CN109948577B - Cloth identification method and device and storage medium - Google Patents
Cloth identification method and device and storage medium Download PDFInfo
- Publication number
- CN109948577B CN109948577B CN201910239784.6A CN201910239784A CN109948577B CN 109948577 B CN109948577 B CN 109948577B CN 201910239784 A CN201910239784 A CN 201910239784A CN 109948577 B CN109948577 B CN 109948577B
- Authority
- CN
- China
- Prior art keywords
- cloth
- image
- color
- pattern
- texture
- 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
- 239000004744 fabric Substances 0.000 title claims abstract description 246
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000003860 storage Methods 0.000 title claims abstract description 11
- 238000000605 extraction Methods 0.000 claims abstract description 55
- 239000013598 vector Substances 0.000 claims description 15
- 238000013135 deep learning Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 12
- 238000005520 cutting process Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 3
- 239000003086 colorant Substances 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000005286 illumination Methods 0.000 description 2
- 239000004753 textile Substances 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 210000003464 cuspid Anatomy 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a cloth identification method, a device and a storage medium, wherein the method comprises the following steps: extracting color features of the cloth image through a color feature extraction model; extracting pattern texture characteristics of the cloth image through a pattern texture characteristic extraction model; expressing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics; and inputting an image of the cloth to be identified, and completing identification of the color characteristic and the pattern texture characteristic of the cloth to be identified. According to the method, the pattern, the texture and the color of the cloth are considered simultaneously during identification, the color characteristic and the pattern texture characteristic of the cloth are processed separately, in order to consider the general pattern and the texture detail of the cloth, a characteristic extraction model is divided into two branches, one branch is used for classification of the general pattern, the other branch is used for identification of the finer characteristic of the cloth, the identification speed is high, and the identification effect is good.
Description
Technical Field
The invention relates to the technical field of cloth identification, in particular to a cloth identification method, a cloth identification device and a storage medium.
Background
Cloth is a necessity for human life. The patterns, textures and colors of the cloth are varied and are not easily described in terms of language, which makes it very difficult for manufacturers, distributors and cloth-searching personnel of textile cloth to find a certain cloth. If the cloth is identified to match the cloth in the cloth library in a mode of scanning or photographing the cloth, the time for finding the cloth can be greatly saved, and the cloth with the pattern and the color similar to those of the target cloth can be matched. If the cloth drawing is associated with the cloth feeding production process, all the information of cloth production can be obtained, the cloth production period can be shortened, and the company benefits are improved. Compared with other object identification, the cloth identification has uniqueness, the structural characteristics of the cloth image are not strong, the characteristics are not easy to extract, the cloth has both large patterns and small textures, and the patterns and the textures and the colors need to be identified during identification. The cloth has various patterns, and the texture is ordered texture and disordered texture. The existing technology is not ideal for recognizing pattern texture and color.
Disclosure of Invention
The present invention aims to solve the problems mentioned in the background section above by a method, device and storage medium for fabric identification.
In order to achieve the purpose, the invention adopts the following technical scheme:
a cloth identification method comprises the following steps:
s101, extracting color features of the cloth image through a color feature extraction model; extracting pattern texture characteristics of the cloth image through a pattern texture characteristic extraction model;
s102, representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics;
s103, inputting an image of the cloth to be identified, and completing identification of the color feature and the pattern texture feature of the cloth to be identified.
Specifically, the extracting of the color feature of the cloth image by the color feature extraction model in step S101 includes:
cutting out a size image with a fixed size from the cloth image, counting the color histogram of the size image, representing the color information of the cloth, and extracting the color characteristics of the cloth image.
Specifically, the extracting, in the step S101, the pattern texture features of the fabric image through a pattern texture feature extraction model includes:
and extracting the pattern and texture information of the cloth image through a pattern texture feature extraction model, and extracting the pattern texture feature of the cloth image by adopting a deep learning network.
Particularly, the extracting of the pattern and the texture information of the cloth image by the pattern texture feature extracting model and the extracting of the pattern texture feature by the deep learning network comprise the following steps:
firstly, classifying cloth images: for the cloth labels, the cloth is divided into N categories according to the difference of cloth patterns and textures, wherein each category comprises m groups;
secondly, preprocessing a cloth image: preprocessing each cloth image, and amplifying one cloth image into a plurality of images, wherein the method comprises the following steps: randomly cutting a cloth image, adjusting the cut image to a fixed size, randomly adding illumination and contrast, and converting the RGB image into a gray-scale image to generate a plurality of images.
Particularly, the extracting of the pattern and the texture information of the cloth image by the pattern texture feature extracting model and the extracting of the pattern texture feature by the deep learning network include:
the cloth image is sent into a depth feature extraction network, the depth feature extraction network is subjected to embedding after being regularized by using features L2, the features after the embedding are divided into two branches for processing, one branch is connected with a full connection layer to perform softmax loss for classifying cloth types, and the second branch is subjected to triplet loss (Tripletloss) for distinguishing cloth groups.
Specifically, the step S103 includes: inputting an image of the cloth to be identified, and measuring the similarity of color characteristics by adopting a Pasteur distance; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer, pattern and texture features of the cloth to be recognized are represented, and the distance of the feature vectors is calculated to represent the similarity of the cloth.
Based on the cloth identification method, the invention also discloses a cloth identification device, which comprises the following steps:
the color feature extraction unit is used for extracting the color features of the cloth images through the color feature extraction model;
the pattern texture feature extraction unit is used for extracting the pattern texture features of the cloth images through a pattern texture feature extraction model;
the cloth characteristic unit is used for representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics;
and the cloth identification unit is used for finishing the identification of the color characteristics and the pattern texture characteristics of the cloth to be identified according to the input image of the cloth to be identified.
In particular, the color feature extraction unit is specifically configured to: cutting out size images with fixed sizes from the cloth image, counting a color histogram of the size images, representing color information of the cloth, and extracting color features of the cloth image;
the flower type texture feature extraction unit is specifically configured to: extracting pattern and texture information of the cloth image through a pattern texture feature extraction model, and extracting pattern texture features of the cloth image by adopting a deep learning network;
the method comprises the steps of firstly, classifying cloth images, dividing cloth into N categories according to different cloth patterns and textures, preprocessing the cloth images, amplifying one cloth image into a plurality of images, extracting pattern and texture information of the cloth images through the pattern texture characteristic extraction model, extracting pattern texture characteristics of the cloth images through the deep learning network, wherein the cloth images are sent into the deep characteristic extraction network, the deep characteristic extraction network is regularized by characteristics L2 and then subjected to embedding, the embedded characteristics are divided into two branches, softloss max is performed after one branch is connected with a full connection layer and used for classifying the cloth categories, and triple loss (Triplet loss) is performed on the second branch and used for distinguishing the cloth categories.
In particular, the cloth identification unit is specifically configured to: inputting an image of the cloth to be identified, and measuring the similarity of color characteristics by adopting a Pasteur distance; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer, pattern and texture features of the cloth to be recognized are represented, and the distance of the feature vectors is calculated to represent the similarity of the cloth.
The invention also provides a computer-readable storage medium on which computer program instructions are stored which, when executed by a processor, implement the method of any one of claims 1-6.
The cloth identification method, the cloth identification device and the storage medium provided by the invention take the flower type, the texture and the color of the cloth into consideration during identification, separate processing is carried out on the color characteristic and the flower type texture characteristic of the cloth, in order to take the general flower type and the texture detail of the cloth into consideration, a characteristic extraction model is divided into two branches, one branch is used for classification of the general flower type, and the other branch is used for identification of the finer characteristic of the cloth, so that the identification speed is high, and the identification effect is good.
Drawings
Fig. 1 is a schematic flow chart of a cloth identification method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the classification of fabric types and groups according to an embodiment of the present invention;
fig. 3 is a schematic view of a flow of preprocessing a cloth image according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of an overall architecture of a feature extraction network according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a triple learning process according to an embodiment of the present invention.
Detailed Description
The invention is further illustrated by the following figures and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It is also to be noted that, for the convenience of description, only a part of the contents, not all of the contents, which are related to the present invention, are shown in the drawings, and unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Example one
Referring to fig. 1, fig. 1 is a schematic flow chart of a cloth identification method according to an embodiment of the present invention.
The cloth identification method in the embodiment specifically comprises the following steps:
s101, extracting color features of the cloth image through a color feature extraction model; and extracting the pattern texture characteristics of the cloth image through a pattern texture characteristic extraction model.
And S102, representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics.
S103, inputting an image of the cloth to be identified, and completing identification of the color feature and the pattern texture feature of the cloth to be identified.
Specifically, in this embodiment, the extracting the color feature of the cloth image by the color feature extraction model includes: cutting out a size image with a fixed size from the cloth image, counting the color histogram of the size image, representing the color information of the cloth, and extracting the color characteristics of the cloth image.
In this embodiment, the extracting the pattern texture features of the fabric image through the pattern texture feature extraction model includes: and extracting the pattern and texture information of the cloth image through a pattern texture feature extraction model, and extracting the pattern texture feature of the cloth image by adopting a deep learning network. The method comprises the following steps of extracting pattern and texture information of a cloth image through a pattern texture feature extraction model, extracting pattern texture features of the cloth image through a deep learning network, and comprising the following steps:
firstly, classifying cloth images: for the cloth label, the cloth is divided into N categories such as plain color, lattices, stripes, canine teeth, hidden strips and the like according to the difference of the pattern and the texture of the cloth, wherein each category comprises m groups. One group is one cloth or several extremely similar cloths. The fabric or the fabrics with extremely similar patterns means that the overall patterns and textures of the fabric are similar, for example, the fabric or the fabrics are all plain color fabrics, the textures are twill, the fabrics are stripe fabrics, the width of the stripes is similar, the textile textures are the same, and the colors can be different. The categories and group classifications are shown in FIG. 2 below.
Secondly, preprocessing a cloth image: regarding one cloth or a plurality of extremely similar cloths as a group, the group is used for performing finer identification on the cloths, preprocessing each cloth image, and amplifying one cloth image into a plurality of images, wherein the specific process is as shown in fig. 3: randomly cutting a cloth image, adjusting (resize) the cut image to a fixed size, randomly adding illumination and contrast, and converting an RGB image into a gray-scale image to generate a plurality of images; in this embodiment, one cloth image is expanded to 30 images.
Specifically, in the embodiment, a pattern and texture information of a cloth image is extracted through a pattern texture feature extraction model, and a deep learning network is adopted to extract pattern texture features of the cloth image, as shown in fig. 4, the cloth image is sent into a depth feature extraction network, the depth feature extraction network is regularized by using features L2 and then is subjected to embedding, the features after embedding are divided into two branches, one branch is connected with a full connection layer and then is subjected to Softmax loss for classifying cloth categories, wherein the embedding refers to the conversion of the extracted cloth features into N-dimensional vectors (the size of N can be set), Softmax refers to the probability that a Softmax layer calculation model divides the cloth into each category, the category with the highest probability is selected as an output category, the second branch is subjected to triplex loss (triplex loss) for distinguishing the cloth categories, and the triplex loss is a loss function:
in the formula (I), the compound is shown in the specification,
is used as a target cloth and is provided with a plurality of cloth layers,
is a positive sample cloth, and is characterized in that,
the cloth image triplet selecting method comprises the steps of firstly selecting a cloth image as a target cloth, randomly selecting a pattern from the same group as the target cloth as a positive sample, then randomly selecting a group as a different group from the target cloth, and randomly selecting a pattern from the group as a negative sample.
In this embodiment, the step S103 includes: inputting an image of the cloth to be identified, measuring the similarity of color characteristics by adopting the Pasteur distance, wherein the closer the distance is, the closer the colors of the cloth are; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer to represent the pattern and texture features of the cloth to be recognized, the similarity of the cloth is represented by calculating the distance of the feature vectors, and the closer the distance between the cloth feature vectors is, the more similar the two cloths are.
Example two
Based on the above cloth identification method, this embodiment provides a cloth identification apparatus, which includes:
and the color feature extraction unit is used for extracting the color features of the cloth images through the color feature extraction model.
And the pattern texture feature extraction unit is used for extracting the pattern texture features of the cloth images through a pattern texture feature extraction model.
And the cloth characteristic unit is used for representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics.
And the cloth identification unit is used for finishing the identification of the color characteristics and the pattern texture characteristics of the cloth to be identified according to the input image of the cloth to be identified.
The method comprises the steps of cutting out size images with fixed sizes from a cloth image, counting a color histogram of the size images, representing color information of cloth, extracting color features of the cloth image, classifying the cloth images, dividing cloth into N categories according to different cloth patterns and textures, preprocessing the cloth images, expanding the cloth images into a plurality of images, extracting pattern and texture information of the cloth images through a pattern texture feature extraction model, extracting positive and negative texture samples of the cloth images through a pattern texture feature extraction model, selecting a negative branch of a triple copy image as a cloth sample, selecting a positive branch of the triple copy image as a cloth sample, selecting a negative branch of the triple copy image as a cloth sample, and selecting a negative branch of the triple copy image as a cloth sample.
In this embodiment, the cloth identification unit is specifically configured to: inputting an image of the cloth to be identified, measuring the similarity of color characteristics by adopting the Pasteur distance, wherein the closer the distance is, the closer the colors of the cloth are; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer to represent the pattern and texture features of the cloth to be recognized, the similarity of the cloth is represented by calculating the distance of the feature vectors, and the closer the distance between the cloth feature vectors is, the more similar the two cloths are.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, which stores computer program instructions, and when the computer program instructions are executed by a processor, the cloth identification method according to the first embodiment is implemented.
According to the technical scheme, the pattern, the texture and the color of the cloth are considered simultaneously during identification, the color characteristic and the pattern texture characteristic of the cloth are processed separately, in order to consider the general pattern and the texture details of the cloth, a characteristic extraction model is divided into two branches, one branch is used for classification of the general pattern, the other branch is used for identification of the finer characteristics of the cloth, the identification speed is high, and the identification effect is good.
It will be understood by those skilled in the art that all or part of the above embodiments may be implemented by the computer program to instruct the relevant hardware, and the program may be stored in a computer readable storage medium, and when executed, may include the procedures of the embodiments of the methods as described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory or a random access memory.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (3)
1. A cloth identification method is characterized by comprising the following steps:
the method comprises the steps of S101, extracting color features of a cloth image through a color feature extraction model, extracting pattern texture features of the cloth image through a pattern texture feature extraction model, wherein the extracting of the color features of the cloth image through the color feature extraction model comprises the steps of cutting out a size image with a fixed size from the cloth image, counting a color histogram of the size image, representing color information of cloth, and extracting the color features of the cloth image, the extracting of the pattern texture features of the cloth image through the pattern texture feature extraction model comprises the steps of extracting pattern and texture information of the cloth image through the pattern texture feature extraction model, extracting pattern texture features of the cloth image through a deep learning network, classifying the cloth image, classifying the cloth into N classes according to the difference of cloth patterns and textures, wherein each class comprises m classes, pre-processing the cloth image, pre-processing each cloth image, expanding the cloth image into a plurality of cloth images, adjusting the size of the cloth image into a plurality of classes according to the cloth patterns and textures, extracting the pattern texture features of the cloth image, and extracting the pattern texture features of the cloth image through a deep learning network, and extracting the texture features of a plurality of the cloth image after the three sub-step of the cloth image, wherein the cloth image is divided into a three sub-step of sub-image extraction, the cloth image extraction step of the three sub-step of the cloth image extraction, the three sub-step of the cloth image extraction, the three sub-step of the cloth image extraction, the sub-image extraction of the sub;
s102, representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics;
s103, inputting an image of the cloth to be identified, and completing identification of color features and pattern texture features of the cloth to be identified; wherein, the step S103 specifically includes: inputting an image of the cloth to be identified, and measuring the similarity of color characteristics by adopting a Pasteur distance; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer, pattern and texture features of the cloth to be recognized are represented, and the distance of the feature vectors is calculated to represent the similarity of the cloth.
2. A cloth recognition device, characterized in that, the device includes:
the color feature extraction unit is used for extracting the color features of the cloth images through the color feature extraction model; wherein the color feature extraction unit is specifically configured to: cutting out size images with fixed sizes from the cloth image, counting a color histogram of the size images, representing color information of the cloth, and extracting color features of the cloth image;
the cloth image preprocessing comprises the steps of preprocessing each cloth image, expanding one cloth image into a plurality of images, extracting pattern and texture information of the cloth image through the pattern texture characteristic extraction model, extracting the pattern and texture characteristics of the cloth image through the deep learning network, and performing preprocessing of the cloth image, wherein the pattern and texture characteristics of the cloth image are extracted through the pattern texture characteristic extraction model, and the pattern texture characteristic extraction unit is used for extracting the pattern texture characteristics of the cloth image through the deep learning network;
the cloth characteristic unit is used for representing the characteristics of the cloth by adopting the color characteristics and the pattern texture characteristics;
the cloth identification unit is used for finishing the identification of the color characteristics and the pattern texture characteristics of the cloth to be identified according to the input image of the cloth to be identified; wherein, the cloth recognition unit is specifically configured to: inputting an image of the cloth to be identified, and measuring the similarity of color characteristics by adopting a Pasteur distance; through a trained pattern texture recognition model, feature vectors are extracted from the embedding layer, pattern and texture features of the cloth to be recognized are represented, and the distance of the feature vectors is calculated to represent the similarity of the cloth.
3. A computer-readable storage medium on which computer program instructions are stored, which when executed by a processor implement the method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910239784.6A CN109948577B (en) | 2019-03-27 | 2019-03-27 | Cloth identification method and device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910239784.6A CN109948577B (en) | 2019-03-27 | 2019-03-27 | Cloth identification method and device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109948577A CN109948577A (en) | 2019-06-28 |
| CN109948577B true CN109948577B (en) | 2020-08-04 |
Family
ID=67012019
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910239784.6A Active CN109948577B (en) | 2019-03-27 | 2019-03-27 | Cloth identification method and device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109948577B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110321850A (en) * | 2019-07-05 | 2019-10-11 | 杭州时趣信息技术有限公司 | Garment material automatic identifying method, device, system, equipment and storage medium |
| CN114253199B (en) * | 2022-03-01 | 2022-05-06 | 江苏苏美达纺织有限公司 | A knitting control method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104331447A (en) * | 2014-10-29 | 2015-02-04 | 邱桃荣 | Cloth color card image retrieval method |
| CN107256246A (en) * | 2017-06-06 | 2017-10-17 | 西安工程大学 | PRINTED FABRIC image search method based on convolutional neural networks |
| WO2018213108A1 (en) * | 2017-05-15 | 2018-11-22 | Siemens Aktiengesellschaft | Domain adaptation and fusion using weakly supervised target irrelevant data |
| CN108960265A (en) * | 2017-05-22 | 2018-12-07 | 阿里巴巴集团控股有限公司 | Optimization method, image classification method, the apparatus and system of image classification process |
| CN108960080A (en) * | 2018-06-14 | 2018-12-07 | 浙江工业大学 | Based on Initiative Defense image to the face identification method of attack resistance |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180330194A1 (en) * | 2017-05-15 | 2018-11-15 | Siemens Aktiengesellschaft | Training an rgb-d classifier with only depth data and privileged information |
| CN108875827B (en) * | 2018-06-15 | 2022-04-12 | 拓元(广州)智慧科技有限公司 | Method and system for classifying fine-grained images |
| CN109117862B (en) * | 2018-06-29 | 2019-06-21 | 北京达佳互联信息技术有限公司 | Image tag recognition methods, device and server |
| CN109344845B (en) * | 2018-09-21 | 2020-06-09 | 哈尔滨工业大学 | Feature matching method based on triple deep neural network structure |
-
2019
- 2019-03-27 CN CN201910239784.6A patent/CN109948577B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104331447A (en) * | 2014-10-29 | 2015-02-04 | 邱桃荣 | Cloth color card image retrieval method |
| WO2018213108A1 (en) * | 2017-05-15 | 2018-11-22 | Siemens Aktiengesellschaft | Domain adaptation and fusion using weakly supervised target irrelevant data |
| CN108960265A (en) * | 2017-05-22 | 2018-12-07 | 阿里巴巴集团控股有限公司 | Optimization method, image classification method, the apparatus and system of image classification process |
| CN107256246A (en) * | 2017-06-06 | 2017-10-17 | 西安工程大学 | PRINTED FABRIC image search method based on convolutional neural networks |
| CN108960080A (en) * | 2018-06-14 | 2018-12-07 | 浙江工业大学 | Based on Initiative Defense image to the face identification method of attack resistance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109948577A (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11681418B2 (en) | Multi-sample whole slide image processing in digital pathology via multi-resolution registration and machine learning | |
| Khairnar et al. | Disease detection and diagnosis on plant using image processing—a review | |
| Su et al. | Region segmentation in histopathological breast cancer images using deep convolutional neural network | |
| CN109657595B (en) | Key feature region matching face recognition method based on stacked hourglass network | |
| CN110674874B (en) | Fine-grained image identification method based on target fine component detection | |
| US8503792B2 (en) | Patch description and modeling for image subscene recognition | |
| CN111310662B (en) | A method and system for flame detection and recognition based on integrated deep network | |
| CN108090406B (en) | Face recognition method and system | |
| CN114140665A (en) | A Dense Small Object Detection Method Based on Improved YOLOv5 | |
| CN108520226A (en) | A kind of pedestrian's recognition methods again decomposed based on body and conspicuousness detects | |
| CN110751089A (en) | A flame target detection method based on digital images and convolutional features | |
| CN110728302A (en) | Method for identifying color textile fabric tissue based on HSV (hue, saturation, value) and Lab (Lab) color spaces | |
| CN109740572A (en) | A kind of human face in-vivo detection method based on partial color textural characteristics | |
| Priyankara et al. | Computer assisted plant identification system for Android | |
| CN111191584A (en) | Face recognition method and device | |
| Lodh et al. | Flower recognition system based on color and GIST features | |
| CN111126401A (en) | License plate character recognition method based on context information | |
| CN104573701B (en) | A kind of automatic testing method of Tassel of Corn | |
| CN109948577B (en) | Cloth identification method and device and storage medium | |
| CN103778428B (en) | Disease area-of-interest exacting method and system based on block labelling | |
| Kumar et al. | Syn2real: Forgery classification via unsupervised domain adaptation | |
| Puttemans et al. | Visual detection and species classification of orchid flowers | |
| CN115170839B (en) | Image data enhancement method and system | |
| Hu et al. | Fast face detection based on skin color segmentation using single chrominance Cr | |
| Ticay-Rivas et al. | Pollen classification based on geometrical, descriptors and colour features using decorrelation stretching method |
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 |