CN108789428A - A kind of smart home robot - Google Patents
A kind of smart home robot Download PDFInfo
- Publication number
- CN108789428A CN108789428A CN201810542349.6A CN201810542349A CN108789428A CN 108789428 A CN108789428 A CN 108789428A CN 201810542349 A CN201810542349 A CN 201810542349A CN 108789428 A CN108789428 A CN 108789428A
- Authority
- CN
- China
- Prior art keywords
- clothes
- clothing
- smooth
- image
- 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.)
- Withdrawn
Links
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of smart home robots, including clothes identification device, clothes sorting equipment and clothes cleaning device, the clothes identification device is for being identified the clothes in room, for being classified to clothes according to the clothes recognition result, the clothes cleaning device is used to carry out classification cleaning to clothes according to clothes classification results the clothes sorting equipment.Beneficial effects of the present invention are:A kind of smart home robot is provided, sorting and the classification cleaning of clothes are realized.
Description
Technical field
The present invention relates to robotic technology fields, and in particular to a kind of smart home robot.
Background technology
As the improvement of people's living standards, robot enters into people's lives, become a part for smart home.
Clothes are the necessitys in people's life daily, as living standard is continuously improved, need of the people to clothes
It asks and is no longer satisfied with simple warming, also unprecedented to emphasize its aesthetic function, then designer and the producer carry out one after another
Innovation so that material, pattern, the style of clothes are unprecedented abundant.Clothes type is various no doubt to be increased for people's lives
Color, but also bring new problem simultaneously.Material, style and pattern it is changeable so that the information of clothes becomes more multiple
Miscellaneous difficulty distinguishes, while the housework activity such as select, clean of clothes also being made than more to spend human and material resources in the past.
Invention content
In view of the above-mentioned problems, the present invention is intended to provide a kind of smart home robot.
The purpose of the present invention is realized using following technical scheme:
A kind of smart home robot, including clothes identification device, clothes sorting equipment and clothes cleaning device are provided,
For the clothes in room to be identified, the clothes sorting equipment is used to be known according to the clothes clothes identification device
Other result classifies to clothes, and the clothes cleaning device is used to carry out classification cleaning to clothes according to clothes classification results.
Beneficial effects of the present invention are:A kind of smart home robot is provided, sorting and the classification for realizing clothes are clear
It washes.
Description of the drawings
Using attached drawing, the invention will be further described, but the embodiment in attached drawing does not constitute any limit to the present 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 the structural schematic diagram of the present invention;
Reference numeral:
Clothes identification device 1, clothes sorting equipment 2, clothes cleaning device 3.
Specific implementation mode
The invention will be further described with the following Examples.
Referring to Fig. 1, a kind of smart home robot of the present embodiment, including clothes identification device 1, clothes sorting equipment 2
With clothes cleaning device 3, the clothes identification device 1 is for being identified the clothes in room, the clothes sorting equipment 2
For being classified to clothes according to the clothes recognition result, the clothes cleaning device 3 is used for according to clothes classification results
Classification cleaning is carried out to clothes.
A kind of smart home robot is present embodiments provided, sorting and the classification cleaning of clothes are realized.
Preferably, the clothes identification device 1 includes first processing units, second processing unit and third processing unit,
The first processing units are to adopt in advance for acquiring non-smooth image of clothing and smooth image of clothing, the smooth image of clothing
Collection, the second processing unit is for extracting the feature of non-smooth image of clothing and smooth image of clothing, the third
Processing unit is matched for non-smooth image of clothing with smooth image of clothing feature, and non-smooth clothes identification is completed.
The present embodiment introduces the smooth image with part clothes, with the image of clothing progress for being in out-of-flatness placement status
Match, to convert the identification of non-smooth image of clothing to image matching problems.
Preferably, the second processing unit includes a feature extraction unit, Further Feature Extraction unit, described primary
Feature extraction unit is used for the extraction of non-smooth image of clothing and smooth image of clothing color characteristic, the Further Feature Extraction list
Textural characteristics of the member for non-smooth image of clothing and smooth image of clothing extract;
This preferred embodiment realizes non-smooth image of clothing and smooth image of clothing feature extraction, is matched for subsequent characteristics
It lays a good foundation with non-smooth clothes identification.
Preferably, a feature extraction unit is used for the extraction of color of image feature:Image is converted from rgb space
To the spaces HSI, wherein H indicates that chrominance component, S indicate saturation degree component, and I indicates luminance component, for arbitrary pixel, as I <
0.2 and S < 0.2+4 (0.2-I), then pixel is black, and as I > 0.8 and S < 0.2+4 [0.2- (1-I)], then the pixel is white
Color, as 0.2≤I≤0.8 and S < 0.2, then pixel is that chroma circle is equally divided into 10 parts, pixel by grey for remaining color
Share 13 kinds of colors;By 13 kinds of colors according to being centainly ranked sequentially, the color characteristic RL of image of clothing is determined using following formula1:RL1
=[p1,p2,…,p13], wherein piIndicate the characterization factor of i-th kind of color,Wherein, niIndicate i-th
The probability that kind of colored pixels occur in the picture, N indicate pixel total amount in image, i=1,2 ..., 13;
This preferred embodiment color characteristic generates clothing surface because of ambient light in 13 kinds of colors of HSI definition spaces
Deviation has preferable robustness, reduces illumination and shade and identifies the influence brought to clothes.
Preferably, the Further Feature Extraction unit is used for the textural characteristics of non-smooth image of clothing and smooth image of clothing
It extracts:Image of clothing is divided into the region with different texture attribute, elliptical grain is extracted from each region and is adopted
Sample block, then extraction elliptical grain uses the textural characteristics of block respectively;The elliptical grain sampling block carries in the following ways
It takes:Using the ellipse that semi-major axis is 1/5th evolution of region area, eccentricity is 0.2 as sliding window, if the sliding window
Mouthful can be completely disposed in region, then keep the constant increase semi-major axis of eccentricity, until with texture region inscribe, by sliding at this time
Window is as elliptical grain sampling block;If the sliding window cannot be completely in region, by sliding window and the region
Position when overlapping area maximum is as elliptical grain sampling block;The elliptical grain use the textural characteristics of block to use with
Under type is extracted:Using the principal direction of Radon transformation calculations elliptical grain sampling blocks, as the texture master of corresponding region
Direction determines the textural characteristics RL of image of clothing using following formula2:RL2=[q1,q2,…,qm], wherein qjIndicate j-th of region
Texture principal direction, m indicate with different texture attribute region number, j=1,2 ..., m;
This preferred embodiment divides the image into the region of different texture attribute, first, since clothes produce a large amount of foldings
And distortion, even same texture may also show a variety of different states in image;Second is that non-smooth image of clothing and
Smooth image of clothing shooting distance and visual angle are inconsistent, and the scale and direction that texture may be caused to present have differences.
Preferably, the third processing unit includes a matching unit, Secondary Match unit and matching unit three times, institute
A matching that matching unit is used for non-smooth image of clothing and smooth image of clothing color characteristic is stated, the first matching knot is obtained
Fruit, textural characteristics of the Secondary Match unit for non-smooth image of clothing and smooth image of clothing are matched, and obtain the
Two matching results, the matching unit three times are used to carry out non-smooth clothes according to the first matching result and the second matching result
Identification;
Matching unit is used for the matching of non-smooth image of clothing and smooth image of clothing color characteristic:Under
Formula calculates the similarity factor of non-smooth image of clothing and smooth image of clothing color characteristic:Formula
In, BZkIndicate the similarity factor of the color characteristic of non-smooth clothes and the smooth clothes of kth part, piIt indicates in non-smooth clothes
The characterization factor of i kind colors, pi,kIndicate that the characterization factor of the smooth i-th kind of color of clothes of kth part, k ∈ [1, M], M indicate clothes
Total quantity;The similarity factor of color characteristic is bigger, indicates that matching degree is higher;The similarity factor for the feature that gets colors is most
First matching result of the big smooth image of clothing as non-smooth clothes;
This preferred embodiment measures the similar of two image of clothing color characteristics by defining the similarity factor of color characteristic
Degree.Realize the accurate acquisition of the first matching result:
Preferably, Secondary Match unit is for the progress of the textural characteristics of non-smooth image of clothing and smooth image of clothing
Match:The similarity factor of non-smooth image of clothing and smooth image of clothing textural characteristics is calculated using following formula:In formula, FVkIndicate the similitudes of the textural characteristics of non-smooth clothes and the smooth clothes of kth part because
Son, piIndicate the texture principal direction of j-th of texture region in non-smooth clothes, qj,kIndicate smooth j-th of the texture of clothes of kth part
The texture principal direction in region, k ∈ [1, M], M indicate the total quantity of clothes;The similarity factor of textural characteristics is bigger, indicates matching
Degree is higher;Choose second matching knot of the maximum smooth image of clothing of similarity factor as non-smooth clothes of textural characteristics
Fruit;
The matching unit three times is for knowing non-smooth clothes according to the first matching result and the second matching result
Not:If the first matching result and the second matching result are identical, using any matching result as recognition result;If the first matching knot
Fruit and the second matching result differ, then larger in the similarity factor for the feature that gets colors and the similarity factor of textural characteristics
Be used as recognition result.
This preferred embodiment realizes clothing texture feature by the similarity factor of definition textural characteristics and textural characteristics
Accurate extraction realize accurate match and the knowledge of clothes by being merged to the first matching result and the second matching result
Not.
5 families are chosen to clean clothes using smart home robots of the present invention, respectively family 1, family 2,
Family 3, family 4, family 5, count cleaning efficiency and user satisfaction, compared with washing machine, the advantageous effect of generation
As shown in the table:
Cleaning efficiency improves | User satisfaction improves | |
Family 1 | 29% | 27% |
Family 2 | 27% | 26% |
Family 3 | 26% | 26% |
Family 4 | 25% | 24% |
Family 5 | 24% | 22% |
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 being explained in detail to the present invention with reference to preferred embodiment, the ordinary skill family of this field answers
Work as understanding, technical scheme of the present invention can be modified or replaced equivalently, without departing from the reality of technical solution of the present invention
Matter and range.
Claims (7)
1. a kind of smart home robot, which is characterized in that clean and fill including clothes identification device, clothes sorting equipment and clothes
It sets, for the clothes in room to be identified, the clothes sorting equipment is used for according to the clothes clothes identification device
Dress recognition result classifies to clothes, and the clothes cleaning device is used to clothes classify according to clothes classification results clear
It washes.
2. smart home robot according to claim 1, which is characterized in that the clothes identification device includes at first
Unit, second processing unit and third processing unit are managed, the first processing units are for acquiring non-smooth image of clothing peace
Whole image of clothing, the smooth image of clothing are acquisition in advance, the second processing unit be used for non-smooth image of clothing and
The feature of smooth image of clothing extracts, and the third processing unit is for non-smooth image of clothing with smooth image of clothing spy
Sign is matched, and non-smooth clothes identification is completed.
3. smart home robot according to claim 2, which is characterized in that the second processing unit includes primary special
Extraction unit, Further Feature Extraction unit are levied, a feature extraction unit is used for non-smooth image of clothing and smooth clothes
The extraction of color of image feature, the Further Feature Extraction unit are used for the texture of non-smooth image of clothing and smooth image of clothing
Feature extracts;
Feature extraction unit is used for the extraction of color of image feature:Image is transformed into the spaces HSI from rgb space,
Wherein, H indicates that chrominance component, S indicate saturation degree component, and I indicates luminance component, for arbitrary pixel, as I < 0.2 and S <
0.2+4 (0.2-I), then pixel is black, and as I > 0.8 and S < 0.2+4 [0.2- (1-I)], then the pixel is white, when 0.2
≤ I≤0.8 and S < 0.2, then pixel is that chroma circle is equally divided into 10 parts, pixel shares 13 kinds by grey for remaining color
Color;By 13 kinds of colors according to being centainly ranked sequentially, the color characteristic RL of image of clothing is determined using following formula1:RL1=[p1,
p2,…,p13], wherein piIndicate the characterization factor of i-th kind of color,Wherein, niIndicate i-th kind of color
The probability that pixel occurs in the picture, N indicate image in pixel total amount, i=1,2 ..., 13.
4. smart home robot according to claim 3, which is characterized in that the Further Feature Extraction unit is for non-
The textural characteristics of smooth image of clothing and smooth image of clothing extract:Image of clothing is divided into different texture attribute
Region, from each region extract elliptical grain sampling block, then respectively extraction elliptical grain use block textural characteristics;
The elliptical grain sampling block extracts in the following ways:By semi-major axis be 1/5th evolution of region area, eccentricity is
0.2 ellipse is as sliding window, if the sliding window can be completely disposed in region, keeps the constant increase half of eccentricity long
Axis, until with texture region inscribe, using sliding window at this time as elliptical grain sampling block;If the sliding window cannot be complete
Entirely in the region, then using sliding window and the area coincidence area maximum when position as elliptical grain sampling block;Institute
Elliptical grain is stated to extract in the following ways using the textural characteristics of block:It is sampled using Radon transformation calculations elliptical grains
The principal direction of block determines the textural characteristics RL of image of clothing using following formula as the texture principal direction of corresponding region2:RL2
=[q1,q2,…,qm], wherein qjIndicate that the texture principal direction in j-th of region, m indicate the region with different texture attribute
Number, j=1,2 ..., m.
5. smart home robot according to claim 4, which is characterized in that the third processing unit includes primary
With unit, Secondary Match unit and matching unit, a matching unit are used for non-smooth image of clothing and smooth clothes three times
The matching of color of image feature is filled, obtains the first matching result, the Secondary Match unit is for non-smooth image of clothing peace
The textural characteristics of whole image of clothing are matched, and the second matching result is obtained, and the matching unit three times is used for according to first
Non- smooth clothes are identified with result and the second matching result;
Matching unit is used for the matching of non-smooth image of clothing and smooth image of clothing color characteristic:Using following formula meter
Calculate the similarity factor of non-smooth image of clothing and smooth image of clothing color characteristic:In formula, BZk
Indicate the similarity factor of the color characteristic of non-smooth clothes and the smooth clothes of kth part, piIndicate i-th kind of face in non-smooth clothes
The characterization factor of color, pi,kIndicate that the characterization factor of the smooth i-th kind of color of clothes of kth part, k ∈ [1, M], M indicate the sum of clothes
Amount;The similarity factor of color characteristic is bigger, indicates that matching degree is higher;The similarity factor for the feature that gets colors is maximum flat
First matching result of the whole image of clothing as non-smooth clothes.
6. smart home robot according to claim 5, which is characterized in that Secondary Match unit is used for non-smooth clothes
The textural characteristics of image and smooth image of clothing are matched:Non- smooth image of clothing and smooth image of clothing are calculated using following formula
The similarity factor of textural characteristics:In formula, FVkIndicate non-smooth clothes and the smooth clothes of kth part
Textural characteristics similarity factor, piIndicate the texture principal direction of j-th of texture region in non-smooth clothes, qj,kIndicate kth
The texture principal direction of part j-th of texture region of smooth clothes, k ∈ [1, M], M indicate the total quantity of clothes;Textural characteristics it is similar
Sex factor is bigger, indicates that matching degree is higher;The maximum smooth image of clothing of similarity factor of textural characteristics is chosen as non-
Second matching result of smooth clothes.
7. smart home robot according to claim 6, which is characterized in that the matching unit three times is used for according to the
Non- smooth clothes are identified in one matching result and the second matching result:If the first matching result and the second matching result phase
Together, then using any matching result as recognition result;If the first matching result and the second matching result differ, get colors
Larger conduct recognition result in the similarity factor of feature and the similarity factor of textural characteristics.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810542349.6A CN108789428A (en) | 2018-05-30 | 2018-05-30 | A kind of smart home robot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810542349.6A CN108789428A (en) | 2018-05-30 | 2018-05-30 | A kind of smart home robot |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108789428A true CN108789428A (en) | 2018-11-13 |
Family
ID=64089501
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810542349.6A Withdrawn CN108789428A (en) | 2018-05-30 | 2018-05-30 | A kind of smart home robot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108789428A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111452035A (en) * | 2019-01-22 | 2020-07-28 | ***通信集团终端有限公司 | Robot control method and robot |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1365067A (en) * | 2001-01-09 | 2002-08-21 | 三星电子株式会社 | Image retrieval method based on color and image characteristic combination |
US20090116698A1 (en) * | 2007-11-07 | 2009-05-07 | Palo Alto Research Center Incorporated | Intelligent fashion exploration based on clothes recognition |
CN104730930A (en) * | 2015-01-16 | 2015-06-24 | 小米科技有限责任公司 | Clothes sorting method and device and clothes washing method and device |
CN108055317A (en) * | 2017-12-11 | 2018-05-18 | 西安科技大学 | Clothing manages the integrated wardrobe of coded system and processing method and application this method |
-
2018
- 2018-05-30 CN CN201810542349.6A patent/CN108789428A/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1365067A (en) * | 2001-01-09 | 2002-08-21 | 三星电子株式会社 | Image retrieval method based on color and image characteristic combination |
US20090116698A1 (en) * | 2007-11-07 | 2009-05-07 | Palo Alto Research Center Incorporated | Intelligent fashion exploration based on clothes recognition |
CN104730930A (en) * | 2015-01-16 | 2015-06-24 | 小米科技有限责任公司 | Clothes sorting method and device and clothes washing method and device |
CN108055317A (en) * | 2017-12-11 | 2018-05-18 | 西安科技大学 | Clothing manages the integrated wardrobe of coded system and processing method and application this method |
Non-Patent Citations (1)
Title |
---|
桂琳: "智能家居中基于视觉的服装分析和推荐研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111452035A (en) * | 2019-01-22 | 2020-07-28 | ***通信集团终端有限公司 | Robot control method and robot |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103824059B (en) | Facial expression recognition method based on video image sequence | |
CN108549891B (en) | Multi-scale diffusion well-marked target detection method based on background Yu target priori | |
US9396412B2 (en) | Machine-learnt person re-identification | |
CN106874884B (en) | Human body recognition methods again based on position segmentation | |
Zhang et al. | A novel graph-based optimization framework for salient object detection | |
CN106960182B (en) | A kind of pedestrian's recognition methods again integrated based on multiple features | |
CN106023151B (en) | Tongue object detection method under a kind of open environment | |
CN107330892A (en) | A kind of sunflower disease recognition method based on random forest method | |
CN105957124B (en) | With the natural image color edit methods and device for repeating situation elements | |
CN107481183B (en) | Low-light-level image colorization method based on multi-dimensional data association rule | |
CN103366178A (en) | Method and device for carrying out color classification on target image | |
CN107194348A (en) | The domain color recognition methods of target object in a kind of image | |
CN109271932A (en) | Pedestrian based on color-match recognition methods again | |
Mythili et al. | Color image segmentation using ERKFCM | |
CN110032932A (en) | A kind of human posture recognition method based on video processing and decision tree given threshold | |
CN108961250A (en) | A kind of object statistical method, device, terminal and storage medium | |
CN110298893A (en) | A kind of pedestrian wears the generation method and device of color identification model clothes | |
CN109800659A (en) | A kind of action identification method and device | |
CN109213886A (en) | Image search method and system based on image segmentation and Fuzzy Pattern Recognition | |
CN108789428A (en) | A kind of smart home robot | |
Dwina et al. | Skin segmentation based on improved thresholding method | |
CN106960188B (en) | Weather image classification method and device | |
CN107358635A (en) | A kind of Color-scale Morphology image processing method based on fuzzy comparability | |
Liu et al. | Fuzzy color recognition and segmentation of robot vision scene | |
CN104484324B (en) | A kind of pedestrian retrieval method of multi-model and fuzzy color |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181113 |