Multi-garment feature fusion method and system

Abstract

The invention relates to a multi-clothing feature fusion method and system, and relates to the technical field of clothing, and the method comprises the steps: obtaining a reference clothing image, a reference clothing image and triggering information; obtaining a mark fusion mode according to the trigger information; when the mark fusion mode is a preset orientation mode, determining a reference garment contour according to the reference garment image; determining a garment color style according to the reference garment image; determining a fused garment image according to the reference garment contour and the garment color style; when the mark fusion mode is a preset creative mode, determining a reference garment type according to the reference garment image; determining reference garment features according to the reference garment image, and determining a creative garment image according to the reference garment features and the reference garment type; and outputting the fused clothing image or the creative clothing image to a preset display area. The garment manufacturing method has the effect of improving the garment manufacturing efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of clothing, and in particular to a method and system for fusing multiple clothing features. Background Art

[0002] Clothing is a general term for fabrics and accessories that cover the human body, and it has both practical functions and cultural connotations.

[0003] In the process of clothing design, designers gain inspiration by observing nature, society and culture, and studying the history of clothing and costumes of different ethnic groups. They also use a variety of channels to collect materials, and then develop design ideas. After determining the theme, they draw sketches, clarify the outline and details of the clothing, and match the colors. The next step is pattern making, combining ergonomic design with pattern design, and selecting appropriate fabrics and accessories for cutting and making according to the purpose and style of the clothing. The finished garments are then let models try on to facilitate subsequent adjustments.

[0004] In the process of clothing design, it takes a lot of time to manually integrate clothing features such as clothing outline, color, and fabric to solve problems such as style conflicts, cultural misunderstandings, and large differences in production processes, which reduces the efficiency of clothing production. Summary of the invention

[0005] In order to improve the efficiency of clothing production, the present invention provides a method and system for fusing multiple clothing features.

[0006] In a first aspect, the present invention provides a method for fusing multiple clothing features, which adopts the following technical solution: A multi-clothing feature fusion method, comprising: Acquire a baseline clothing image, a reference clothing image, and trigger information; Acquire a marking fusion mode according to the trigger information; When the marker fusion mode is a preset directional mode, determining a reference clothing outline according to the reference clothing image; Determine clothing color style based on reference clothing images; Determine a fused clothing image according to a reference clothing outline and clothing color style; When the tag-based fusion mode is a preset creative mode, a reference clothing type is determined according to a reference clothing image; Determine reference clothing features according to the reference clothing image, and determine a creative clothing image according to the reference clothing features and a benchmark clothing type; The fused garment image or the creative garment image is output to a preset display area.

[0007] By adopting the above technical solution, the baseline clothing image and the reference clothing image are analyzed in the directional mode or the creative mode to obtain a fused clothing image or a creative clothing image, and output to the display area, so that clothing can be automatically designed through pictures, reducing the time required for manual design and improving the efficiency of clothing production.

[0008] Optionally, the method for determining a reference garment outline further comprises: When the reference clothing image only includes preset line features, determining the line drawing outline according to the reference clothing image and the line features; Delineate the contours based on the lines to determine contours with different numbers of line combinations; Screening is performed based on the line combination contours and the preset portrait line contours, and the line combination contours that are not portrait line contours are used as marked combination contours; Combine contours based on markers to determine contour distribution range; The line drawing clothing outline is determined according to the outline distribution range and the preset clothing wearing range, and the line drawing clothing outline is used as the reference clothing outline.

[0009] By adopting the above technical solution, the lines in the benchmark clothing image are analyzed to obtain the line combination contour, and the line combination contour is screened according to the portrait line contour and the clothing wearing range to obtain the benchmark clothing contour, so that the clothing contour can be extracted from the image with only lines.

[0010] Optionally, methods for determining creative clothing images include: Determine the marked reference garment according to the reference garment characteristics and the benchmark garment type; According to the marked reference clothing input into the preset clothing storage library, similar clothing images with similar features are matched; Reference garments based on markings to determine matching garment types; Determine creative clothing features based on labeled reference clothing and similar clothing images; Determine the matching clothing features based on the matching clothing type and creative clothing features; A creative clothing image is formed according to matching clothing features, matching clothing types, creative clothing features and marked reference clothing.

[0011] By adopting the above technical solution, the marked reference clothing and similar clothing images are analyzed to obtain the matching clothing type and matching clothing features, and the matching clothing features, matching clothing types, creative clothing features and marked reference clothing are integrated to form a creative clothing image, thereby making it possible to produce a creative clothing image.

[0012] Optionally, the method of outputting to a preset display area includes: Get model body information; Determine the character model based on the model's body information; defining the fused clothing image or the creative clothing image output to the display area as the labeled image information; Determine the reference contour size according to the marked image information; Determine the size of the garment to be tested based on the figure model and the reference outline size; The clothing model is determined according to the detected clothing size and the marked image information, and the clothing model is output to the display area.

[0013] By adopting the above technical solution, the model's body information and the marked image information are analyzed to obtain the detected clothing size, and the clothing model is obtained by detecting the clothing size and the model's body information, and the clothing model is output to the display area, so that the operator can easily view the upper body effect of the clothing.

[0014] Optionally, the method after determining the clothing model includes: Get the clothing wearing scene; Determine the wearing model based on the character model and clothing model; Determine the motion trajectory according to the clothing wearing scene and the character model, and control the wearing model to perform simulated movement according to the motion trajectory; Determine the garment contraction position based on the motion trajectory and wearing model; Fabric inspection based on wearing models, preset reference fabrics, and motion trajectories to identify markers where garments shrink; The fabric is detected according to the markers to update the wearing model, and the updated wearing model is output to the display area.

[0015] By adopting the above technical solution, through further analyzing the effect of the clothing on the body, when the model wears the clothing and moves along a motion trajectory, the fabric is marked and detected through the wearing model, the preset reference fabric and the motion trajectory to update the wearing model, and then the updated wearing model is output to the display area, so that the clothing can be adjusted in time to reduce the impact of the clothing on the person's activities.

[0016] Optionally, the method for determining the marked detection fabric includes: Determine the test extension coefficient and test stretch distance based on the wearing model, reference fabric and motion trajectory; Determine the contact pressure value according to the detection extension coefficient and the detection stretching distance; When the contact pressure value exceeds a preset reference pressure value, a marking pressure value is determined according to a wearing model, a preset detection fabric, and a motion trajectory; A detection fabric with a smaller pressure value than a reference pressure value is selected from the marked pressure values ​​as the marked detection fabric.

[0017] By adopting the above technical solution, the contact pressure value is obtained by analyzing the wearing model, the reference fabric and the motion trajectory. When the contact pressure value exceeds the preset reference pressure value, the marked detection fabric is selected through the wearing model, the preset detection fabric and the motion trajectory, so that the clothing can be adjusted in time to reduce the restraining effect of the clothing on the activities of the person.

[0018] Optionally, after determining the clothing model, the method further includes: When the contact pressure value does not exceed the reference pressure value, the degree of deformation of the garment contraction position is determined according to the wearing model, the reference fabric and the motion trajectory; Determine the total force applied in the motion based on the degree of deformation detected; Determine the fabric contraction force based on the total force and contact pressure value applied in the motion; Apply total force based on the reference fabric and motion to determine the reference friction force; Determine the fabric shrinkage distance based on the reference friction force and the fabric shrinkage force; When the fabric shrinkage distance exceeds the preset reference retraction distance, the target retraction distance is determined according to the wearing model, the preset detection fabric and the motion trajectory; Selecting a detection fabric corresponding to a target shrinkage distance greater than a reference shrinkage distance from the target shrinkage distance as the target detection fabric; Fabric detection based on target to update the wear model.

[0019] By adopting the above technical solution, when a model is wearing clothes and performing activities, the fabric shrinkage distance is obtained by analyzing the wearing model, the reference fabric and the motion trajectory. When the fabric shrinkage distance exceeds the reference retraction distance, the wearing model is used to select the target detection fabric to update the wearing model, the detection fabric and the motion trajectory, so as to timely adjust the fabric of the clothes to reduce the impact on the visual effect of the clothes caused by the excessive shrinkage distance.

[0020] Optionally, the method after updating the wearing model further includes: The target detection fabric and the marker detection fabric are used as adjustment fabrics; Adjust colors based on updated wearing models and adjusted fabrics; Determine the baseline ambient light intensity based on the clothing wearing scenario; Adjust gloss according to baseline ambient light, adjust color, and adjust fabric; Determine the baseline gloss level based on the baseline fabric; When the adjusted glossiness exceeds the reference glossiness, the detection glossiness is determined according to the adjusted color, the detection fabric and the reference ambient light illumination; The test fabric corresponding to the test glossiness that does not exceed the reference glossiness is selected from the test glossiness and defined as the marking adjustment fabric; Adjust fabrics based on markings to update the wear model.

[0021] By adopting the above technical solution, the glossiness of the adjusted fabric is analyzed. When the adjusted glossiness exceeds the baseline glossiness, it means that the effect presented by the clothing has a visual deviation. Then, the marked adjusted fabric is selected and updated through analysis of the adjusted color, detected fabric and baseline ambient light intensity to update the wearing model, thereby reducing the probability of visual deviation of clothing when adjusting the fabric of the clothing.

[0022] Optionally, the method after updating the wearing model further includes: Determine the gloss gradient area based on the model being worn; According to the gloss gradient position and adjusting the glossiness to determine the gloss variation range; Determine the selection range based on the gloss variation range; Selecting the adjusted glossiness that falls within the selected range from the adjusted glossiness as the marked glossiness; Determine continuous fabrics based on marking glossiness; Update the wearing model based on continuous fabric and gloss gradient areas.

[0023] By adopting the above technical solution, the gloss gradient is analyzed, and continuous fabrics are selected according to the selection range to adjust the wearing model, thereby improving the diversity of clothing production when selecting fabrics for clothing.

[0024] In the second aspect, the present application provides a multi-clothing feature fusion system, which adopts the following technical solution: A multi-clothing feature fusion system, comprising: An acquisition module is used to acquire a benchmark clothing image, a reference clothing image, trigger information, a marker fusion mode, model body information, and clothing wearing scenes; A memory, used for storing a multi-clothing feature fusion method; The processor is used to load, execute and implement the program stored in the memory.

[0025] In summary, the present application includes at least one of the following beneficial technical effects: 1. By analyzing the benchmark clothing image and the reference clothing image in the directional mode or creative mode to obtain a fused clothing image or a creative clothing image, and outputting it to the display area, clothing can be automatically designed through pictures, reducing the time required for manual design and improving the efficiency of clothing production; 2. Analyze the lines in the reference clothing image to obtain the line combination contour, and filter the line combination contour according to the portrait line contour and the clothing wearing range to obtain the reference clothing contour, so that the clothing contour can be extracted from the image with only lines; 3. By analyzing the gloss gradient and selecting continuous fabrics according to the selection range to adjust the wearing model, the diversity of clothing production can be improved when selecting clothing fabrics. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a method flow chart of a multi-clothing feature fusion method according to an embodiment of the present invention; Figure 2 is a flow chart of a method for determining a reference garment contour according to an embodiment of the present invention; Figure 3 is a flow chart of a method for determining a creative clothing image according to an embodiment of the present invention; Figure 4 is a flow chart of a method after outputting to a preset display area according to an embodiment of the present invention; Figure 5 This is the method flow after determining the clothing model in the embodiment of the present invention. Figure 1 ; Figure 6 is a flow chart of a method for determining a marking and detecting a fabric according to an embodiment of the present invention; Figure 7 This is the method flow after determining the clothing model in the embodiment of the present invention. Figure 2 ; Figure 8 This is the method flow after updating the wearing model in the embodiment of the present invention. Figure 1 ; Fig. 9 This is the method flow after updating the wearing model in the embodiment of the present invention. Figure 2 . DETAILED DESCRIPTION

[0027] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0028] Reference Figure 1 The present application embodiment discloses a method for fusing multiple clothing features, comprising the following steps: Step S100: Acquire a benchmark clothing image, a reference clothing image, and trigger information.

[0029] The benchmark clothing image refers to the benchmark image required by the operator when designing clothing. The benchmark acquisition module is a module set by the technician to acquire the benchmark clothing image, and the image retrieved from the benchmark acquisition module is used as the benchmark clothing image. The reference clothing image refers to the clothing image used as a reference when the operator designs clothing with the benchmark clothing image. The reference acquisition module is a module set by the technician to acquire the reference clothing image, and the image retrieved from the reference acquisition module is used as the reference clothing image. In this embodiment, the benchmark acquisition module and the reference acquisition module are set in the clothing software platform.

[0030] The trigger information refers to the control information of the clothing software platform when the operator clicks on it, and the trigger information is retrieved from the system of the clothing software platform.

[0031] Step S101: Acquire a marker fusion mode according to trigger information.

[0032] The preset reference mode is a mode used for designing clothing in a clothing software platform set by technical personnel.

[0033] The directional mode is a mode set by the technicians to retain the outline of the clothing in the reference clothing image and integrate the style of the reference clothing for design. The creative mode is a mode set by the technicians to retain the type of clothing in the reference clothing image and integrate the clothing features of the reference clothing for design. Clothing features are the color, style, outline, and pattern of clothing.

[0034] The basic modes include Directional Mode and Creative Mode.

[0035] Step S102: when the marker fusion mode is a preset orientation mode, a reference clothing outline is determined according to a reference clothing image.

[0036] The reference clothing contour refers to the contour of clothing in the reference clothing image. When the marker fusion mode is the directional mode, it is necessary to retrieve the contour of clothing in the reference clothing image, and the reference clothing image is analyzed to obtain the reference clothing contour.

[0037] Step S103: Determine the clothing color style according to the reference clothing image.

[0038] The clothing color style refers to the style of color arrangement of clothing in the reference clothing image, and the style of color arrangement of clothing in the reference clothing image is retrieved as the clothing color style.

[0039] Step S104: determining a fused clothing image according to the reference clothing outline and clothing color style.

[0040] The fused clothing image refers to an image formed by fusing the benchmark clothing image and the reference clothing image in the directional mode, and the image combining the benchmark clothing outline and the clothing color style is used as the fused clothing image.

[0041] Step S105: when the marker fusion mode is a preset creative mode, a reference clothing type is determined according to the reference clothing image.

[0042] The clothing repository is a library designed by technicians to store all the clothing that can be queried. The reference clothing type refers to the type of clothing in the reference clothing image. When the tag fusion mode is the creative mode, it is necessary to produce clothing that has not appeared in the clothing repository, and the type of clothing is identified from the reference clothing image. Image recognition technology is common knowledge among technicians in this field and will not be described in detail here.

[0043] Step S106: determining reference clothing features according to the reference clothing image, and determining a creative clothing image according to the reference clothing features and the benchmark clothing type.

[0044] The reference clothing feature refers to the feature of the clothing in the reference clothing image, and the reference clothing feature is identified from the reference clothing image. Image recognition technology is common knowledge among those skilled in the art and will not be described in detail here. The creative clothing image refers to a clothing image that does not appear in the clothing storage library, and the creative clothing image is obtained by analyzing the reference clothing feature and the benchmark clothing type.

[0045] Step S107: outputting the fused clothing image or the creative clothing image to a preset display area.

[0046] The display area is an area of ​​the clothing software platform set by the technician for displaying fused clothing images or creative clothing images, and the fused clothing images or creative clothing images are output to the display area for selection by the operator.

[0047] Reference Figure 2 , the method for determining the baseline garment contour further comprises: Step S200: when the reference clothing image only includes preset line features, determining a line drawing outline according to the reference clothing image and the line features.

[0048] The line features are the features such as the color and shape of the lines in the reference clothing image set by the technician. The line-drawing contour refers to the contour surrounded by the lines in the reference clothing image. When the reference clothing image only contains line features, it means that the reference clothing image is an image drawn by lines, and the contour surrounded by the line features is identified from the reference clothing image as the line-drawing contour.

[0049] Step S201: Describe the contour according to the lines to determine the contours combined by different numbers of lines.

[0050] The line combination contour refers to a contour composed of different numbers of adjacent line drawing contours, and the contour after combining different numbers of adjacent line drawing contours is used as the line combination contour.

[0051] Step S202: screening is performed based on the line combination contours and the preset portrait line contours, and the line combination contours that are not portrait line contours are used as marked combination contours.

[0052] The portrait line outline is the outline of the portrait drawn by the lines in the image set by the technician. The marked combined outline refers to the line outline of the clothing in the reference clothing image. By screening different line combined outlines, the line combined outlines that are not portrait line outlines are used as marked combined outlines. For example, the line between the neck and the shoulder is a continuous line. When the portrait in the line draft has the outline of the top, there is a line gap between the neck and the shoulder, so it is divided into two outlines, and the line outline of the neck is removed.

[0053] Step S203: combining contours according to the markers to determine the contour distribution range.

[0054] The contour distribution range refers to the distribution range of the line-drawn contour in the marked combined contour in the reference clothing image, and the distribution range of the line-drawn contour in the marked combined contour in the reference clothing image is used as the contour distribution range.

[0055] Step S204: determining the outline of the line drawing clothing according to the outline distribution range and the preset clothing wearing range, and using the line drawing clothing outline as the reference clothing outline.

[0056] The clothing wearing range is the range of clothing worn on the human body set by the technician. For example, the human body is divided into upper and lower parts by setting a reference ratio, and the range contained in the divided outline is used as the clothing wearing range.

[0057] Line drawing clothing outline refers to the outline that can identify the complete clothing type. For example, if the area of ​​a clothing range in the upper segmentation range exceeds the area of ​​the lower segmentation range, it is defined as a top, otherwise, it is defined as a bottom. If an overall outline is spread over the upper and lower segmentation outlines, the clothing is considered a one-piece type of clothing. The adjacent line drawing outlines within a segmentation range are recombined, and the combined outline is used as the line drawing clothing outline, and the line drawing clothing outline is used as the reference clothing outline.

[0058] Reference Figure 3 ,The methods for determining creative clothing images include: Step S300: Determine a marked reference garment according to the reference garment feature and the benchmark garment type.

[0059] The marked reference garment refers to a garment formed by fusing the reference garment feature with the benchmark garment type, and the garment formed by fusing the benchmark garment type with the reference garment feature is used as the marked reference garment.

[0060] Step S301: inputting the marked reference clothing into a preset clothing storage library to match similar clothing images with similar features.

[0061] Similar clothing images refer to images that are similar to clothing features in the marked reference clothing. The marked reference clothing is input into the clothing repository, and the clothing features in the marked reference clothing are extracted for retrieval and matching in the clothing repository, and the retrieved and matched clothing images are used as similar clothing images.

[0062] Step S302: Determine the matching garment type according to the marked reference garment.

[0063] Matching clothing types refer to clothing types that can be sewn and matched with marked reference clothing, and matching clothing types are matched from a preset matching database through marked reference clothing. For example, a pair of jeans can be cut with holes and multiple layers of cotton thread can be sewn on the holes to obtain new jeans, and the cotton thread is a matching clothing type.

[0064] The matching database stores matching clothing types corresponding to different marked reference clothing. The matching database is set in advance by humans and will not be described in detail here.

[0065] Step S303: determining creative clothing features according to the marked reference clothing and similar clothing images.

[0066] Creative clothing features refer to clothing features that can be matched on the marked reference clothing and do not overlap with similar clothing images. The clothing features are matched from the preset creative feature library by the marked reference clothing, and clothing features that do not overlap with clothing features in similar clothing images are selected from the matched clothing features as creative clothing features. The creative feature library stores clothing features corresponding to different marked reference clothing. The creative feature library is set in advance by humans and will not be described in detail here.

[0067] Step S304: determining matching clothing features according to the matching clothing type and the creative clothing features.

[0068] Matching clothing features refer to clothing features in creative clothing features that can be matched with clothing types as matching clothing features. The corresponding matching clothing features are matched from the matching database through the matching clothing type, and clothing features consistent with the creative clothing features are selected as matching clothing features.

[0069] The matching database also stores clothing features corresponding to different matching clothing types, which will not be described in detail here.

[0070] Step S305: forming a creative clothing image according to the matching clothing features, matching clothing types, creative clothing features and the marked reference clothing.

[0071] The clothing image formed by fusing the matching clothing features, matching clothing types, creative clothing features and labeled reference clothing is used as the creative clothing image.

[0072] Reference Figure 4 , the method after outputting to the preset display area includes: Step S400: Obtaining model body information.

[0073] Model body information refers to the body parameters of the model wearing the clothing on the clothing software platform, which can be obtained through pre-input by the operator.

[0074] Step S401: determining a character model according to the model's body information.

[0075] The character model refers to a three-dimensional model formed by the virtual data of the model. The character model is obtained by analyzing the model's body information. The method for generating the model is common knowledge to those skilled in the art and will not be elaborated here.

[0076] Step S402: defining the fused clothing image or creative clothing image output to the display area as marked image information.

[0077] The tagged image information refers to the fused clothing image or the creative clothing image output to the display area.

[0078] Step S403: Determine the reference outline size according to the marked image information.

[0079] The reference outline size refers to the size of the clothing in the marked image information displayed in the image. The reference outline size is obtained by analyzing the marked image information. The method of image recognition size is common knowledge to those skilled in the art and will not be described in detail here.

[0080] Step S404: determining the size of the clothing to be detected according to the character model and the reference outline size.

[0081] Detecting clothing size refers to the size of the clothing in the marked image information when worn on the human model. The ratio value of the clothing size change is obtained by identifying the overlap between the clothing and the human model. For example, the distance corresponding to the position where the human shoulder contacts the clothing is calculated and the size of the clothing at the shoulder position in the reference outline size is calculated to obtain the ratio value. Then the ratio value is calculated with the reference outline size to obtain the detected clothing size.

[0082] Step S405: determining a clothing model according to the detected clothing size and the marked image information, and outputting the clothing model to the display area.

[0083] The clothing model refers to the three-dimensional model corresponding to the virtual data of the clothing. The clothing model is obtained by analyzing the detected clothing size and the marked image information, and the clothing model is output to the display area. The method of generating the model is common knowledge known to those skilled in the art and will not be described in detail here.

[0084] Reference Figure 5 ,The methods after determining the clothing model include: Step S500: Acquire clothing wearing scene.

[0085] The clothing wearing scene refers to the scene where the character is wearing the clothing, which can be obtained through pre-input by the operator.

[0086] Step S501: Determine a wearing model according to the character model and the clothing model.

[0087] The wearing model refers to the model formed after the character model wears the clothing model. The wearing model is formed by combining the character model and the clothing model.

[0088] Step S502: determining a motion trajectory according to the clothing wearing scene and the character model, and controlling the wearing model to perform simulated motion according to the motion trajectory.

[0089] The motion trajectory refers to the trajectory of the character's movement in the clothing wearing scene. The clothing wearing scene is input into the preset motion database to match the motion amplitude, and the motion amplitude is combined with the character model to obtain the motion trajectory, and the wearing model is controlled to simulate the movement according to the motion trajectory. The motion database stores the body amplitude that the character needs to move in different clothing wearing scenes. The motion database is set by humans in advance and will not be elaborated here.

[0090] Step S503: Determine the garment contraction position according to the motion trajectory and the wearing model.

[0091] The clothing shrinkage position refers to the position where the clothing shrinks when the model is wearing it for simulated motion. The clothing shrinkage position can be the shoulder joint, elbow joint, wrist joint, etc. The position where the clothing wrinkles is retrieved from the model wearing it for simulated motion, and the position point where the clothing wrinkles shrink is used as the clothing shrinkage position.

[0092] Step S504: determining the marking detection fabric of the garment contraction position according to the wearing model, the preset reference fabric and the motion trajectory.

[0093] The reference fabric is the fabric used when the clothing model is initially generated and set by the technician. The marked detection fabric refers to the fabric sewn at the contraction position of the clothing, and the marked detection fabric is obtained by analyzing the wearing model, the reference fabric, and the motion trajectory. In this embodiment, the operator can select other fabrics as the reference fabric.

[0094] Step S505: Detect the fabric according to the mark to update the wearing model, and output the updated wearing model to the display area.

[0095] The shrinkage positions of fabrics and clothing are detected by marking and input into the wearing model to form a new wearing model, and the updated wearing model is output to the display area.

[0096] Reference Figure 6 , methods for determining the marking detection fabric include: Step S600: determining a detection extension coefficient and a detection stretching distance according to a wearing model, a reference fabric, and a motion trajectory.

[0097] The detection extension coefficient refers to the coefficient of the extension of the reference fabric at the contraction position of the clothing when the wearing model simulates the movement with the motion trajectory. The detection extension coefficient is matched by inputting the reference fabric and the body amplitude in the motion trajectory into the preset extension database. The detection stretch distance refers to the distance that the reference fabric at the contraction position of the clothing is stretched when the wearing model simulates the movement with the motion trajectory. When the wearing model simulates the movement with the motion trajectory, the clothing size that changes at the contraction position of the clothing is recorded, and the recorded size is compared with the detection clothing size to obtain the stretch distance.

[0098] Step S601: determining a contact pressure value according to a detected elongation coefficient and a detected stretching distance.

[0099] The contact pressure value refers to the pressure value that a person is subjected to when wearing clothing. The stretching pressure value is obtained by detecting the elongation coefficient and the stretching distance, and then the movement direction is retrieved from the movement trajectory. The motion component of the movement direction is calculated from the stretching pressure value as the contact pressure value.

[0100] Step S602: when the contact pressure value exceeds a preset reference pressure value, a marking pressure value is determined according to a wearing model, a preset detection fabric, and a motion trajectory.

[0101] The reference pressure value is the minimum pressure value of the clothing set by the technician. The detection fabric is other types of fabrics set by the technician and stored in the clothing software platform. When the contact pressure value exceeds the reference pressure value, it means that the clothing is too tight for the person to wear when performing activities. Then, the pressure value obtained by the detection fabric is referred to step S600 and step S601 as the marking pressure value.

[0102] Step S603: selecting a detection fabric with a smaller pressure value than a reference pressure value from the marked pressure values ​​as the marked detection fabric.

[0103] The marking detection fabric refers to the detection fabric whose marking pressure value is less than the reference pressure value, and the detection fabric whose marking pressure value is less than the reference pressure value is selected from the marking pressure values ​​as the marking detection fabric.

[0104] Reference Figure 7 , the method after determining the clothing model also includes: Step S700: When the contact pressure value does not exceed the reference pressure value, the degree of deformation of the contraction position of the garment is determined according to the wearing model, the reference fabric and the motion trajectory.

[0105] The degree of deformation detection refers to the degree of wrinkles on the clothing at the contraction position of the clothing. When the contact pressure value does not exceed the reference pressure value, it means that the clothing is not likely to be too tight when the person is wearing the clothing for activities. The number of wrinkles at the contraction position of the clothing is retrieved from the wearing model simulating movement, and the number of wrinkles and the reference fabric are input into the preset deformation database to match the degree of deformation detection.

[0106] The deformation database stores different numbers of wrinkles and the detected deformation degrees corresponding to the reference fabric. The deformation database is set in advance by humans and will not be described in detail here.

[0107] Step S701: Determine the total force applied by the motion according to the detected deformation degree.

[0108] The total force applied by the movement refers to the total force applied to the contraction position of the clothing when the wearing model simulates the movement with the movement trajectory. The total force applied by the movement is matched by inputting the detected deformation degree and the reference fabric into the preset movement database. The movement database stores the total force applied by the movement corresponding to different detected deformation degrees and fabrics. The movement database is set by humans in advance and will not be described in detail here.

[0109] Step S702: Determine the fabric contraction force according to the total force applied by the motion and the contact pressure value.

[0110] Fabric shrinkage force refers to the force of fabric shrinkage when the wearing model performs simulated movement. The difference between the total force applied in the movement and the contact pressure value is calculated as the fabric shrinkage force.

[0111] Step S703: Determine a reference friction force based on the reference fabric and the total force applied during the motion.

[0112] The reference friction force refers to the friction force between the reference fabric and the skin when the model is wearing it for simulated exercise. The friction coefficient is matched from the preset friction database through the reference fabric, and then the reference friction force is calculated by the friction coefficient and the total force applied by the movement. The calculation method of friction force is common knowledge for those skilled in the art and will not be described in detail here. The friction database stores the reference friction forces corresponding to different reference fabrics and the total force applied by the movement. The friction database is set in advance by humans and will not be described in detail here.

[0113] Step S704: determining the fabric shrinkage distance according to the reference friction force and the fabric shrinkage force.

[0114] The fabric shrinkage distance refers to the distance that the fabric will shrink when the model is wearing it for simulated exercise. The fabric shrinkage distance is matched from a preset shrinkage database through the reference friction force and the fabric shrinkage force. The shrinkage database stores fabric shrinkage distances corresponding to different reference friction forces and fabric shrinkage forces. The shrinkage database is set in advance by humans and will not be described in detail here.

[0115] Step S705: when the fabric shrinkage distance exceeds the preset reference shrinkage distance, a target shrinkage distance is determined according to the wearing model, the preset detection fabric and the motion trajectory.

[0116] The reference retraction distance is the maximum distance at which the fabric of the garment shrinks set by the technician. When the fabric shrinkage distance exceeds the reference retraction distance, it indicates that the garment needs to be adjusted, and the shrinkage distance obtained by detecting the fabric is taken as the target shrinkage distance by referring to step S700 and step S704.

[0117] Step S706: selecting a detection fabric corresponding to a target shrinkage distance greater than the reference shrinkage distance from the target shrinkage distances as the target detection fabric.

[0118] The target detection fabric refers to the detection fabric whose target shrinkage distance is greater than the reference retraction distance, and the detection fabric corresponding to the target shrinkage distance greater than the reference retraction distance is selected from the target shrinkage distance as the target detection fabric.

[0119] Step S707: Detect the fabric according to the target to update the wearing model.

[0120] The shrinkage positions of fabrics and clothing are detected by target and input into the wearing model to obtain a new wearing model.

[0121] Reference Figure 8 , the method after updating the wearing model also includes: Step S800: using the target detection fabric and the mark detection fabric as adjustment fabrics.

[0122] Adjustment fabric refers to fabric that needs to adjust the shrinkage position of clothing, and the target detection fabric and the mark detection fabric are used as adjustment fabric.

[0123] Step S801: Determine and adjust the color according to the updated wearing model and the adjusted fabric.

[0124] Adjusting the color means adjusting the color on the fabric by retrieving the color of the position where the fabric is to be adjusted from the updated wearing model as the adjustment color.

[0125] Step S802: Determine a reference ambient light intensity according to the clothing wearing scene.

[0126] The reference ambient illuminance refers to the illuminance displayed in the clothing wearing scene, and the reference ambient illuminance is retrieved from the clothing wearing scene.

[0127] Step S803: adjusting the glossiness according to the reference ambient light, adjusting the color and adjusting the fabric.

[0128] Adjusting the glossiness refers to adjusting the glossiness of the fabric. The reference ambient light illumination and the adjustment color are input into a preset glossiness database to match the adjusted glossiness. The glossiness database stores the adjusted glossiness corresponding to different reference ambient light illuminations and adjustment colors. The glossiness database is set by humans in advance and will not be described in detail here.

[0129] Step S804: Determine a reference glossiness according to a reference fabric.

[0130] The reference glossiness refers to the glossiness of fabrics that are not easy to match with the reference fabrics. The reference glossiness is matched from the gloss database through the reference fabrics. The gloss database also stores reference glossiness corresponding to different fabrics, which will not be described in detail here.

[0131] Step S805: when the adjusted glossiness exceeds the reference glossiness, the detection glossiness is determined according to the adjusted color, the detection fabric and the reference ambient light illumination.

[0132] When the adjusted glossiness exceeds the reference glossiness, it means that the fabric is not easy to match with the reference fabric, and the glossiness obtained by referring to step S801 and step S803 of the fabric is used as the detection glossiness.

[0133] Step S806: Select the detected fabric corresponding to the glossiness that does not exceed the reference glossiness from the detected glossiness and define it as the marked adjustment fabric.

[0134] The marked adjustment fabric refers to the test fabric corresponding to the test glossiness that does not exceed the reference glossiness. The test fabric corresponding to the test glossiness that does not exceed the reference glossiness is selected from the test glossiness and defined as the marked adjustment fabric.

[0135] Step S807: Adjust the fabric according to the mark to update the wearing model.

[0136] The shrinkage positions of fabrics and garments are adjusted by marking and input into the wearing model to form a new wearing model.

[0137] Reference Fig. 9 , the method after updating the wearing model also includes: Step S900: determining the gloss gradient position according to the wearing model.

[0138] The gloss gradient part refers to the part on the garment that is easy to be gloss gradient. For example, if the garment is a coat, the long sleeve part of the coat can be the gloss gradient part. The gloss gradient part is retrieved by the wearing model, and the gloss gradient part where the garment shrinks is used as the gloss gradient part.

[0139] Step S901: determining the gloss variation range according to the gloss gradient position and adjusting the glossiness.

[0140] The gloss variation range refers to the range formed by the glossiness of the reference fabric and the adjusted glossiness. The glossiness is obtained by analyzing the adjusted color and the reference fabric with reference to step S803, and the glossiness is combined with the adjusted glossiness to form the gloss variation range.

[0141] Step S902: determining a selection range according to the gloss variation range.

[0142] The selection range refers to the range used to select fabrics. The minimum range formed by extending the maximum and minimum values ​​of the gloss change range is used as the selection range. For example, if the gloss change range is 1-11, the selection range is 5-5, and when the gloss change range is 1-10, the selection range is 4-6.

[0143] Step S903: selecting the adjusted glossiness that falls within the selected range from the adjusted glossiness as the marked glossiness.

[0144] Marked gloss refers to fabrics that can have a gradual gloss change. The adjusted gloss within the selected range is selected as the marked gloss.

[0145] Step S904: Determine the continuous fabric according to the marked glossiness.

[0146] The continuous fabric refers to the adjusted fabric corresponding to the marked glossiness, and the adjusted fabric with the marked glossiness is used as the continuous fabric.

[0147] Step S905: updating the wearing model according to the continuous fabric and the gloss gradient area.

[0148] A new wearing model is formed by inputting continuous fabrics and gloss gradient areas into the wearing model.

[0149] Based on the same inventive concept, an embodiment of the present invention provides a multi-clothing feature fusion system, including: An acquisition module is used to acquire a benchmark clothing image, a reference clothing image, trigger information, a marker fusion mode, model body information, and clothing wearing scenes; A memory, used for storing a multi-clothing feature fusion method; The processor is used to load, execute and implement the program stored in the memory.

[0150] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0151] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A multi-clothing feature fusion method, characterized in that: include: Acquire a baseline clothing image, a reference clothing image, and trigger information; Acquire a marking fusion mode according to the trigger information; When the marker fusion mode is a preset directional mode, determining a reference clothing outline according to the reference clothing image; Determine clothing color style based on reference clothing images; Determine a fused clothing image according to a reference clothing outline and clothing color style; When the tag-based fusion mode is a preset creative mode, a reference clothing type is determined according to a reference clothing image; Determine reference clothing features according to the reference clothing image, and determine a creative clothing image according to the reference clothing features and a benchmark clothing type; The fused garment image or the creative garment image is output to a preset display area.

2. A multi-clothing feature fusion method according to claim 1, characterized in that: Methods for determining a baseline garment outline also include: When the reference clothing image only includes preset line features, determining the line drawing outline according to the reference clothing image and the line features; Delineate the contours based on the lines to determine contours with different numbers of line combinations; Screening is performed based on the line combination contours and the preset portrait line contours, and the line combination contours that are not portrait line contours are used as marked combination contours; Combine contours based on markers to determine contour distribution range; The line drawing clothing outline is determined according to the outline distribution range and the preset clothing wearing range, and the line drawing clothing outline is used as the reference clothing outline.

3. A multi-clothing feature fusion method according to claim 1, characterized in that: Methods for determining creative clothing images include: Determine the marked reference garment according to the reference garment characteristics and the benchmark garment type; According to the marked reference clothing input into the preset clothing storage library, similar clothing images with similar features are matched; Reference garments based on markings to determine matching garment types; Determine creative clothing features based on labeled reference clothing and similar clothing images; Determine the matching clothing features based on the matching clothing type and creative clothing features; A creative clothing image is formed according to matching clothing features, matching clothing types, creative clothing features and marked reference clothing.

4. A multi-clothing feature fusion method according to claim 1, characterized in that: The method after outputting to the preset display area includes: Get model body information; Determine the character model based on the model's body information; defining the fused clothing image or the creative clothing image output to the display area as the labeled image information; Determine the reference contour size according to the marked image information; Determine the size of the garment to be tested based on the figure model and the reference outline size; The clothing model is determined according to the detected clothing size and the marked image information, and the clothing model is output to the display area.

5. A multi-clothing feature fusion method according to claim 4, characterized in that: The methods after determining the clothing model include: Get the clothing wearing scene; Determine the wearing model based on the character model and clothing model; Determine the motion trajectory according to the clothing wearing scene and the character model, and control the wearing model to perform simulated movement according to the motion trajectory; Determine the garment contraction position based on the motion trajectory and wearing model; Fabric inspection based on wearing models, preset reference fabrics, and motion trajectories to identify markers where garments shrink; The fabric is detected according to the markers to update the wearing model, and the updated wearing model is output to the display area.

6. A multi-clothing feature fusion method according to claim 5, characterized in that: Methods for determining the marking of detected fabrics include: Determine the test extension coefficient and test stretch distance based on the wearing model, reference fabric and motion trajectory; Determine the contact pressure value according to the detection extension coefficient and the detection stretching distance; When the contact pressure value exceeds a preset reference pressure value, a marking pressure value is determined according to a wearing model, a preset detection fabric, and a motion trajectory; A detection fabric with a smaller pressure value than a reference pressure value is selected from the marked pressure values ​​as the marked detection fabric.

7. A multi-clothing feature fusion method according to claim 6, characterized in that: The method after determining the clothing model also includes: When the contact pressure value does not exceed the reference pressure value, the degree of deformation of the garment contraction position is determined according to the wearing model, the reference fabric and the motion trajectory; Determine the total force applied in the motion based on the degree of deformation detected; Determine the fabric contraction force based on the total force and contact pressure value applied in the motion; Apply total force based on the reference fabric and motion to determine the reference friction force; Determine the fabric shrinkage distance based on the reference friction force and the fabric shrinkage force; When the fabric shrinkage distance exceeds the preset reference retraction distance, the target retraction distance is determined according to the wearing model, the preset detection fabric and the motion trajectory; Selecting a detection fabric corresponding to a target shrinkage distance greater than a reference shrinkage distance from the target shrinkage distance as the target detection fabric; Fabric detection based on target to update the wear model.

8. A method for fusion of multiple clothing features according to claim 7, characterized in that: The method after updating the wearing model also includes: The target detection fabric and the marker detection fabric are used as adjustment fabrics; Adjust colors based on updated wearing models and adjusted fabrics; Determine the baseline ambient light intensity based on the clothing wearing scenario; Adjust gloss according to baseline ambient light, adjust color, and adjust fabric; Determine the baseline gloss level based on the baseline fabric; When the adjusted glossiness exceeds the reference glossiness, the detection glossiness is determined according to the adjusted color, the detection fabric and the reference ambient light illumination; The test fabric corresponding to the test glossiness that does not exceed the reference glossiness is selected from the test glossiness and defined as the marking adjustment fabric; Adjust fabrics based on markings to update the wear model.

9. A method for fusion of multiple clothing features according to claim 8, characterized in that: The method after updating the wearing model also includes: Determine the gloss gradient area based on the model being worn; According to the gloss gradient position and adjusting the glossiness to determine the gloss variation range; Determine the selection range based on the gloss variation range; Selecting the adjusted glossiness that falls within the selected range from the adjusted glossiness as the marked glossiness; Determine continuous fabrics based on marking glossiness; Update the wearing model based on continuous fabric and gloss gradient areas.

10. A multi-clothing feature fusion system, characterized in that: include: An acquisition module is used to acquire a benchmark clothing image, a reference clothing image, trigger information, a marker fusion mode, model body information, and clothing wearing scenes; A memory, used to store a multi-clothing feature fusion method according to any one of claims 1 to 9; The processor is used to load, execute and implement the program stored in the memory.

Images