Character outfit display method, system, terminal device and medium

Abstract

The present invention discloses a method, system, terminal device, and computer-readable storage medium for displaying a person's outfit. The method comprises: extracting a target person's body outline from a preset video data stream and determining the clothing items corresponding to the body outline; obtaining color information of the clothing items, and displaying the target person's outfit based on the color information. The present invention can automatically generate graphics and text summarizing a person's outfit, improving the efficiency of displaying a person's outfit.

Description

Technical Field

[0001] The present invention relates to the field of intelligent recognition technology, and in particular to a method, system, terminal device, and computer-readable storage medium for displaying a person's outfit. Background Art

[0002] Fashion dramas involve a large number of popular costumes in the clothing industry, presenting the latest popular combinations to the audience and leading the fashion trends, and thus deriving various fashion dressing tutorials or inventory pictures and texts.

[0003] Currently, the generation of outfit inventory pictures and texts is mainly done manually, which requires manual screenshots, sorting and adding text descriptions of the cast and crew's outfits in the play. This outfit inventory method has problems such as large workload, labor cost consumption and low efficiency. Summary of the Invention

[0004] The main purpose of the present invention is to provide a method, system, terminal device and computer-readable storage medium for displaying character outfits, aiming to automatically generate character outfit inventory graphics and texts to improve the efficiency of character outfit display.

[0005] To achieve the above-mentioned object, the present invention provides a method for displaying a person's outfit, the method comprising:

[0006] Extracting the target person's body outline from a preset video data stream and determining the clothing corresponding to the body outline;

[0007] Obtain color information of the clothing, and display the target person's clothing according to the color information.

[0008] The step of displaying the target person's outfit according to the color information includes:

[0009] Determine the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in a preset cast and crew image library, and obtain the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing;

[0010] Weighted scores are given to the outfit pictures of the cast and crew in the same type of outfit collection library, and the weighted scored outfit pictures of the cast and crew are displayed.

[0011] Optionally, the clothing corresponding to the human body silhouette includes: human body upper clothing and human body lower clothing;

[0012] The step of determining the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in the cast and crew image library, and obtaining the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing includes:

[0013] Determining, by a preset neural network, a first similarity between the color information of the upper garment of the human body and the color information of the upper garment of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of the lower garment of the human body and the color information of the lower garment of the other human silhouettes to be detected;

[0014] When both the first similarity and the second similarity reach a preset similarity threshold, the overall outfit image corresponding to the human body contour and the overall outfit images corresponding to the other human body contours to be detected are imported into a database to obtain a collection library of the same type of outfits.

[0015] Optionally, before the step of extracting the human body contour of the target person from the preset video data stream, the method further includes:

[0016] Obtain a facial skeleton simulation image corresponding to the facial image of the target person, and mark the facial orientation parameters in the facial image according to the facial parameters of the facial skeleton simulation image, so as to classify and display the target person's outfit in a graphic and text manner according to the facial orientation parameters.

[0017] Optionally, the step of determining clothing corresponding to the human body contour includes:

[0018] The human body contour is segmented according to a preset segmentation ratio to obtain a human upper garment region and a human lower garment region, and the human upper garment clothing corresponding to the human upper garment region and the human lower garment clothing corresponding to the human lower garment region are determined respectively.

[0019] Optionally, the step of displaying the weighted-scored outfit pictures of the cast and crew includes:

[0020] Based on the facial orientation parameters, and in combination with the video name, actor role name and shot type, the weighted scored cast and crew outfit pictures are classified and displayed according to the preset outfit graphic template.

[0021] Optionally, the color information includes: a color feature value, and the step of obtaining the color information of the clothing includes:

[0022] Determine the color space RGB of the upper body clothing area and the lower body clothing area respectively, and convert the color space RGB into the color space YUV and the color space HSV respectively;

[0023] The color feature value of the upper garment of the human body and the color feature value of the lower garment of the human body are determined according to the color space RGB, the color space YUV, and the color space HSV.

[0024] To achieve the above-mentioned purpose, the present invention further provides a character clothing display system, the character clothing display system comprising:

[0025] A determination module is used to extract the human body outline of the target person from the preset video data stream and determine the clothing corresponding to the human body outline;

[0026] The display module is used to obtain the color information of the clothing and display the clothing of the target person according to the color information.

[0027] Among them, each functional module of the character clothing display system of the present invention implements the steps of the character clothing display method described above during operation.

[0028] To achieve the above-mentioned purpose, the present invention also provides a terminal device, which includes: a memory, a processor, and a character outfit display program stored in the memory and runnable on the processor, and when the character outfit display program is executed by the processor, the steps of the character outfit display method described above are implemented.

[0029] In addition, to achieve the above-mentioned purpose, the present invention also proposes a computer-readable storage medium, on which a character dressing display program is stored. When the character dressing display program is executed by a processor, the steps of the character dressing display method as described above are implemented.

[0030] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the character clothing display method as described above.

[0031] The present invention provides a method, system, terminal device, computer-readable storage medium and computer program product for displaying a person's outfit. The method for displaying a person's outfit includes: extracting the human body contour of a target person from a preset video data stream, and determining the clothing corresponding to the human body contour; obtaining color information of the clothing, and displaying the outfit of the target person according to the color information.

[0032] Compared to the existing method of manually classifying the overall outfit of people in videos, this method automatically extracts the human body outline from the video data stream and displays the overall outfit according to the color information of the clothing corresponding to the human body outline. Therefore, this method can save the labor cost required for the repetitive work of manual screenshot comparison and editing, and improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of the hardware operating environment involved in an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of the first process of an embodiment of a method for displaying a person's outfit according to the present invention;

[0035] Figure 3 This is a second flow chart of an embodiment of a method for displaying a person's outfit according to the present invention;

[0036] Figure 4 This is a schematic diagram showing the classification of clothing types according to an embodiment of a method for displaying clothing of a character according to the present invention;

[0037] Figure 5 This is a schematic diagram of a clothing graphic template according to an embodiment of a method for displaying a character's clothing outfit according to the present invention;

[0038] Figure 6 This is a schematic diagram of human body contour segmentation according to an embodiment of a method for displaying a person's outfit according to the present invention;

[0039] Figure 7 This is a schematic diagram of the functional modules of an embodiment of a character outfit display system of the present invention.

[0040] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0041] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0042] like Figure 1 As shown, Figure 1 It is a schematic diagram of the device structure of the hardware operating environment involved in the embodiment of the present invention.

[0043] It should be noted that the terminal device in the embodiment of the present invention can be a terminal device for automatically generating outfit inventory pictures and texts. The terminal device can specifically be a mobile phone, a computer, a server, or a network device.

[0044] like Figure 1As shown, the device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0045] Those skilled in the art will understand that Figure 1 The device structure shown in the figure does not constitute a limitation of the device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0046] like Figure 1 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a character outfit display program. The operating system is a program that manages and controls the hardware and software resources of the device and supports the operation of the character outfit display program and other software or programs. Figure 1 In the device shown, the user interface 1003 is mainly used to communicate data with the client; the network interface 1004 is mainly used to establish a communication connection with the server; and the processor 1001 can be used to call the character outfit display program stored in the memory 1005 and perform the following operations:

[0047] Extracting the target person's body outline from a preset video data stream and determining the clothing corresponding to the body outline;

[0048] Obtain color information of the clothing, and display the target person's clothing according to the color information.

[0049] Furthermore, the processor 1001 may also be configured to call a character outfit display program stored in the memory 1005 and perform the following operations:

[0050] Determine the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in a preset cast and crew image library, and obtain the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing;

[0051] Weighted scores are given to the outfit pictures of the cast and crew in the same type of outfit collection library, and the weighted scored outfit pictures of the cast and crew are displayed.

[0052] Furthermore, the clothing corresponding to the human body silhouette includes: human body upper clothing and human body lower clothing. The processor 1001 can also be used to call the character clothing display program stored in the memory 1005, and perform the following operations:

[0053] The step of determining the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in the cast and crew image library, and obtaining the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing includes:

[0054] Determining, by a preset neural network, a first similarity between the color information of the upper garment of the human body and the color information of the upper garment of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of the lower garment of the human body and the color information of the lower garment of the other human silhouettes to be detected;

[0055] When both the first similarity and the second similarity reach a preset similarity threshold, the overall outfit image corresponding to the human body contour and the overall outfit images corresponding to the other human body contours to be detected are imported into a database to obtain a collection library of the same type of outfits.

[0056] Furthermore, before the step of extracting the human body contour of the target person from the preset video data stream, the processor 1001 may also be configured to call a character outfit display program stored in the memory 1005, and perform the following operations:

[0057] Obtain a facial skeleton simulation image corresponding to the facial image of the target person, and mark the facial orientation parameters in the facial image according to the facial parameters of the facial skeleton simulation image, so as to classify and display the target person's outfit in a graphic and text manner according to the facial orientation parameters.

[0058] Furthermore, the processor 1001 may also be configured to call a character outfit display program stored in the memory 1005 and perform the following operations:

[0059] The human body contour is segmented according to a preset segmentation ratio to obtain a human upper garment region and a human lower garment region, and the human upper garment clothing corresponding to the human upper garment region and the human lower garment clothing corresponding to the human lower garment region are determined respectively.

[0060] Furthermore, the processor 1001 may also be configured to call a character outfit display program stored in the memory 1005 and perform the following operations:

[0061] Based on the facial orientation parameters, and in combination with the video name, actor role name and shot type, the weighted scored cast and crew outfit pictures are classified and displayed according to the preset outfit graphic template.

[0062] Furthermore, the color information includes: color feature values. The processor 1001 can also be used to call the character outfit display program stored in the memory 1005, and perform the following operations:

[0063] Determine the color space RGB of the upper body clothing area and the lower body clothing area respectively, and convert the color space RGB into the color space YUV and the color space HSV respectively;

[0064] The color feature value of the upper garment of the human body and the color feature value of the lower garment of the human body are determined according to the color space RGB, the color space YUV, and the color space HSV.

[0065] Reference Figure 2 , Figure 2 The figure is a first flow chart of the first embodiment of the character outfit display method of the present invention. In this embodiment, the character outfit display method proposes to automatically split the outfits of the characters in the video to generate outfit inventory graphics, and the character outfit display method can be used in scenes such as film and television dramas and short videos. In this embodiment, a facial feature library of the cast and crew is first established, and then the video data stream is input. The character images of each frame in the video data stream are subjected to character recognition to determine the facial orientation, body contour and clothing of the cast and crew in each frame, so as to realize automatic classification of the outfits of each cast and crew. At the same time, it is also necessary to perform weighted scoring on the quality of the outfit pictures to generate disseminable graphic information, which saves the repetitive work of manual screenshot comparison and editing, thereby saving part of the labor cost and improving production efficiency.

[0066] Reference Figure 3 , Figure 3 This is a second flow chart of the first embodiment of the character outfit display method of the present invention.

[0067] Step S10, extracting the target person's body outline from the preset video data stream, and determining the clothing corresponding to the body outline;

[0068] After the terminal device obtains the character's clothing pictures in each frame from the preset video stream, in order to classify the character's overall clothing in each frame, it is necessary to extract the character's body contour from the character's clothing picture in advance and determine the clothing corresponding to the body contour.

[0069] It should be noted that, in this embodiment, before extracting the human body contour of the target person from the preset video data stream, a cast and crew gallery of each cast and crew member must be pre-built, and the cast and crew gallery includes all the outfit pictures of the cast and crew members in a certain video stream.

[0070] Specifically, for example, based on the known cast and crew list of a video file or live stream, images of the cast and crew are obtained from the Internet, and a preset neural network is constructed. The neural network is trained with a large number of facial images to extract feature values ​​and establish a facial feature library for each cast and crew member. The facial features of each frame in the preset video data stream are then compared with the facial feature parameters in the facial feature library to obtain facial feature similarity. When the facial feature similarity reaches a preset facial feature similarity threshold, the outfit images of the cast and crew member corresponding to the facial feature similarity are imported into the corresponding cast and crew library. The characters in the outfit images in the cast and crew library are then segmented, and an edge extraction operator, such as the Canny operator, is used to extract edges from the segmented character images to obtain a contour image. Finally, the contour image is optimized using morphological and related algorithms, including operations such as removing noise and connecting discontinuous edges, to obtain the human body contours of each cast and crew member.

[0071] Step S20, obtain the color information of the clothing, and display the clothing of the target person according to the color information.

[0072] It should be noted that, in this embodiment, taking into account that the actual clothing of the characters, including the upper and lower garments, is not a single color, but a mixture of multiple colors, in this embodiment, color information is used as the basis for dividing the clothing types of the cast and crew in each video frame, so as to classify the clothing types of the cast and crew in each video frame.

[0073] Specifically, for example, after obtaining the clothing corresponding to the human body outline, it is necessary to further obtain the color information of the clothing, and then classify the overall clothing of the cast and crew according to the color information to obtain multiple clothing types, and finally realize the display of the clothing of each cast and crew in the video stream, specifically as follows Figure 4 As shown, for example, the various types of outfits of cast member A are classified and displayed.

[0074] In this embodiment, after obtaining the character outfit images for each frame from a preset video stream, the terminal device extracts the character's body outline from the image and determines the clothing item corresponding to that body outline in order to classify the character's overall outfit in each frame. Furthermore, the terminal device obtains color information for the clothing item and, based on this color information, categorizes the cast and crew's overall outfit into multiple outfit types, ultimately displaying the outfit of each cast and crew member in the video stream.

[0075] Compared to the existing method of manually classifying the overall outfit of people in videos, this method automatically extracts the human body outline from the video data stream and displays the overall outfit according to the color information of the clothing corresponding to the human body outline. Therefore, this method can save the labor cost required for the repetitive work of manual screenshot comparison and editing, and improve production efficiency.

[0076] Furthermore, based on the first embodiment of the character clothing display method of the present invention, a second embodiment of the character clothing display method of the present invention is proposed.

[0077] In this embodiment, in the above step S20, “displaying the target person's outfit according to the color information” may include:

[0078] Step S201: Determine the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in a preset cast and crew image library, and based on the similarity, obtain the same type of clothing of the target person from the cast and crew image library to obtain a collection library of similar clothing;

[0079] Step S202: weightedly score the outfit pictures of the cast and crew in the same type outfit collection library, and display the weightedly scored outfit pictures of the cast and crew.

[0080] It should be noted that, in this embodiment, each cast member has a corresponding cast member gallery. For example, all the outfit pictures of cast member A in each frame of the video stream are in the cast member gallery a corresponding to actor A.

[0081] Furthermore, the terminal device first obtains the target person, such as the color information of the clothing of cast member A, and the similarity between the color information of the clothing of other human silhouettes to be detected in the cast member gallery a, and then determines the same type of clothing of cast member A in the video stream based on the similarity, so as to construct a collection library of similar clothing based on the same type of clothing of cast member A. After the terminal device classifies the clothing pictures of cast member A in each video frame to form multiple collection libraries of similar clothing of cast member A, it further weights and scores each clothing picture in the collection library of similar clothing based on parameters such as picture brightness, character screen ratio, and whether the character is centered, so as to obtain the clothing picture with the best quality, and displays each clothing picture after weighted scoring according to the preset clothing graphic template to obtain the clothing inventory graphic of the cast member corresponding to each human silhouette.

[0082] Specifically, for example, Figure 5As shown, the preset outfit graphic template includes the video name, role name, lens type and face orientation parameters. The terminal device will generate the outfit classification graphics for each frame of a certain video stream of each cast member according to the preset outfit graphic template, that is, Figure 4 When automatically generating outfit images and text according to the preset outfit image and text template settings, if the parameters of the outfit image and text template settings are met, the outfit images with higher weighted image quality scores will be displayed first. If the weighted scores are the same, the outfit images with the earlier order in the video timeline will be displayed.

[0083] Furthermore, in the above step S201, "determining the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in a preset cast and crew image library, and obtaining the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing", may include:

[0084] Step S2011, determining, through a preset neural network, a first similarity between the color information of the upper garment of the human body and the color information of the upper garments of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of the lower garment of the human body and the color information of the lower garments of the other human silhouettes to be detected;

[0085] Step S2012: When both the first similarity and the second similarity reach a preset similarity threshold, the overall outfit image corresponding to the human body contour and the overall outfit images corresponding to the other human body contours to be detected are imported into a database to obtain a collection library of the same type of outfits.

[0086] It should be noted that in this embodiment, in order to achieve classified display of the target person's clothing, the terminal device needs to pre-determine the clothing corresponding to the human body contour, including the upper and lower body clothing. The color information of the clothing can include color feature values.

[0087] Furthermore, on this basis, starting from the first picture in the cast and crew gallery where cast and crew member A is located, the terminal device calculates the similarity between the color feature values ​​of the upper garment of the human body in the first outfit picture and the color feature values ​​of the upper garment of the human body in the second outfit picture, that is, the first similarity, and at the same time calculates the similarity between the color feature values ​​of the lower garment of the human body in the first outfit picture and the color feature values ​​of the lower garment of the human body in the second outfit picture, that is, the second similarity, through a convolutional neural network. If the first similarity corresponding to the upper garment and the second similarity corresponding to the lower garment both reach the preset similarity threshold, the overall outfit in the first outfit picture and the overall outfit in the second outfit picture are classified into one type of outfit, and then the first outfit picture and the second picture are imported into the database to obtain a first outfit collection library of the same type. In this way, the first outfit image is compared with all other outfit images in the cast and crew image library to obtain the first outfit image library. The remaining images are then compared in sequence to form the second outfit collection library of the same type, the third outfit collection library of the same type, and so on. In addition, if the outfit image includes both tops and bottoms, the shot type is recorded as full body; if the outfit image only includes tops, the shot type is recorded as half body.

[0088] Furthermore, before the above step S10, "extracting the human body contour of the target person from the preset video data stream", the following steps may be further included:

[0089] Step S30, obtain the facial skeleton simulation image corresponding to the facial image of the target person, and mark the facial orientation parameters in the facial image of the person according to the facial parameters of the facial skeleton simulation image, so as to classify and display the target person's outfit in a graphic and text manner according to the facial orientation parameters.

[0090] It should be noted that, in this embodiment, Figure 5 As shown, since the preset outfit graphic template includes parameters such as video name, role name, lens type, and face orientation, the terminal device needs to pre-determine the facial orientation of each cast member's outfit picture in the cast and crew gallery.

[0091] Specifically, for example, the terminal device first determines the angle between the face of the character in the cast and crew gallery and the plane, and rotates it to the plane to generate a simulated skeleton image, identifies facial parameters such as the face contour, facial feature position and facial feature contour in the simulated skeleton image, and marks the facial orientation in each facial image according to the facial parameters such as the face contour, facial feature position and facial feature contour. The facial orientation can be marked as front, left and right.

[0092] Furthermore, in the above step S10, "determining the clothing corresponding to the human body contour" may include:

[0093] Step S101 , segmenting the human body contour according to a preset segmentation ratio to obtain a human upper garment region and a human lower garment region, and determining the human upper garment corresponding to the human upper garment region and the human lower garment corresponding to the human lower garment region respectively.

[0094] The terminal device can use a preset segmentation ratio to segment the human body contour to obtain the corresponding upper body area and lower body area, and then determine the upper body clothing corresponding to the upper body area and the lower body clothing corresponding to the lower body area.

[0095] Specifically, for example, Figure 6 As shown, the terminal device first determines the head position according to facial features, and then intercepts the upper and lower body parts according to different ratios to obtain the upper and lower body parts. Two-fifths of the upper body part can be selected as the upper body part of the human body. In addition, the height and width of the human body can be obtained based on the detected human body contour information. On this basis, according to a preset ratio, for example, based on the height value, the 60% area from the soles of the feet can be the lower body, and the next 30% area can be the upper body. For another example, from the left to the right side of the human body contour image, 30% of the width value can be used as the width value of the upper body part of the human body and the width value of the lower body part of the human body. Based on the above segmentation ratio, the height and width values ​​of the upper and lower body parts are determined, thereby realizing the regional image of the upper and lower body parts of the human body. In addition, images with only facial features and unable to identify the upper and lower body parts can be deleted.

[0096] Furthermore, in the above step S202, "displaying the weighted-scored outfit pictures of the cast and crew" may include:

[0097] Step S2021, based on the facial orientation parameters, and in combination with the video name, actor role name and shot type, the weighted scored cast and crew outfit pictures are classified and displayed according to the preset outfit graphic template.

[0098] It should be noted that, in this embodiment, Figure 5 As shown, the preset outfit graphic template includes: facial orientation parameters, video name, actor role name and lens type, etc.

[0099] Furthermore, after the terminal device obtains a collection library of similar outfits for the target person and weightedly scores the cast and crew outfit pictures in the collection library, in order to display the similar outfits for the target person in detail and improve the user experience, in this embodiment, the terminal device classifies and displays the cast and crew outfit pictures after the weighted scoring in combination with the video name, actor role name, shot type, and facial orientation parameters of the character in the cast and crew outfit picture corresponding to the cast and crew outfit picture.

[0100] Furthermore, in the above step S20, “obtaining the color information of the clothing” may include:

[0101] Step S203, respectively determining the color space RGB of the upper body clothing area and the lower body clothing area, and converting the color space RGB into the color space YUV and the color space HSV respectively;

[0102] Step S204 : determining the color feature value of the upper garment and the color feature value of the lower garment according to the color space RGB, the color space YUV, and the color space HSV.

[0103] The terminal device converts the RGB color space of the upper body area and the lower body area into the YUV color space and the HSV color space, respectively, and obtains the channel value corresponding to each channel of the color space for each pixel point of the image in the area, where:

[0104] The calculation formula for converting RGB color space to YUV color space is as follows:

[0105] Y=0.30R+0.59G+0.11B

[0106] U=0.493(BY)

[0107] V=0.877(RY) (1)

[0108] Among them, R, G, and B represent the red channel, green channel, and blue channel respectively, and Y, U, and V represent brightness perception, hue, and color saturation respectively.

[0109] The calculation formula for converting color space RGB to color space HSV is as follows:

[0110] max=max(R,G,B)

[0111] min=min(R,G,B)

[0112] if R=max,H=(GB) / (max-min)

[0113] if G=max,H=2+(BR) / (max-min)

[0114] if B=max,H=4+(RG) / (max-min)

[0115] H=H*60

[0116] if H<0,H=H+360

[0117] V=max(R,G,B)

[0118] S=(max-min) / max

[0119] Among them, H, S, and V represent hue, saturation, and tone, respectively. Then, the nine channels of the RGB, YUV, and HSV color spaces of the above regional images are fused to form new regional image data as the color feature values ​​of the upper and lower clothing of the human body.

[0120] In this embodiment, a preset neural network is used to determine a first similarity between the color information of a person's upper garment and the color information of the upper garments of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of a person's lower garment and the color information of the lower garments of other human silhouettes to be detected. If both the first similarity corresponding to the upper garment and the second similarity corresponding to the lower garment reach a preset similarity threshold, the overall outfit in the first outfit picture and the overall outfit in the second outfit picture are classified as one type of outfit, and the first outfit picture and the second outfit picture are then imported into the database to obtain a library of the same type of first outfit. In this way, the first outfit picture is compared with all other outfit pictures in the cast and crew image library to obtain a library of outfit one, and the remaining pictures are sequentially compared to form a library of the same type of outfit second outfit, a library of the same type of outfit third outfit, and so on. After the terminal device classifies the outfit images of cast member A in each video frame to form multiple outfit collection libraries of cast member A of the same type, it will further weight and score each outfit image in the same outfit collection library based on parameters such as image brightness, character screen ratio, and whether the character is centered, in order to obtain the outfit images with the best quality. It will then display each outfit image after weighting and scoring according to a preset outfit graphic template to obtain outfit inventory graphics of the cast member corresponding to each human body contour. Since the preset outfit graphic template includes parameters such as the video name, character name, lens type, and face orientation, the terminal device needs to pre-determine the facial orientation of each cast member's outfit image in the cast and crew library, and combine the video name, actor role name, and lens type to classify and display the weighted cast and crew outfit images according to the outfit graphic template.

[0121] In the present invention, the regional feature value is determined by the color space of each area to automatically classify the outfit types of the cast and crew based on the regional feature value. The color feature value of the upper body area corresponding to the human body contour and the similarity of the color feature value of the upper body area of ​​the human body contour to be detected are used to determine whether the outfit type corresponding to the human body contour is the same type as the outfit type of the human body contour to be detected. This realizes the classification of all the outfits of each cast and crew in the video stream, and automatically generates the outfit inventory graphics of each cast and crew according to the preset outfit graphic template. This aspect can replace the manual classification and inventory of the characters' outfits, reducing the operating cost and improving the operating efficiency. At the same time, the present invention can also be applied to various scenes such as film and television dramas and short videos.

[0122] In addition, the embodiment of the present invention also proposes a character clothing display system, referring to Figure 7 , Figure 7 This is a functional module diagram showing an embodiment of the character outfit of the present invention. Figure 7 As shown, the character clothing display system of the present invention includes:

[0123] A determination module 10 is configured to extract the human body outline of a target person from a preset video data stream and determine the clothing corresponding to the human body outline;

[0124] The display module 20 is used to obtain the color information of the clothing and display the clothing of the target person according to the color information.

[0125] Furthermore, the determining module 10 includes:

[0126] a similar outfit collection library acquisition unit, configured to determine the similarity between the color information of the clothing corresponding to the human body silhouette and the color information of the clothing of other human body silhouettes to be detected in a preset cast and crew image library, and based on the similarity, acquire similar outfits of the target person from the cast and crew image library to obtain a similar outfit collection library;

[0127] The display unit is used to weight the cast and crew outfit pictures in the same type of outfit collection library and display the cast and crew outfit pictures after the weighted scoring.

[0128] Furthermore, the clothing corresponding to the human body silhouette includes: human body upper clothing and human body lower clothing;

[0129] The unit for acquiring a collection library of outfits of the same type includes:

[0130] a similarity acquisition subunit, configured to determine, through a preset neural network, a first similarity between the color information of the upper garment of the human body and the color information of the upper garment of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of the lower garment of the human body and the color information of the lower garment of the other human silhouettes to be detected;

[0131] The subunit for obtaining a collection library of the same type of outfits is used to import the overall outfit image corresponding to the human body contour and the overall outfit images corresponding to the other human body contours to be detected into the database to obtain a collection library of the same type of outfits when both the first similarity and the second similarity reach a preset similarity threshold.

[0132] Furthermore, the character outfit display system further includes:

[0133] A facial orientation parameter acquisition unit is used to obtain a facial skeleton simulation image corresponding to the facial image of the target person, and mark the facial orientation parameters in the facial image of the person according to the facial parameters of the facial skeleton simulation image, so as to classify and display the target person's outfit in a graphic and textual manner according to the facial orientation parameters.

[0134] Furthermore, the determining module 10 includes:

[0135] The ratio segmentation module is used to segment the human body contour according to a preset segmentation ratio to obtain a human upper clothing area and a human lower clothing area, and respectively determine the human upper clothing corresponding to the human upper clothing area and the human lower clothing corresponding to the human lower clothing area.

[0136] Furthermore, the display unit includes:

[0137] The classification display sub-unit is used to classify and display the weighted-scored outfit pictures of the cast and crew according to the preset outfit graphic template based on the facial orientation parameters and in combination with the video name, actor role name and shot type.

[0138] Furthermore, the color information includes: color feature values, and the display module 20 includes:

[0139] a color space determining unit, configured to determine the color space RGB of the upper body clothing region and the lower body clothing region, respectively, and convert the color space RGB into a color space YUV and a color space HSV;

[0140] The color feature value determining unit is configured to determine the color feature value of the upper garment and the color feature value of the lower garment according to the color space RGB, the color space YUV, and the color space HSV.

[0141] The specific implementation methods of the various functional modules of the character outfit display system of the present invention are basically the same as the various embodiments of the above-mentioned character outfit display method, and will not be repeated here.

[0142] In addition, an embodiment of the present invention also proposes a computer-readable storage medium, on which a character outfit display program is stored. When the character outfit display program is executed by a processor, the steps of the character outfit display method as described above are implemented.

[0143] The various embodiments of the character outfit display system and computer-readable storage medium of the present invention can refer to the various embodiments of the character outfit display method of the present invention, and will not be repeated here.

[0144] In addition, an embodiment of the present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the steps of the character dressing display method as described in any embodiment of the above character dressing display method.

[0145] The specific embodiments of the computer program product of the present invention are basically the same as the embodiments of the above-mentioned character clothing display method, and will not be described in detail here.

[0146] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0147] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0148] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0149] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for displaying a person's outfit, characterized in that: The character outfit display method includes: Extracting the target person's body outline from a preset video data stream and determining the clothing corresponding to the body outline; Obtaining color information of the clothing, and displaying the clothing of the target person according to the color information; The step of displaying the target person's outfit according to the color information includes: Determine the similarity between the color information of the clothing and the color information of clothing of other human silhouettes to be detected in a preset cast and crew image library, and obtain similar clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing; Weighted scores are given to the outfit pictures of the cast and crew in the same type of outfit collection library, and the weighted scored outfit pictures of the cast and crew are displayed.

2. The method for displaying a person's outfit according to claim 1, wherein: The clothing corresponding to the human body silhouette includes: human body upper clothing and human body lower clothing; The step of determining the similarity between the color information of the clothing and the color information of the clothing of other human silhouettes to be detected in the cast and crew image library, and obtaining the same type of clothing of the target person from the cast and crew image library based on the similarity to obtain a collection library of similar clothing includes: Determining, by a preset neural network, a first similarity between the color information of the upper garment of the human body and the color information of the upper garment of other human silhouettes to be detected in the cast and crew image library, and a second similarity between the color information of the lower garment of the human body and the color information of the lower garment of the other human silhouettes to be detected; When both the first similarity and the second similarity reach a preset similarity threshold, the overall outfit image corresponding to the human body contour and the overall outfit images corresponding to the other human body contours to be detected are imported into a database to obtain a collection library of the same type of outfits.

3. The method for displaying a person's outfit according to any one of claims 1 to 2, wherein: Before the step of extracting the human body contour of the target person from the preset video data stream, the method further includes: Obtain a facial skeleton simulation image corresponding to the facial image of the target person, and mark the facial orientation parameters in the facial image according to the facial parameters of the facial skeleton simulation image, so as to classify and display the target person's outfit in a graphic and text manner according to the facial orientation parameters.

4. The method for displaying a person's outfit according to claim 2, wherein: The step of determining clothing corresponding to the human body contour includes: The human body contour is segmented according to a preset segmentation ratio to obtain a human upper garment region and a human lower garment region, and the human upper garment clothing corresponding to the human upper garment region and the human lower garment clothing corresponding to the human lower garment region are determined respectively.

5. The method for displaying a person's outfit according to claim 3, wherein: The step of displaying the weighted-scored outfit images of the cast and crew includes: Based on the facial orientation parameters, and in combination with the video name, actor role name and shot type, the weighted scored cast and crew outfit pictures are classified and displayed according to the preset outfit graphic template.

6. The method for displaying a person's outfit according to claim 5, wherein: The color information includes: a color feature value, and the step of obtaining the color information of the clothing includes: Determine the color space RGB of the upper body clothing area and the lower body clothing area respectively, and convert the color space RGB into the color space YUV and the color space HSV respectively; The color feature value of the upper garment of the human body and the color feature value of the lower garment of the human body are determined according to the color space RGB, the color space YUV, and the color space HSV.

7. A character clothing display system, characterized in that: The character outfit display system includes: A determination module is used to extract the human body outline of the target person from the preset video data stream and determine the clothing corresponding to the human body outline; A display module is used to obtain the color information of the clothing and display the clothing of the target person according to the color information; The display module is further configured to determine the similarity between the color information of the clothing and the color information of clothing of other human silhouettes to be detected in a preset cast and crew image library, and based on the similarity, obtain the same type of clothing of the target person from the cast and crew image library to obtain a collection library of similar clothing; Weighted scores are given to the outfit pictures of the cast and crew in the same type of outfit collection library, and the weighted scored outfit pictures of the cast and crew are displayed.

8. A terminal device, characterized in that: The terminal device includes a memory, a processor, and a character outfit display program stored in the memory and executable on the processor. When the character outfit display program is executed by the processor, the steps of the character outfit display method as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a character outfit display program, and when the character outfit display program is executed by the processor, the steps of the character outfit display method according to any one of claims 1 to 6 are implemented.

Images