CN109918783B - Intelligent clothing design system - Google Patents
Intelligent clothing design system Download PDFInfo
- Publication number
- CN109918783B CN109918783B CN201910169133.4A CN201910169133A CN109918783B CN 109918783 B CN109918783 B CN 109918783B CN 201910169133 A CN201910169133 A CN 201910169133A CN 109918783 B CN109918783 B CN 109918783B
- Authority
- CN
- China
- Prior art keywords
- module
- component
- clothing
- diagram
- vectorization
- 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
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Processing Or Creating Images (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of clothing design and discloses an intelligent clothing design system which comprises a processor, wherein the processor is connected with an importing module, an identifying module, a modifying module and a database, the database is arranged in a way that all component diagrams in clothing style diagrams of a plurality of types are in one-to-one correspondence with structure diagrams thereof, the importing module is used for acquiring the clothing style diagrams or the component diagrams to be identified, the identifying module is used for identifying the clothing style diagrams or the component diagrams to be identified, the corresponding structure diagrams are acquired by combining with the database, the modifying module is used for virtually fitting all the structure diagrams with a human body model, and modifying the structure diagrams according to fitting results. The system provides an intelligent design platform for a clothing designer, realizes the conversion from a style chart to a structural chart, saves the cost of clothing design and plate making process, and can also improve the reaction speed of the whole manufacturing process.
Description
Technical Field
The invention relates to the technical field of clothing design, in particular to an intelligent clothing design system.
Background
In recent years, the influence of international economic forms and the continuous rise of domestic raw materials and labor costs make the market competition of the domestic clothing textile industry increasingly stronger. In the traditional garment manufacturing process, a templet is generated by a manual edition or garment CAD platemaking system according to the demands of clients or designers, so that a garment is manufactured, the clients or designers can check the effect, the whole process takes a long time, the cost is high, and the experience dependence on the templet is strong. The clothing CAD is used for plate making, so that the plate making speed is accelerated to a certain extent, but the clothing CAD system only replaces the traditional manual plate making program by means of a computer. Aiming at the PDS technology of the existing clothing template making system, the automation degree is low, the operation is complex, the efficiency is low, the inspiration of a designer, the technology and the like cannot be completely replaced by the PDS technology, the dependence on the experience of a paper sample designer is not separated, the complicated printing process is not basically changed, the design process is not intelligent, the rapid improvement of the artificial intelligence, big data and computer application technology and the continuous improvement and improvement of the planar clothing aided design system technology are not realized, and the intelligent clothing CAD has become the research direction of a plurality of students. The clothing style pattern recognition and the application thereof in the structure diagram search are all intelligent design services for the clothing industry.
Many universities have been studied in recent years to address this area of technology. Beijing clothing college Liu developed an intelligent system for clothing deformation with the construction of virtual models of 3D clothing, and the development of this system brought great convenience to designers in deforming some styles, achieving the effect of intelligent design, but it only parameterized the deformation of the specified clothing. The method is characterized in that the automatic identification of the clothing style images is researched by the university of east China Dong Chenxue, the characteristics of the clothing style images are extracted by adopting an image processing technology, MATLAB programming software is used as a technical platform, and image processing functions are called to extract and identify the style image information, so that the automatic identification and measurement of key template data such as the clothing length, sleeve length, chest circumference, waistline, hip circumference and the like in the style images are finally realized. The Shanghai engineering university Liu Weimin and the like construct BP models between the human waistline and hip size variation and the corresponding transformation rules by adopting a reverse neural network algorithm, and automatically generate the fit template by calling the corresponding transformation rules of the target size.
In summary, there is an increasing demand for an intelligent garment design system, from design to construction, to virtual display of ready-made garments, and to modify the construction.
Disclosure of Invention
The invention provides an intelligent clothing design system, which solves the problems of low automation degree, complex operation, low efficiency and the like of the existing clothing design system.
The invention can be realized by the following technical scheme:
the intelligent clothing design system comprises a processor, wherein the processor is connected with an importing module, an identifying module, a modifying module and a database, the database is arranged in a one-to-one correspondence between each part drawing in clothing style drawings of a plurality of types and the structure drawing thereof, the importing module is used for acquiring the clothing style drawings or the part drawings to be identified, the identifying module is used for identifying the clothing style drawings or the part drawings to be identified, the corresponding structure drawing is acquired by combining the database, and the modifying module is used for virtually fitting each structure drawing with a human body model and modifying the structure drawing according to the fitting result.
Further, the identification module comprises a segmentation module, a vectorization module, a feature extraction module and a classification module,
the segmentation module is used for segmenting the clothing style graph to be identified to obtain corresponding part graphs;
the vectorization module is used for carrying out edge identification on each segmented part drawing or the clothing part drawing to be identified, analyzing the identified edges to obtain a point set sequence, and carrying out vectorization on the point set sequence to obtain a vector part drawing to be detected;
the feature extraction module is used for carrying out chord feature extraction and normalization processing on the vector component diagram to be detected;
the classifying module is used for calling the vectorizing module and the feature extracting module to process the component images in the database, classifying the processed component images by adopting a support vector machine SVM-based classifying method, finding out the category of the normalized component images to be tested, accurately identifying the component images to be tested in the category by adopting a nearest neighbor method 1NN, finding out the component image closest to the component images to be tested, and obtaining the structure diagram corresponding to the component images to be tested.
Further, the chord characteristics are described by an average projection length matrix PM, an outer chord length matrix OM and an absolute value matrix IODM of the difference between the inner chord length and the outer chord length, the vectorization comprises straight line segment vectorization and curve segment vectorization, and the curve segment vectorization is carried out by adopting a piecewise cubic Bezier curve.
Further, the recognition module further comprises a preprocessing module, the preprocessing module is connected with the segmentation module and the vectorization module and is used for preprocessing the clothing style graph or the component graph to be recognized, the preprocessing module comprises a sharpening module and a binarization module, the sharpening module adopts a Laplace operator to sharpen the clothing style graph or the component graph to be recognized, and the binarization module is used for carrying out binarization processing on the sharpened clothing style graph or the component graph to be recognized.
Further, the modification module adopts CLO3D software to carry out virtual try-on and structural diagram modification.
Further, the importing module adopts a universal tool of CAD software to draw or adopts a scanning instrument to scan, so as to obtain a garment pattern diagram or a part diagram to be identified.
The beneficial technical effects of the invention are as follows:
the method has the advantages that the database is built, the pattern diagram or the part diagram of the garment to be identified can be drawn by oneself or directly obtained by virtue of the importing module, the pattern diagram or the part diagram of the garment to be identified is identified by virtue of the identifying module, the corresponding structure diagram is obtained by combining the database, the modification module is used for virtually fitting each structure diagram with the human body model, and the structure diagram is modified according to the fitting result, so that an intelligent design platform is provided for a garment designer, the conversion from the pattern diagram to the structure diagram is realized, the cost of the garment design and the plate making process is saved, the reaction speed of the whole manufacturing process is improved, the designer is started from the design diagram, the process is shortened, links are reduced, a great deal of manpower and time are saved, the development cost of the garment in the early stage is reduced, and the competitive advantage of a garment enterprise is improved.
Drawings
FIG. 1 is a block diagram of the overall circuit connection of the present invention;
FIG. 2 is a schematic diagram of a system framework of the present invention.
Detailed Description
The following detailed description of the invention refers to the accompanying drawings and preferred embodiments.
As shown in fig. 1 and 2, the present invention provides an intelligent clothing design system, including a processor, the processor is connected with an importing module, an identifying module, a modifying module and a database, the database is set to be that each component diagram in a plurality of types of clothing style diagrams corresponds to a structure diagram thereof one by one, the importing module is used for obtaining the clothing style diagram or the component diagram to be identified, the identifying module is used for identifying the clothing style diagram or the component diagram to be identified, and combining with the database to obtain the corresponding structure diagram thereof, the modifying module is used for virtually fitting each structure diagram with a human body model, modifying the structure diagram according to the fitting result, thereby providing an intelligent design platform for clothing designers, fully utilizing the existing clothing style diagrams and the structure diagrams corresponding thereto, avoiding repeated design, realizing conversion from the style diagrams to the structure diagrams, saving the cost of clothing design and plate making, and improving the reaction speed of the whole manufacturing process, starting from the design diagram, allowing the designer to directly participate in the plate making process, shortening the process, saving a great amount of manpower and time, and reducing the competition of clothing manufacturers in the front stage, and reducing the cost of clothing manufacturers.
A part of samples of the database are made clothing pattern drawings and template structure drawings provided by Shanghai PGM company, the pattern drawings and the templates are in one-to-one correspondence, but the sample amount is not great; the other part is a garment part drawing made by the inventor, and according to the requirements on the style drawing, the style drawing is made in CORELLARW X6 according to drawing standards. The collar can be divided into a lapel, a turnup collar, a standing collar, a continuous collar, a non-collar and the like, wherein the underside of each roughly classified collar can be subdivided into a plurality of types of collars, various common collar parts such as turnup collar parts are manufactured, the turnup collar clothing structure similar to a suit is mainly considered, 60 collar parts of each type are respectively manufactured, and the accumulation and collection can be continued later.
The leading-in module adopts a general tool of the existing CAD software to draw or uses a scanner to scan or uses USB to transmit data, and a clothing style diagram or a part diagram to be identified is obtained.
The identification module comprises a segmentation module, a vectorization module, a feature extraction module and a classification module.
The segmentation module is used for segmenting the clothing style graph to be identified to obtain each corresponding part graph, preprocessing is needed before segmentation in order to improve segmentation accuracy, the segmentation module specifically comprises a sharpening module and a binarization module, the sharpening module is used for sharpening the clothing style graph to be identified, laplace is adopted for carrying out Laplace, the boundary of the graph can be more obvious, the detail of the graph is enhanced, and the binarization module is used for carrying out binarization processing on the sharpened clothing style graph to be identified, wherein the threshold value of the binarization module is preferably 0.6. The segmentation module adopts a region growing method to segment a binarized clothing style diagram to be identified, firstly, growing points are manually selected on the binarized clothing style diagram, the difference between gray values of the points to be detected and the growing points is set to be 1 or 0 as a growing rule, if pixels with gray values of 8 are used as initial growing points, the part segmentation is carried out according to the growing rule.
Of course, before vectorization, the sharpening module and the binarization module are also required to be called to preprocess the clothing component diagram to be identified.
The vectorization module is used for carrying out edge recognition on each segmented component diagram or clothing component diagram to be recognized, analyzing the recognized edges to obtain a point set sequence, and carrying out vectorization on the point set sequence to obtain a vector component diagram to be detected, and the vectorization module is concretely as follows:
firstly, edge detection is carried out on a component diagram by adopting an edge detection method based on a Canny operator, then the detected edge is analyzed, points on the edge are stored into a point set sequence according to an analysis sequence, feature judgment is carried out on all points in the point set sequence by utilizing Hough transformation, if the features are consistent with collinear features, straight-line segment vectorization is carried out, otherwise, curve vectorization is carried out by adopting a piecewise three-time Bezier curve, and vectorization of the whole component diagram is completed, so that a vector component diagram to be detected is obtained.
The feature extraction module is used for carrying out chord feature extraction and normalization processing on the vector component diagram to be detected, and specifically comprises the following steps:
firstly, extracting the chord characteristics of each vector component diagram to be detected by adopting a chord characteristic extraction method to obtain three chord characteristic matrixes, namely an average projection length matrix PM, an outer chord length matrix OM and an absolute value matrix IODM of the difference between the inner chord length and the outer chord length.
Let the point set sequence be set c= { P i (x i ,y i ) I=1,.. T And T is a positive integer, the function expression of the binarized part graph is set as
x and y respectively represent the abscissa of the pixel point in the component diagram, D represents the region where the edge of the component diagram is located in the component diagram,
the expression of the outer chord matrix OM is set as:
the expression of the absolute value matrix IODM of the inner and outer chord length differences is set as:
the expression of the average projection length matrix PM is set as:
wherein, the liquid crystal display device comprises a liquid crystal display device,
represents the outer chord length->
Represents the inner chord length, a= (y) s+i -y i )/l s,i ,B＝(x s+i -x i )/l s,i ,C＝(x i y s+i -y i x s+i )/l s,i Delta (delta) represents the dirac function, x 1 And x 2 Respectively represent x i And x i+s Minimum and maximum of (2), y 1 And y 2 Respectively represent y i And y i+s S represents the number of unit length and takes on a value of 2 1 ,2 2 ,...,2 T-1 The unit length is defined as the distance between a point on an edge and its nearest neighbor, which distance is related to the size of the point set sequence,/o>
Representing the point P on the slave edge i (x i ,y i ) Starting from the edge, it reaches another point P in a counter-clockwise direction i+s (x i+s ,y i+s ) The passing chord length, ax+by+c=0, represents the point P i And P i+s Normal equation of the straight line determined, +.>
Representing the average projected length of the arc to the chord,
representing chord->
Point P on the corresponding arc i+t Projection distance to the chord.
From the mathematical model, it is known that all three matrices are (T-1) N, and that each value of s has a scale corresponding to it, i.e. the descriptor has (T-1) scales. It follows that each row of the chord feature matrix of this method corresponds to a scale level specific to each row. Where row 1 of the feature matrix is the smallest scale level, s=2 1 The last row of the matrix is the largest scale level, 2 T-1 Where t=log 2 N. The geometric shape feature information of the feature descriptions of different scale levels is different, the feature information of the features of the descriptions with small scale is finer, the shape information of the features of the descriptions with large scale is coarser, from the construction of the string feature matrix,it is from 2 1 Scale of (2) in an equal-ratio array T-1 Thus, the shape characteristics of the overall component profile can be fully described. The string feature extraction method used by the invention is used for describing the features of the component graph, the method is also called CFM, the feature description method can describe the convexity of the contour of the collar component, wherein the convexity of the edge of the collar component is described by extracting the inner and outer chord lengths of strings of a plurality of scale levels as feature descriptors of the style component, and the inner and outer chord lengths describe the convexity of the edge of the collar component well, so that the concavo-convex feature of the contour curve of the component is described more accurately.
Secondly, the chord characteristic descriptor normalization processing of the clothing component diagram to be identified is carried out, and the geometric shape of the plane style diagram is not changed under the transformation of translation, rotation and scaling, so that normalization processing is needed before classification is carried out, the situation that descriptors are different due to the occurrence of the geometric transformation is avoided, and finally component classification errors and inaccurate identification are caused.
(1) Component map translation
From the above OM, IODM, PM formula, it can be found that when the component translates, the values in the outer chord matrix, the absolute value of the difference between the inner chord and the outer chord and the value in the three matrices of the average projection length remain unchanged, so that the normalization processing on the translation change is not needed.
(2) Component map scaling
When the part map is scaled, the value of the chord changes. Assuming that the chord characteristic function of the component is changed from g (x, y) to g (ax, ay), where a > 0 is the scaling factor, the maximum value of each row in the three matrices of the characteristic descriptors OM, IODM, PM is used to normalize each element of the row. The constitution of the feature matrix shows that the feature elements in the same row correspond to the same scale level, and the different rows correspond to different scale levels, so that the invention respectively normalizes each row, and can ensure that the features of T-1 scale levels, namely the features of each row have the same contribution in the component recognition, and the failure that the feature value of the contour line of one row is too large and the feature value of the contour line of other rows is smaller can not occur.
(3) Rotation of part drawings
Taking the collar component diagram as an example, when the collar component diagram rotates, the strings on the edge lines of the component also rotate in the same way, and the inner chord length, the outer chord length and the average projection length from the arc to the strings of the edge lines can be known to be the same as those before rotation according to the expression OM, IODM, PM. However, rotation of the collar segment map causes a change in the initial position points of the extracted edge lines, which in turn causes sequential shifting of the rows of the three matrices of chord feature descriptors. If t represents the translation amount, wherein t is equal to or more than 1 and equal to or less than N is the translation amount, the characteristic value of the mth row and the kth column in the three matrixes is moved to the mth row and the kth+t column. In order to solve the problem, the invention uses one-dimensional discrete Fourier transform to normalize the rotation of the collar part graph.
The classification module is used for calling the vectorization module and the feature extraction module to process the component images in the database, classifying the component images in the processed database by adopting a support vector machine SVM-based classification method, finding out the category of the normalized component images to be tested, accurately identifying the component images to be tested in the category by adopting a nearest neighbor method 1NN, finding out the component image closest to the component images to be tested, and obtaining a structure diagram corresponding to the component images to be tested. The method combines the two classification modes to identify the clothing component diagram, classifies the unknown multi-dimensional multi-feature class component database for identifying the sample to be detected based on SVM, and then carries out 1NN classification identification mainly on the known class, calculates the distance between the detected sample and the component style of each class to judge which style component belongs to, and more specifically identifies the best matching component diagram, thereby being convenient for a designer to quickly and accurately find out the required style, improving the design efficiency, and identifying other clothing components as well as the collar components. In addition, the component diagrams can be arranged according to the difference degree order, so that various choices can be provided in the intelligent design of the clothing.
The modification module adopts CLO3D software to carry out virtual try-in and modification, and of course, other software capable of realizing the function can be adopted. The CLO3D software is very specialized clothing design software, is industry software developed by Korea CLO, can easily design a set of 3D model clothing through the software, and in the design process, a user only needs to guide the finished clothing into the model to know whether the clothing is designed to fit or not, the steps of real typing, material selection, body measurement, sewing and the like are not needed, and the clothing can be repeatedly modified according to fitting results, so that a great amount of time of the user is saved, and beautiful fashion can be designed only through simple mouse clicking actions. The system of the invention performs virtual sewing and fitting by combining the clothing designer with the structure diagram designed by the prior style diagram and structure diagram by means of the virtual fitting and modifying functions of the software, and can repeatedly modify the structure diagram according to fitting results until satisfaction.
By means of the system, a clothing designer can draw a style chart by himself or directly import an existing style chart when carrying out clothing design, no matter whether the style chart is an electronic draft or a hand draft, when drawing a specific part, the identification module can be utilized to check related parts in a database to finish drawing the style chart or the part chart, then the identification module can be utilized to obtain a corresponding structure chart, then the modification module is utilized to carry out virtual fitting on the designed structure chart, and the modification can be carried out according to fitting results, of course, if the obtained structure chart does not meet the requirements of the designer, the designer can utilize the modification module to modify the style chart so as to meet the requirements, thus, the existing style chart and structure chart resources can be fully utilized to provide reference for the clothing designer, the design efficiency is improved, the conversion from the style chart to the structure chart can be realized, and the combination with the printing plate, the sewing and the fitting can be shortened, links are reduced, a large amount of manpower and time are saved, the development cost of the clothing earlier stage is reduced, and the function level of a clothing design system is improved.
While particular embodiments of the present invention have been described above, it will be appreciated by those skilled in the art that these are merely illustrative, and that many changes and modifications may be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims.
Claims (4)
1. An intelligent garment design system, characterized in that: the device comprises a processor, an importing module, an identifying module, a modifying module and a database, wherein the database is arranged in a way that each component diagram in a plurality of types of clothing style diagrams corresponds to a structure diagram one by one, the importing module is used for acquiring the clothing style diagram or the component diagram to be identified, the identifying module is used for identifying the clothing style diagram or the component diagram to be identified, the corresponding structure diagram is acquired by combining the database, and the modifying module is used for virtually fitting each structure diagram with a human body model and modifying the structure diagram according to the fitting result;
the identification module comprises a segmentation module, a vectorization module, a feature extraction module and a classification module,
the segmentation module is used for segmenting the clothing style graph to be identified to obtain corresponding part graphs;
the vectorization module is used for carrying out edge identification on each segmented part drawing or the clothing part drawing to be identified, analyzing the identified edges to obtain a point set sequence, and carrying out vectorization on the point set sequence to obtain a vector part drawing to be detected;
the feature extraction module is used for carrying out chord feature extraction and normalization processing on the vector component diagram to be detected;
the classifying module is used for calling the vectorization module and the feature extraction module to process the component images in the database, classifying the processed component images by adopting a support vector machine SVM-based classifying method, finding out the category of the normalized component images to be tested, accurately identifying the component images to be tested in the category by adopting a nearest neighbor method 1NN, finding out the component image closest to the component images to be tested, and obtaining a structure diagram corresponding to the component images;
the chord characteristics are described by an average projection length matrix PM, an outer chord length matrix OM and an absolute value matrix IODM of the difference between the inner chord length and the outer chord length, the vectorization comprises straight line segment vectorization and curve segment vectorization, and the curve segment vectorization adopts a piecewise cubic Bezier curve for vectorization.
2. The intelligent garment design system of claim 1, wherein: the recognition module further comprises a preprocessing module, the preprocessing module is connected with the segmentation module and the vectorization module and used for preprocessing the clothing style graph or the component graph to be recognized, the preprocessing module comprises a sharpening module and a binarization module, the sharpening module adopts Laplace to sharpen the clothing style graph or the component graph to be recognized, and the binarization module is used for carrying out binarization processing on the sharpened clothing style graph or the component graph to be recognized.
3. The intelligent garment design system of claim 1, wherein: and the modification module adopts CLO3D software to modify virtual try-on and structural drawings.
4. The intelligent garment design system of claim 1, wherein: and the importing module adopts a universal tool of CAD software to draw or adopts a scanning instrument to scan so as to obtain a garment pattern diagram or a part diagram to be identified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910169133.4A CN109918783B (en) | 2019-03-06 | 2019-03-06 | Intelligent clothing design system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910169133.4A CN109918783B (en) | 2019-03-06 | 2019-03-06 | Intelligent clothing design system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109918783A CN109918783A (en) | 2019-06-21 |
| CN109918783B true CN109918783B (en) | 2023-06-02 |
Family
ID=66963458
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910169133.4A Active CN109918783B (en) | 2019-03-06 | 2019-03-06 | Intelligent clothing design system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109918783B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113505410A (en) * | 2021-07-13 | 2021-10-15 | 高瑞彤 | Intelligent garment design system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102044038A (en) * | 2010-12-27 | 2011-05-04 | 上海工程技术大学 | Three-dimensional virtual dressing method of clothes for real person |
| CN103106586A (en) * | 2011-11-11 | 2013-05-15 | 江南大学 | Three-dimensional virtual fitting system |
| CN103324860A (en) * | 2013-07-08 | 2013-09-25 | 刘优然 | Intelligent integrated costume designing and fitting system |
-
2019
- 2019-03-06 CN CN201910169133.4A patent/CN109918783B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102044038A (en) * | 2010-12-27 | 2011-05-04 | 上海工程技术大学 | Three-dimensional virtual dressing method of clothes for real person |
| CN103106586A (en) * | 2011-11-11 | 2013-05-15 | 江南大学 | Three-dimensional virtual fitting system |
| CN103324860A (en) * | 2013-07-08 | 2013-09-25 | 刘优然 | Intelligent integrated costume designing and fitting system |
Non-Patent Citations (3)
| Title |
|---|
| 弦特征矩阵：一种有效的用于植物叶片图像分类和检索的形状描述子;王斌 等;《计算机学报》;20171130;第40卷(第11期);第2559-2574页 * |
| 服装样片图像中直线与曲线矢量化的研究;杨杰 等;《天津工业大学学报》;20040229;第23卷(第01期);第55-58页 * |
| 服装款式图提取及其模式识别的研究;安立新;《中国博士学位论文全文数据库》;20160715(第07期);第21-39、43-50、55-69、90-106页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109918783A (en) | 2019-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106384126B (en) | Clothes fashion recognition methods based on contour curvature characteristic point and support vector machines | |
| Yue et al. | Automatic detection of feather defects using Lie group and fuzzy Fisher criterion for shuttlecock production | |
| CN110021028B (en) | Automatic clothing making method based on clothing style drawing | |
| Shi et al. | Automatic road crack detection using random structured forests | |
| CN105389593B (en) | Image object recognition methods based on SURF feature | |
| CN114930408A (en) | System, method and computer program product for automatically extracting information from a flowchart image | |
| CN110378225A (en) | The recognition methods of pointer instrument automatic reading and identification device | |
| CN107230203A (en) | Casting defect recognition methods based on human eye vision attention mechanism | |
| CN111858997B (en) | Cross-domain matching-based clothing template generation method | |
| CN108090913B (en) | Image semantic segmentation method based on object-level Gauss-Markov random field | |
| CN110648312A (en) | Method for identifying wool and cashmere fibers based on scale morphological characteristic analysis | |
| CN114445268A (en) | Garment style migration method and system based on deep learning | |
| De | Automatic data extraction from 2D and 3D pie chart images | |
| CN105354405A (en) | Machine learning based immunohistochemical image automatic interpretation system | |
| CN109918783B (en) | Intelligent clothing design system | |
| CN110232337B (en) | Chinese character image stroke extraction method and system based on full convolution neural network | |
| CN108090461A (en) | Three-dimensional face identification method based on sparse features | |
| CN107590806B (en) | Detection method and system based on brain medical imaging | |
| CN113793385A (en) | Method and device for positioning fish head and fish tail | |
| Iqbal et al. | Automated Meter Reading Detection Using Inception with Single Shot Multi-Box Detector. | |
| JP2019021100A (en) | Image search device, merchandise recognition device, and image search program | |
| CN109460781B (en) | Beefsteak grade division method based on decision tree induction learning | |
| CN112232119A (en) | Remote sensing texture image segmentation method and device | |
| Rathore et al. | CBISC: a novel approach for colon biopsy image segmentation and classification | |
| Blackledge et al. | Texture classification using fractal geometry for the diagnosis of skin cancers |
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 |