Human body measurement method, device, electronic device and storage medium

Abstract

The present invention discloses a human body measurement method, device, electronic device and storage medium. The method obtains an image containing a target object, determines the human body skeleton point position information, a first human body mask map and a clothing mask map of the target object based on the image, and constructs a first three-dimensional human body model of the target object. Then, an adjustment parameter is determined according to the gap distance information corresponding to the measurement position of the target object, and the clothing mask map is adjusted according to the adjustment parameter, and then a second human body mask map is determined according to the adjusted clothing mask map. The first three-dimensional human body model is then rendered to obtain a third human body mask map; a target rendering loss is determined according to the third human body mask map and the second human body mask map, and finally the first three-dimensional human body model is adjusted according to the target rendering loss to obtain a second three-dimensional human body model; and human body measurement is performed according to the second three-dimensional human body model. Even for objects wearing ordinary clothes, more accurate measurement results can be obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of human body measurement, and in particular to a human body measurement method, device, electronic equipment and storage medium. Background Art

[0002] With the development of current technologies such as computer vision, human body measurement has become a research point and cutting-edge direction that has attracted much attention in the industry. Its applications in virtual reality, virtual fitting, virtual social interaction, etc. are increasing. Vision-based human body measurement solutions generally measure human bodies based on RGB images to detect information such as the chest, waist, and hip circumference of the human body.

[0003] In related technologies, if you want to obtain more accurate human body measurement results, the scanned human subject needs to wear no or only a small amount of close-fitting clothes, and then use the RGB camera to capture the front and side images for three-dimensional modeling, and calibrate the model according to the actual height. Then the final human body detection is achieved. The current problem is that if the scanned human subject wears ordinary clothes for human body measurement, due to the different fits of different clothes, it is very easy to cause large errors in the measurements of chest circumference, waist circumference, hip circumference, etc., resulting in inaccurate human body measurement. Summary of the invention

[0004] The embodiments of the present invention provide a human body measurement method, device, electronic device and storage medium, which are used to solve the problem of inaccurate measurement results when performing human body measurement on a subject wearing ordinary clothes in the prior art.

[0005] In a first aspect, an embodiment of the present invention provides a method for measuring a human body, the method comprising:

[0006] Acquire an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object;

[0007] determining an adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjusting the clothing mask image according to the adjustment parameter, and determining a second human body mask image according to the adjusted clothing mask image;

[0008] The first three-dimensional human body model is rendered to obtain a third human body mask image; a target rendering loss is determined according to rendering losses of the third human body mask image and the second human body mask image; the first three-dimensional human body model is adjusted according to the target rendering loss to obtain a second three-dimensional human body model; and human body measurement is performed according to the second three-dimensional human body model.

[0009] On the other hand, an embodiment of the present invention provides a human body measurement device, the device comprising:

[0010] A first determination module is used to obtain an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; and construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object;

[0011] a second determination module, configured to determine an adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjust the clothing mask image according to the adjustment parameter, and determine a second human body mask image according to the adjusted clothing mask image;

[0012] The measurement module is used to render the first three-dimensional human body model to obtain a third human body mask image; determine a target rendering loss according to the rendering losses of the third human body mask image and the second human body mask image; adjust the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model; and perform human body measurement according to the second three-dimensional human body model.

[0013] On the other hand, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0014] Memory, used to store computer programs;

[0015] The processor is used to implement any of the method steps described above when executing the program stored in the memory.

[0016] On the other hand, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any of the method steps described above is implemented.

[0017] The embodiment of the present invention provides a human body measurement method, device, electronic device and storage medium. The method includes: acquiring an image containing a target object, determining the human body skeleton point position information, a first human body mask map and a clothing mask map of the target object based on the image; constructing a first three-dimensional human body model based on the human body skeleton point position information and the first human body mask map of the target object; determining an adjustment parameter according to gap distance information corresponding to the measurement position of the target object, adjusting the clothing mask map according to the adjustment parameter, and determining a second human body mask map according to the adjusted clothing mask map; rendering the first three-dimensional human body model to obtain a third human body mask map; determining a target rendering loss according to the rendering loss of the third human body mask map and the second human body mask map; adjusting the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model; and performing human body measurement according to the second three-dimensional human body model.

[0018] In the embodiment of the present invention, an image containing a target object is obtained, and the human skeleton point position information, the first human mask map and the clothing mask map of the target object are determined based on the image, and after the first three-dimensional human model of the target object is constructed, an adjustment parameter is determined according to the gap distance information corresponding to the measured position of the target object, and the clothing mask map is adjusted according to the adjustment parameter, and then a second human mask map is determined according to the adjusted clothing mask map. The first three-dimensional human model is then rendered to obtain a third human mask map; the target rendering loss is determined according to the third human mask map and the second human mask map, and finally the first three-dimensional human model is adjusted according to the target rendering loss to obtain a second three-dimensional human model; and human body measurement is performed according to the second three-dimensional human model. The second three-dimensional human body model obtained by adjusting the target rendering loss is a three-dimensional human body model that fits the real human body better, so even if the object is wearing ordinary clothes, a three-dimensional human body model that fits the real human body better can be obtained, and then human body measurement is performed to obtain a more accurate measurement result. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A schematic diagram of a human body measurement process provided in Example 1 of the present invention;

[0021] Figure 2 A schematic diagram of a detailed process of human body measurement provided in Example 4 of the present invention;

[0022] Figure 3A schematic diagram of another human body measurement process provided in Embodiment 4 of the present invention;

[0023] Figure 4 A schematic diagram of a human body mask correction process provided by Embodiment 4 of the present invention;

[0024] Figure 5 A schematic diagram of a standing three-dimensional model provided in Example 4 of the present invention;

[0025] Figure 6 A schematic diagram of the structure of a human body measurement device provided in Example 5 of the present invention;

[0026] Figure 7 This is a schematic diagram of the structure of an electronic device provided in Example 6 of the present invention. DETAILED DESCRIPTION

[0027] The present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] Embodiment 1:

[0029] Figure 1 A schematic diagram of a human body measurement process provided by an embodiment of the present invention, the process includes the following steps:

[0030] S101: Acquire an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; and construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object.

[0031] S102: determining an adjustment parameter according to gap distance information corresponding to the measurement position of the target object, adjusting the clothing mask image according to the adjustment parameter, and determining a second human body mask image according to the adjusted clothing mask image.

[0032] S103: Rendering the first three-dimensional human body model to obtain a third human body mask image; determining a target rendering loss according to rendering losses of the third human body mask image and the second human body mask image; adjusting the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model; and performing human body measurement according to the second three-dimensional human body model.

[0033] The human body measurement method provided by the embodiment of the present invention is applied to an electronic device, which may be a PC, a tablet computer, or other device.

[0034] After the camera captures an image containing the target object, the electronic device obtains the image containing the target object captured by the camera. Based on the image, the human skeleton point position information, the first human mask image and the clothing mask image of the target object are determined. Then, based on the human skeleton point position information and the first human mask image of the target object, a first three-dimensional human model is constructed. The three-dimensional human model includes shape parameters, posture parameters and camera parameters, etc.

[0035] In an embodiment of the present invention, determining the human skeleton point position information, the first human mask image and the clothing mask image of the target object based on the image includes:

[0036] The position information of the human skeleton points of the target object in the image is determined based on the skeleton point detection model; the first human mask image of the target object is determined based on the human body segmentation model; and the clothing mask image of the target object is determined based on the clothing segmentation model.

[0037] The electronic device is pre-trained with a skeleton point detection model, a human body segmentation model and a clothing segmentation model. The skeleton point detection model is obtained by training with sample images labeled with human skeleton point position information. After the skeleton point detection model is trained, the image containing the target object is input into the skeleton point detection model to obtain the human skeleton point position information of the target object. The human body segmentation model is obtained by training with sample images labeled with human body mask images. After the human body segmentation model is trained, the image containing the target object is input into the human body segmentation model to obtain the first human body mask image of the target object. The clothing segmentation model is obtained by training with sample images labeled with clothing mask images. After the clothing segmentation model is trained, the image containing the target object is input into the clothing segmentation model to obtain the clothing mask image of the target object.

[0038] The electronic device obtains the gap distance information corresponding to the measurement position of the target object, and the measurement position is, for example, the position corresponding to the chest circumference, the position corresponding to the hip circumference, the position corresponding to the waist circumference, etc. The gap distance information corresponding to the measurement position of the target object refers to the gap distance information between the clothes and the human body corresponding to the measurement position. The gap distance information can be estimated by the target object and sent to the electronic device. The electronic device determines the adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjusts the clothes mask image according to the adjustment parameter, and determines the second human body mask image according to the adjusted clothes mask image.

[0039] The electronic device renders the first 3D human body model to obtain a third human body mask image, then calculates the rendering loss of the third human body mask image and the second human body mask image, and determines the target rendering loss according to the rendering loss of the third human body mask image and the second human body mask image. Finally, the first 3D human body model is adjusted according to the target rendering loss to obtain a second 3D human body model; and human body measurement is performed according to the second 3D human body model.

[0040] In the embodiment of the present invention, an image containing a target object is obtained, and the human skeleton point position information, the first human mask map and the clothing mask map of the target object are determined based on the image, and after the first three-dimensional human model of the target object is constructed, an adjustment parameter is determined according to the gap distance information corresponding to the measured position of the target object, and the clothing mask map is adjusted according to the adjustment parameter, and then a second human mask map is determined according to the adjusted clothing mask map. The first three-dimensional human model is then rendered to obtain a third human mask map; the target rendering loss is determined according to the third human mask map and the second human mask map, and finally the first three-dimensional human model is adjusted according to the target rendering loss to obtain a second three-dimensional human model; and human body measurement is performed according to the second three-dimensional human model. The second three-dimensional human body model obtained by adjusting the target rendering loss is a three-dimensional human body model that fits the real human body better, so even if the object is wearing ordinary clothes, a three-dimensional human body model that fits the real human body better can be obtained, and then human body measurement is performed to obtain a more accurate measurement result.

[0041] Embodiment 2:

[0042] The posture of the target object in the image may be varied, and there may be a problem of inaccurate human body measurement due to part occlusion. In order to solve the above problem, based on the above embodiment, in an embodiment of the present invention, after constructing the first three-dimensional human body model, before determining the adjustment parameter according to the gap distance information corresponding to the measured position of the target object, the method further includes:

[0043] The first three-dimensional human body model is transformed into a standing human body model by adjusting posture parameters, and the standing human body model is transformed according to the human body model height in the standing human body model and the real height information of the target object to obtain a third three-dimensional human body model, and the third three-dimensional human body model is used to replace the first three-dimensional human body model.

[0044] The three-dimensional human body model includes shape parameters, posture parameters, and camera parameters. By adjusting the posture parameters of the first three-dimensional human body model, the first three-dimensional human body model is transformed into a standing human body model. Performing human body measurement based on the standing human body model can avoid the problem of body part occlusion, thereby improving the accuracy of human body measurement. Further, according to the human body model height in the standing human body model and the real height information of the target object, the standing human body model is transformed to obtain a third three-dimensional human body model, and the first three-dimensional human body model is replaced with the third three-dimensional human body model. Then, the subsequent human body measurement steps are performed according to the first three-dimensional human body model obtained after the replacement. That is, the third three-dimensional human body model is the first three-dimensional human body model obtained after the replacement. It should be noted that the real height information of the target object can be the target object input into the electronic device. The human body model height in the standing human body model can be actively measured by the electronic device after obtaining the standing human body model.

[0045] Embodiment 3:

[0046] On the basis of the above embodiments, in an embodiment of the present invention, determining the adjustment parameter according to the gap distance information corresponding to the measured position of the target object includes:

[0047] Determine first length information of a measurement position of the target object according to a slice of the first three-dimensional human body model, and determine second length information in the clothing mask image corresponding to the measurement position; determine adjustment parameters according to the first length information, the second length information and predetermined gap distance information corresponding to the measurement position.

[0048] In an embodiment of the present invention, the first length information of the measurement position of the target object is determined according to the slice of the first three-dimensional human model. For example, taking the determination of the chest circumference of the target object as an example, the armpit point position is first determined according to the slice of the first three-dimensional human model, and then the maximum circumference polygon is searched nearby to determine the chest position, and the first length information of the chest position is determined to be N centimeters according to the slice of the first three-dimensional human model, and the second length information in the clothing mask corresponding to the chest position is determined to be M pixels, and the gap distance information corresponding to the predetermined chest position is offset. Then, the adjustment parameter is determined according to the first length information, the second length information and the gap distance information corresponding to the predetermined measurement position. Specifically, the adjustment parameter P = offset*M / N. The process of adjusting the clothing mask according to the adjustment parameter is to perform morphological operations on the clothing mask by selecting a specific kernel size, so that the corresponding part of the clothing mask is reduced to be consistent with the calculated P pixels. Similar operations are also used for other parts such as waist circumference and hip circumference.

[0049] In the embodiment of the present invention, determining the second human body mask image according to the adjusted clothing mask image includes: performing an AND operation on the adjusted clothing mask image and the first human body mask image to obtain the second human body mask image.

[0050] Embodiment 4:

[0051] On the basis of the above embodiments, in an embodiment of the present invention, determining the target rendering loss according to the rendering losses of the third human body mask image and the second human body mask image includes:

[0052] A target rendering loss is determined according to a first rendering loss between the third human body mask map and the first human body mask map, and a second rendering loss between the third human body mask map and the second human body mask map.

[0053] In the embodiment of the present invention, a first rendering loss of the third human body mask image and the first human body mask image, and a second rendering loss of the third human body mask image and the second human body mask image are determined respectively. Then, a target rendering loss is determined according to the first rendering loss and the second rendering loss. For example, corresponding weight values ​​are assigned to the first rendering loss and the second rendering loss respectively, and then the first rendering loss and the second rendering loss and the corresponding weight values ​​are weightedly summed to obtain a target rendering loss.

[0054] The step of adjusting the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model comprises:

[0055] According to the target rendering loss and the minimization loss principle, target shape parameters and target posture parameters are determined; according to the target shape parameters and target posture parameters, a second three-dimensional human body model is determined.

[0056] Figure 2 A detailed flow chart of human body measurement provided by an embodiment of the present invention, the process includes the following steps:

[0057] S201: Acquire an image containing a target object through a camera, and determine human skeleton point position information, a first human mask image, and a clothing mask image of the target object based on the image.

[0058] S202: Constructing a first three-dimensional human body model based on the human skeleton point position information of the target object and a first human body mask image, wherein the first three-dimensional human body model includes camera parameters, shape parameters, and posture parameters.

[0059] S203: transforming the first three-dimensional human body model into a standing human body model by adjusting the posture parameters, transforming the standing human body model according to the human body model height in the standing human body model and the real height information of the target object to obtain a third three-dimensional human body model; and replacing the first three-dimensional human body model with the third three-dimensional human body model.

[0060] S204: determining first length information of a measurement position of the target object according to the replaced slice of the first three-dimensional human body model, and determining second length information in the clothing mask image corresponding to the measurement position; determining an adjustment parameter according to the first length information, the second length information and predetermined gap distance information corresponding to the measurement position; adjusting the clothing mask image according to the adjustment parameter, and performing an AND operation on the adjusted clothing mask image and the first human body mask image to obtain a second human body mask image.

[0061] S205: Render the replaced first three-dimensional human body model to obtain a third human body mask image; determine a target rendering loss according to a first rendering loss between the third human body mask image and the first human body mask image, and a second rendering loss between the third human body mask image and the second human body mask image; determine a target shape parameter and a target posture parameter according to the target rendering loss and a loss minimization principle; determine a second three-dimensional human body model according to the camera parameters, the target shape parameter and the target posture parameter; perform plane interception analysis according to the second three-dimensional human body model to obtain various measurement parameters of the human body.

[0062] The human body measurement process provided by the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0063] Figure 3 The human body measurement process diagram provided by the embodiment of the present invention mainly includes 5 links, among which the key links for solving the improvement of clothing measurement error are S3 and S4. The following describes the 5 links respectively:

[0064] S1: Preprocessing the input image for subsequent use, including determining the position information of the human skeleton points of the target object in the image based on the skeleton point detection model; determining the first human mask of the target object based on the human segmentation model; and determining the clothing mask of the target object based on the clothing segmentation model.

[0065] S2: Based on the human skeleton point position information and the first human mask of the target object, the rough 3D human body parameters are obtained, whose posture and shape roughly match, but the scale is different from the real human body. Specifically, it includes using a deep learning model to regress the input human skeleton points and human mask, and mapping them to a parameterized 3D human model expressed in the camera parameters and shape and posture space.

[0066] S3: Scale the rough 3D human body model obtained in S2 and generate a corrected mask based on the offsets. Figure 4 Schematic diagram of the human body mask correction process.

[0067] a. First, different postures may block each other, which is not conducive to accurate measurement. In S2, we have obtained the shape parameters {β} and posture parameters {θ} of the 3D human body model, where θ expresses different human body postures. By fixing some parameter values ​​in θ, the 3D human body model is transformed into a standard A-shape, that is, converted into Figure 5 The standing 3D model shown in the figure has a scale difference from the real human body because the depth information of the human body is lost during 2D imaging, where scale = real human body height / reconstructed 3D human body model height. The scale is used to scale the coordinates of each vertex of the reconstructed 3D human body model to obtain a 3D human body model that roughly matches the real human body height.

[0068] b. Determine the armpit point position through A-shaped 3D human body model slicing, then search for the polygon with the largest circumference nearby to determine the chest position, and then estimate the number of pixels corresponding to the offset according to the chest cross-sectional width. Then use different sizes of checkpoints to perform corrosion operations on the corresponding parts of the clothing mask, and then align the mask with the human body mask to obtain a revised human body mask that better matches the real human body.

[0069] For example, the length of the chest section on the 3D human model is N centimeters, the length of the chest section of the clothing mask is M pixels, and offset represents the distance between the clothing and the real body estimated by the user, in centimeters. Then the number of pixels that should be reduced on the mask is P = offset*M / N. By selecting a specific kernel size to perform morphological operations on the mask, the corresponding part of the mask is reduced to be consistent with the calculated P pixels.

[0070] Similar operations are performed on other parts such as waist circumference and hip circumference.

[0071] S4: Use an optimization method to obtain 3D human model parameters that match the corrected human mask.

[0072] Since the 3D human body model can be expressed as a function f(θ, β) of shape {β} and posture {θ}, we want to obtain the optimal {β} and {θ} so that the reconstructed 3D human body model rendered back to the 2D image is closest to the corrected mask. If the rendering process is represented by K, the rendering loss L optimized_mask It can be expressed as:

[0073] L optimized_mask =Ioptimized_mask -I render

[0074] Among them I render =K(f(θ,β)) is the rendered mask image, I optimized_mask is the corrected mask image. By minimizing the loss through the quasi-Newton method, LM optimization algorithm, etc., the optimal shape parameter {β} is obtained, that is, the 3D human body model that best fits the corrected human body mask is obtained.

[0075] In order to back-propagate the gradients during the rendering process, the human body model obtained by S3 is rendered using the existing differentiable renderer. For two or more different human mask inputs of the front and side, the same set of shape parameters {β} are used for different inputs during the optimization process, ensuring the stability and consistency of the final 3D human body shape results when multiple input images are used.

[0076] At the same time, you can optionally add the uncorrected human mask loss and use α to control the weight, so that users can dynamically control it. The final rendering loss L render :

[0077] L render =(1-α)*L origin_mask +α*L optimized_mask

[0078] Where L optimized_mask is the corrected human mask rendering loss, L origin_mask is the initial human mask rendering loss, calculated in the same way as L optimized _ mask Similarly. When α is 0, the 3D human body model is fitted directly using the human body mask with clothes; when α is 1, the fitting is based on the user's subjective evaluation of the degree of clothing fit. By dynamically controlling the α value and displaying it in the optional visualization results, the user can also fine-tune the previously evaluated gap offset through the visualization results.

[0079] S5: Perform plane interception analysis on the human body Mesh obtained in S4 to obtain various parameters of the human body.

[0080] Embodiment 5:

[0081] Figure 6 This is a schematic diagram of the structure of a human body measurement device provided by an embodiment of the present invention. The device:

[0082] A first determination module 61 is used to obtain an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; and construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object;

[0083] A second determination module 62 is used to determine an adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjust the clothing mask image according to the adjustment parameter, and determine a second human body mask image according to the adjusted clothing mask image;

[0084] The measurement module 63 is used to render the first three-dimensional human body model to obtain a third human body mask image; determine a target rendering loss according to the rendering losses of the third human body mask image and the second human body mask image; adjust the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model; and perform human body measurement according to the second three-dimensional human body model.

[0085] The device also includes:

[0086] The updating module 64 is used to transform the first three-dimensional human body model into a standing human body model by adjusting posture parameters, transform the standing human body model according to the human body model height in the standing human body model and the real height information of the target object to obtain a third three-dimensional human body model, and use the third three-dimensional human body model to replace the first three-dimensional human body model.

[0087] The first determination module 61 is specifically used to determine the human skeleton point position information of the target object in the image based on the skeleton point detection model; determine the first human mask image of the target object based on the human segmentation model; and determine the clothing mask image of the target object based on the clothing segmentation model.

[0088] The second determination module 62 is specifically used to determine the first length information of the measurement position of the target object according to the slice of the first three-dimensional human body model, and determine the second length information of the measurement position corresponding to the clothing mask image; determine the adjustment parameter according to the first length information, the second length information and the predetermined gap distance information corresponding to the measurement position.

[0089] The second determining module 62 is specifically configured to perform an AND operation on the adjusted clothing mask image and the first human body mask image to obtain a second human body mask image.

[0090] The measuring module 63 is specifically configured to determine a target rendering loss according to a first rendering loss between the third human body mask image and the first human body mask image, and a second rendering loss between the third human body mask image and the second human body mask image.

[0091] The measuring module 63 is specifically used to determine the target shape parameters and the target posture parameters according to the target rendering loss and the minimization loss principle; and determine the second three-dimensional human body model according to the target shape parameters and the target posture parameters.

[0092] Embodiment 7:

[0093] Based on the above embodiments, an electronic device is also provided in an embodiment of the present invention, such as Figure 7 As shown, it includes: a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304;

[0094] The memory 303 stores a computer program. When the program is executed by the processor 301, the processor 301 performs the following steps:

[0095] Acquire an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object;

[0096] determining an adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjusting the clothing mask image according to the adjustment parameter, and determining a second human body mask image according to the adjusted clothing mask image;

[0097] The first three-dimensional human body model is rendered to obtain a third human body mask image; a target rendering loss is determined according to rendering losses of the third human body mask image and the second human body mask image; the first three-dimensional human body model is adjusted according to the target rendering loss to obtain a second three-dimensional human body model; and human body measurement is performed according to the second three-dimensional human body model.

[0098] Based on the same inventive concept, an electronic device is also provided in an embodiment of the present invention. Since the principle of solving the problem by the above electronic device is similar to that of the human body measurement method, the implementation of the above electronic device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0099] The electronic device provided in the embodiment of the present invention may specifically be a desktop computer, a portable computer, a smart phone, a tablet computer, a personal digital assistant (PDA), a network side device, etc.

[0100] The communication bus mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0101] The communication interface 302 is used for communication between the electronic device and other devices.

[0102] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0103] The above-mentioned processor can be a general-purpose processor, including a central processing unit, a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc.

[0104] Embodiment 8:

[0105] On the basis of the above embodiments, an embodiment of the present invention further provides a computer storage readable storage medium, wherein the computer readable storage medium stores a computer program executable by an electronic device, and when the program is run on the electronic device, the electronic device implements the following steps when executing:

[0106] Acquire an image containing a target object, determine the human skeleton point position information, a first human mask map and a clothing mask map of the target object based on the image; construct a first three-dimensional human body model based on the human skeleton point position information and the first human mask map of the target object;

[0107] determining an adjustment parameter according to the gap distance information corresponding to the measurement position of the target object, adjusting the clothing mask image according to the adjustment parameter, and determining a second human body mask image according to the adjusted clothing mask image;

[0108] The first three-dimensional human body model is rendered to obtain a third human body mask image; a target rendering loss is determined according to rendering losses of the third human body mask image and the second human body mask image; the first three-dimensional human body model is adjusted according to the target rendering loss to obtain a second three-dimensional human body model; and human body measurement is performed according to the second three-dimensional human body model.

[0109] Based on the same inventive concept, a computer-readable storage medium is also provided in an embodiment of the present invention. Since the principle of solving the problem when the processor executes the computer program stored on the above-mentioned computer-readable storage medium is similar to the human body measurement method, the implementation of the processor when executing the computer program stored on the above-mentioned computer-readable storage medium can refer to the implementation of the method, and the repeated parts will not be repeated.

[0110] The above-mentioned computer-readable storage medium can be any available medium or data storage device that can be accessed by the processor in the electronic device, including but not limited to magnetic storage such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc., optical storage such as CD, DVD, BD, HVD, etc., and semiconductor storage such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD), etc.

[0111] An embodiment of the present invention also provides a computer storage readable storage medium, which stores a computer program that can be executed by an electronic device. When the program runs on the electronic device, it can be used to execute the interface call timing detection method provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0112] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for measuring human body, characterized in that: The method comprises: Acquire an image containing a target object, determine the position information of human skeleton points of the target object in the image based on a skeleton point detection model; determine a first human mask image of the target object based on a human segmentation model; determine a clothing mask image of the target object based on a clothing segmentation model; construct a first three-dimensional human model based on the human skeleton point position information of the target object and the first human mask image; Determine first length information of a measurement position of the target object according to the slice of the first three-dimensional human body model, and determine second length information in the clothing mask image corresponding to the measurement position; determine adjustment parameters according to the first length information, the second length information, and predetermined gap distance information corresponding to the measurement position, adjust the clothing mask image according to the adjustment parameters, and determine a second human body mask image according to the adjusted clothing mask image; The first three-dimensional human body model is rendered to obtain a third human body mask image; a target rendering loss is determined according to rendering losses of the third human body mask image and the second human body mask image; the first three-dimensional human body model is adjusted according to the target rendering loss to obtain a second three-dimensional human body model; and human body measurement is performed according to the second three-dimensional human body model.

2. The method according to claim 1, characterized in that After constructing the first three-dimensional human body model and before determining the adjustment parameter according to the gap distance information corresponding to the measured position of the target object, the method further includes: The first three-dimensional human body model is transformed into a standing human body model by adjusting posture parameters, and the standing human body model is transformed according to the human body model height in the standing human body model and the real height information of the target object to obtain a third three-dimensional human body model, and the third three-dimensional human body model is used to replace the first three-dimensional human body model.

3. The method according to claim 1, characterized in that Determining the second human body mask image according to the adjusted clothing mask image comprises: The adjusted clothing mask image is ANDed with the first human body mask image to obtain a second human body mask image.

4. The method according to claim 1, characterized in that The determining a target rendering loss according to the rendering losses of the third human body mask image and the second human body mask image comprises: A target rendering loss is determined according to a first rendering loss between the third human body mask map and the first human body mask map, and a second rendering loss between the third human body mask map and the second human body mask map.

5. The method according to claim 1, characterized in that The step of adjusting the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model comprises: According to the target rendering loss and the minimization loss principle, target shape parameters and target posture parameters are determined; according to the target shape parameters and target posture parameters, a second three-dimensional human body model is determined.

6. A human body measurement device, characterized in that: The device comprises: A first determination module is used to obtain an image containing a target object, determine the position information of the human skeleton points of the target object in the image based on a skeleton point detection model; determine a first human mask image of the target object based on a human segmentation model; determine a clothing mask image of the target object based on a clothing segmentation model; and construct a first three-dimensional human model based on the human skeleton point position information of the target object and the first human mask image; a second determination module, configured to determine first length information of a measurement position of the target object according to the slice of the first three-dimensional human body model, and determine second length information in the clothing mask image corresponding to the measurement position; determine an adjustment parameter according to the first length information, the second length information, and predetermined gap distance information corresponding to the measurement position, adjust the clothing mask image according to the adjustment parameter, and determine a second human body mask image according to the adjusted clothing mask image; The measurement module is used to render the first three-dimensional human body model to obtain a third human body mask image; determine a target rendering loss according to the rendering losses of the third human body mask image and the second human body mask image; adjust the first three-dimensional human body model according to the target rendering loss to obtain a second three-dimensional human body model; and perform human body measurement according to the second three-dimensional human body model.

7. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor is used to implement the human body measurement method described in any one of claims 1 to 5 when executing the program stored in the memory.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the human body measurement method according to any one of claims 1 to 5 is implemented.

Images