixels, obtains central pixel point pAColor complexity FA:
In above-mentioned formula, FAIndicate pixel pAColor complexity；Second step calculates the color weight of each sub-block, in anyon
Block obtains the color weight Q of sub-block by calculating each pixel color weighti:Above-mentioned formula
In, UiIndicate any sub-block of image, ZXiIndicate the color weight of sub-block, NiIndicate the number of pixels that sub-block includes, γ is indicated
The color complexity standard deviation of all pixels point, F in sub-blockAAnd FkIt is the pixel in sub-block；It is described according to image similarity
Comparison is to realize image recognition, specifically, defining two images I according to the color weight of sub-block and histogram1And I2Similarity
MH:In above-mentioned formula, MH (I1, I2) indicate
Two images I1And I2Similarity,WithRespectively indicate image I1And I2J-th of color of i-th of sub-block
The pixel distribution situation in section calculates images to be recognized and sample image similarity, chooses the high sample image conduct of similarity
Recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710184827.6A CN106808475B (en) | 2017-03-24 | 2017-03-24 | A kind of intelligent robot that visual performance is excellent |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710184827.6A CN106808475B (en) | 2017-03-24 | 2017-03-24 | A kind of intelligent robot that visual performance is excellent |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106808475A CN106808475A (en) | 2017-06-09 |
| CN106808475B true CN106808475B (en) | 2019-05-17 |
Family
ID=59115344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710184827.6A Active CN106808475B (en) | 2017-03-24 | 2017-03-24 | A kind of intelligent robot that visual performance is excellent |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106808475B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109079825A (en) * | 2017-06-14 | 2018-12-25 | 天津玛斯特车身装备技术有限公司 | Robot automation's visual grasping system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101947788A (en) * | 2010-06-23 | 2011-01-19 | 焦利民 | Intelligent robot |
| CN104400785A (en) * | 2014-12-02 | 2015-03-11 | 湖南城市学院 | Interactive intelligent home service robot |
| CN106023238A (en) * | 2016-06-30 | 2016-10-12 | 北京大学 | Color data calibration method for camera module |
| CN106182023A (en) * | 2016-07-11 | 2016-12-07 | 天津艾思科尔科技有限公司 | A kind of household service robot with attack early warning function |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5867268B2 (en) * | 2011-06-21 | 2016-02-24 | ソニー株式会社 | Unevenness inspection apparatus and unevenness inspection method |
-
2017
- 2017-03-24 CN CN201710184827.6A patent/CN106808475B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101947788A (en) * | 2010-06-23 | 2011-01-19 | 焦利民 | Intelligent robot |
| CN104400785A (en) * | 2014-12-02 | 2015-03-11 | 湖南城市学院 | Interactive intelligent home service robot |
| CN106023238A (en) * | 2016-06-30 | 2016-10-12 | 北京大学 | Color data calibration method for camera module |
| CN106182023A (en) * | 2016-07-11 | 2016-12-07 | 天津艾思科尔科技有限公司 | A kind of household service robot with attack early warning function |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106808475A (en) | 2017-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Peng et al. | Rgb-t crowd counting from drone: A benchmark and mmccn network | |
| CN104765768B (en) | The quick and precisely search method of magnanimity face database | |
| CN104599275B (en) | The RGB-D scene understanding methods of imparametrization based on probability graph model | |
| CN111798539B (en) | Adaptive camouflage camouflage online design method and system | |
| CN109816714A (en) | A point cloud object type recognition method based on 3D convolutional neural network | |
| CN114255357B (en) | A method for group-rearing pig identity recognition and health monitoring based on computer vision | |
| CN110956099B (en) | Dynamic gesture instruction identification method | |
| CN108509920A (en) | The face identification method of the multichannel combined feature selecting study of more patch based on CNN | |
| CN109635636A (en) | The pedestrian that blocking characteristic based on attributive character and weighting blends recognition methods again | |
| CN111797840B (en) | Online evaluation method and system for adaptive camouflage camouflage effect | |
| CN105976379A (en) | Fuzzy clustering color image segmentation method based on cuckoo optimization | |
| CN115731579A (en) | Individual recognition method of terrestrial animals based on cross-attention Transformer network | |
| CN106808475B (en) | A kind of intelligent robot that visual performance is excellent | |
| Mao et al. | Semantic segmentation for point cloud scenes via dilated graph feature aggregation and pyramid decoders | |
| Bao et al. | E-unet++: A semantic segmentation method for remote sensing images | |
| CN111488951B (en) | Method for generating countermeasure metric learning model for RGB-D image classification | |
| CN105469392A (en) | High spectral image significance detection method based on regional spectrum gradient characteristic comparison | |
| Song et al. | Non-exemplar domain incremental object detection via learning domain bias | |
| CN111191659B (en) | Multi-shape clothes hanger identification method for clothing production system | |
| CN112836593B (en) | Emotion recognition method and system integrating priori and automatic electroencephalogram features | |
| CN118135266A (en) | A Target Clustering Method for Synthetic Aperture Radar Images | |
| CN111160414A (en) | A high-precision identification method of crop disease and insect pest images | |
| CN110020673A (en) | A kind of method of HSV color space color identification and noise filtering | |
| Qiu et al. | Hydrophobicity Classification of Composite Insulators Based on Light‐Weight Convolutional Neural Networks | |
| CN106815860B (en) | A kind of method for tracking target based on orderly comparison feature |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190425 Address after: 211100 Rheinda Road 699-1, Jiangning High-tech Park, Nanjing City, Jiangsu Province Applicant after: Nanjing Chuangzhi Caicai Technology Co., Ltd. Address before: 518000 Guangdong, Shenzhen, Nanshan District, Guangdong Province, Nanshan District high tech Park South Road, 6 floor, 601 room, building. Applicant before: Shenzhen magic Joint Technology Co. Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |