CN114943705B - Image data acquisition method for planar work or product attached to deformed substrate - Google Patents

Abstract

The invention discloses a method and a device for acquiring image data of a planar work or an article attached to a deformed substrate, comprising the following steps: the planar work or product is unfolded and fixed to the maximum extent by means of not changing the state of the substrate material and not damaging the substrate material; acquiring a digital three-dimensional model with high reduction degree of the planar work or product which is actually in a three-dimensional state due to the deformation of the substrate material; flattening the high-reduction-degree digital three-dimensional model of the planar work or the product which is actually in a three-dimensional state due to the deformation of the base material; and storing the image and the size information of the planar work or the product which is in the absolute two-dimensional space after flattening for later use. The invention fundamentally solves the problem of information distortion which is difficult to avoid when the camera or the two-dimensional scanning equipment is simply relied on to acquire the image data, thereby providing powerful guarantee for digital protection, research, appreciation and collection of planar works or products.

Description

Image data acquisition method for planar work or product attached to deformed substrate

Technical Field

The invention belongs to the technical field of digital information acquisition and protection of planar works or products, and particularly relates to a method for carrying out non-physical and digital flattening on the planar works or products which are actually in a three-dimensional state due to deformation of a base material, so that the planar works or products are in an absolute two-dimensional space representation state, namely an original representation state of the planar works or products in an creation or manufacturing stage, and protecting and displaying the planar works or products by utilizing big data and a computer vision technology.

Background

The development of digital technology brings new spring for the digital information acquisition and protection of planar works or products. With the increasing maturity of high-precision digital photography, high-precision planar scanning, big data technology and image comparison technology, the collection, protection, research, appreciation and display of planar works or products are served with great convenience and guarantee.

However, the planar work or article, whether it is drawn or written, printed or rendered, is attached to a particular substrate material, such as paper, silk, iron sheet, glass, etc., and any substrate material, either due to expansion and contraction with heat, or due to rubbing, pressing, folding, globing, it is easy to cause a certain degree of deformation such as curling, swelling, wrinkling, and cracking, etc., and thus the planar properties exhibited at the stage of creation or production of a planar work or article are lost, i.e., the state is changed from a purely two-dimensional space state to a three-dimensional space state. Inevitably, planar works or articles attached to these base materials are also changed in a practical sense into a solid article.

If a planar work or a product which is curled, swelled, wrinkled or broken due to deformation of a base material is not flattened, the real appearance of the planar work or the product in the creation or manufacturing stage cannot be recorded truly by simply adopting a photographing or planar scanning method, namely, the expression of patterns, colors, characters and the like in an absolute two-dimensional space state is not real and reliable, and the direct consequence is that the image information used for protection, research, appreciation and display is not real and reliable. Moreover, if the repair and research of the museum on the planar collection is based on erroneous data information, the repair or research results obtained thereby become worthless.

For a long time, appreciation and research of planar works or articles depend on the physical object itself rather than digitized images, and therefore, it is not a trouble that planar works or articles actually become three-dimensional due to deformation of a base material.

Unfortunately, despite the advent of the digital age, the current state of the art of the fact that planar works or articles are three-dimensional due to deformation of the substrate material has not received sufficient attention, at most, to make some fixation during shooting or scanning without damaging the planar work or article, or to cover the planar work or article that has been deformed with large glass in order to obtain a flattening effect. However, this is not an effective solution because the glass is capable of reflecting and refracting light, compromising the high quality reduction effect on flat works or articles during shooting and scanning, and even because of the weight of the glass, which instead stresses certain substrate materials such as soft textiles and the like, causing new deformations. Moreover, some deformations, such as wrinkles, bulges, or recesses and protrusions of the sheet, are generally difficult to recover and cannot be flattened with a single glass sheet.

The invention aims to explore a new way for thoroughly flattening the planar work or product on the premise of not changing and damaging the material state and the material of the planar work or product. Since the planar work or article is actually rendered three-dimensional by deformation of the base material, it is not necessary to physically flatten it. Under the guidance of the new technical thought, a high-precision and digital color three-dimensional model capable of truly restoring the current situation of a planar work or a product can be firstly obtained by means of the three-dimensional scanning technology and equipment which are already mature at present and the three-dimensional model+photographing and texture fusion technology which can serve the research needs of a museum, and then the real appearance of the planar work or the product in the creation or the manufacturing stage is rebuilt and restored by the related computer geometric technology, the deep learning technology and the three-dimensional curved surface flattening technology, so that the reliable guarantee is provided for truly collecting the image data of the planar work or the product.

Disclosure of Invention

The invention provides a non-physical and digital data acquisition method for planar works or products which actually show a three-dimensional state due to the deformation of a base material, which can fundamentally solve the problem of unavoidable information distortion caused by image data acquisition by relying solely on photographing or two-dimensional scanning technology, and meets the requirement of acquiring image data according to the original state of the works or products in the creation or manufacturing stage on the premise of not damaging or changing the material state of the works or products.

The whole steps include:

the planar work or the product is unfolded and fixed to the maximum extent by means of not changing the state of the substrate material and not damaging the substrate material;

acquiring a full-color high-reduction digital three-dimensional model of a planar work or product in a three-dimensional state due to deformation of a base material in a three-dimensional color scanning or three-dimensional scanning+photographing and texture fusion mode;

flattening the full-color high-reduction digital three-dimensional model of the planar work or product in a three-dimensional state due to deformation of the base material by using three-dimensional curved surface flattening software or other related software to enable the planar work or product to be in an absolute two-dimensional space representation state, namely an original representation state of the planar work or product in an creation or manufacturing stage;

if the original pattern is permanently covered, distorted or deformed due to deformation of the substrate, the original pattern can be reasonably corrected according to the characteristics of the substrate material, the logic of the pattern creation or design and other related information, so that the original pattern is more similar to the original expression of the planar work or product in the creation and manufacturing stages. Techniques in which application is required include:

(1) and calculating the Gaussian curvature of each discrete curved surface obtained by the color three-dimensional scanning. For a region with a corresponding gaussian curvature of 0, the method comprises:

a straight line surface piecewise approximation method;

positive mapping/equal area method;

a guide wire/ribbon approach;

based on the expansion of triangles in the discrete surface;

based on the expansion of vertices in the discrete surface.

(2) If the color three-dimensional scan has a region with Gaussian curvature other than 0, including a region which cannot be absolutely unfolded due to stretching, shrinkage, wrinkling or tearing generated during the storage of a planar work or product, correcting the region to restore the original appearance state to the maximum extent, including:

performing foreground segmentation (segment) and edge detection (edge detection) on the three-dimensional curved surface by using an image processing technology or deep learning to obtain an outer edge contour and an inner line of key contents on a planar work or product;

flattening the complex curved surface which cannot be unfolded absolutely into an absolute two-dimensional space, wherein the method comprises the following steps:

the boundary control method comprises flattening the outer edge contour and the inner line of the key content obtained by performing foreground segmentation (segmentation) and edge detection (edge detection) on the three-dimensional curved surface by using an image processing technology or deep learning to an absolute two-dimensional space without distortion and deformation;

the global optimization method based on geometric attribute constraint comprises the steps of requiring the length of lines or the angle between the lines which are expanded to an absolute two-dimensional space to be topologically equivalent to an original curved surface, and simultaneously requiring the consistency of geometric characteristics of the lines in a deployable curved surface area with the Gaussian curvature of 0 and the lines in a complex curved surface area with the Gaussian curvature of not 0 to be maintained after the lines are flattened to the absolute two-dimensional space, wherein the consistency of the lines is included;

if the internal areas of the outer edge contours and the internal lines of the key contents obtained by performing foreground segmentation (segmentation) and edge detection (edge detection) on the three-dimensional curved surface by using an image processing technology or deep learning are not flattened to an absolute two-dimensional state without distortion, a spring-particle model is adopted to minimize the unavoidable elastic deformation in the flattening process.

(3) And re-performing UV mapping on the flattened planar work or product in the absolute two-dimensional space according to the geometric corresponding relation between the planar work or product in the absolute two-dimensional space and the original three-dimensional curved surface, and mapping the texture information of the original planar work or product onto the planar work or product in the absolute two-dimensional space.

And marking the corrected pattern part, distinguishing the corrected pattern part from the pattern part which does not need correction after flattening, and respectively expressing the corrected pattern part and the pattern part by using different weighting functions to reflect different weights in future research, tracing and other processes. Meanwhile, the related information is stored together with the digital image information and the size information of the planar work or the product in an absolute two-dimensional space state.

If the full-color three-dimensional model of the planar work or product obtained by three-dimensional color scanning or three-dimensional scanning+photographing and texture mapping loses uniformity and relevance in terms of parameters of color brightness, hue and purity due to the influence of scanning or photographing light rays or environment and the like, so that each small area on the planar work or product obtained after flattening in absolute two-dimensional space lacks internal connection, the connection is very hard, so that the work or product lacks a feeling of one-step care in the whole, namely the integral charm characteristic of the work or product, the planar work or product obtained after flattening in absolute two-dimensional space needs to be subjected to integral fine adjustment, and organic fusion in terms of color brightness, hue and purity is realized on each curved surface after flattening of the three-dimensional model according to reasonable parameters, so that the original expression state of the work or product in the creation or manufacturing stage is restored to the greatest extent, and the method adopted comprises the following steps:

an image restoration method (deep image inpaiting) based on deep learning is characterized in that image features around an area needing restoration are utilized to train a deep neural network, and a local restoration patch is generated to be naturally fused with an overall image;

image transfer (style transfer) to automatically apply color, lines, art styles on one photograph to another photograph;

super resolution image generation (super resolution), which uses deep learning to build a relation between a low-definition photo and a high-definition photo of the same object, so as to help to clear the blurred photo;

creating a classifier (classifier) between the generated image and the actual photographed picture by using a generated countermeasure network (Generative Adversarial Network), wherein when the classifier cannot distinguish whether the image is generated or photographed, the generated image patch can be integrated with the subject picture;

an image smoothing method (image smoothing) for adjusting uneven light, color, hue, and reducing image noise, comprising:

an interpolation method, a linear smoothing method and a convolution method.

The base material includes: paper, wood, metal, plastic, rubber, leather, cloth, satin, silk, slate, tile, glass, ceramic tile, bone, and any other material that can be stretched to form a two-dimensional space and on which a planar work or article can be created or fabricated in a manner that depicts, writes, prints, topology, renders, and the like.

The three-dimensional scanning and three-dimensional color scanning method comprises the following steps:

(1) scanning the planar work or the product which is actually in a three-dimensional state by using a three-dimensional scanner;

(2) photographing the planar work or the product which is actually in a three-dimensional state by using a binocular depth camera, and generating a digital three-dimensional model by processing distance information;

(3) and continuously photographing or shooting a video of the planar work or the product which is actually in a stereoscopic state by using a common monocular camera, and reconstructing a digital three-dimensional model by using a photogrammetry algorithm according to the corresponding relation of the overlapped parts of the plurality of photos and the camera projection geometric relation of each photo.

The non-physical, digitized flattening may take other similar forms, including:

(1) by writing specific software, the full-color high-reduction digital three-dimensional model of the planar work or product in a three-dimensional state due to deformation of the base material is obtained, and meanwhile, the high-reduction digital image and the size information of the planar work or product in an absolute two-dimensional space state are directly calculated and obtained;

(2) the intelligent camera is implanted with software or connected with related software through a computer or a network to realize the purpose of directly calculating and obtaining the high-reduction digital image and the size information of the planar work or the product in an absolute two-dimensional space state in the photographing process.

Drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. In the drawings:

fig. 1 shows a schematic diagram of the principle of non-physical, digital flattening of a planar work or article of the invention, which has been in a three-dimensional state in practice due to deformation of the substrate, to obtain image data and dimensions in its absolute two-dimensional state and to store it in an information repository.

Detailed Description

When a planar work or article is considered to be worth collecting, displaying, studying, it is preferable to digitally collect and protect the information. If it is found that the planar work or the product is actually in a three-dimensional state due to deformation of the base material, and the planar work or the product cannot be flattened under the premise of natural expansion or with the help of harmless tools such as a clamp, paperweight and the like, the non-physical digital flattening method provided by the invention is adopted first, and then data are collected and stored, wherein the specific steps include:

the planar work or the product is unfolded and fixed to the maximum extent by means of not changing the state of the substrate material and not damaging the substrate material, and the fixing method comprises clamping, paperweight pressing and the like;

and collecting the full-color high-reduction digital three-dimensional model of the actual three-dimensional planar work or product. The full-color high-reduction-degree digital three-dimensional model can be directly acquired by three-dimensional scanning equipment, or the high-precision digital three-dimensional model can be acquired first, then a plurality of high-precision color photos are shot in a multi-angle and all-dimensional mode by using a camera, and the photos are attached to the high-precision three-dimensional model by using texture fusion software to form the full-color high-reduction-degree digital three-dimensional model;

by the absolute two-dimensional image generation module provided by the invention, the full-color high-reduction digital three-dimensional model is flattened by means of software based on a geometric algorithm, curved surface flattening and deep learning development, so that the image information and the size of a planar work or product in an absolute two-dimensional space state are obtained;

if the original pattern is permanently changed due to deformation of the substrate, the original pattern can be reasonably corrected according to the characteristics of the substrate material, the logic of the pattern creation or design and other related information, so that the original pattern is more similar to the original expression of a planar work or product in the creation and manufacturing stages. Techniques in which application is required include:

deep learning, namely obtaining key points, external contours and internal lines on a three-dimensional model;

computer geometry, ensuring that the geometric features on the three-dimensional model are consistent with the feature topology of the two-dimensional image;

an image analysis algorithm, which obtains surface texture features;

and (3) expanding the complex curved surface, and controlling the distortion deformation of the three-dimensional model in the process of expanding the three-dimensional model to the two-dimensional plane to the minimum extent.

The specific implementation scheme comprises the following steps:

if in the three-dimensional scanning process, the folds of the surface of a planar work or product are shielded, so that errors occur in the numerical calculation process of using a 3d mesh (triangular mesh) to represent the curved surface, and the three-dimensional model curved surface is possibly not 100% Gaussian, the Gaussian curvature of each part of the three-dimensional model curved surface can be calculated first, and the curved surface with the Gaussian curvature of 0 and the surface texture thereof are flattened into the same two-dimensional plane;

for very few complex surfaces that are not absolutely flattened, other methods of flattening may be employed, including:

controlling the unfolding process through the boundary and the grid lines, and ensuring that the geometric relationship of the boundary and the grid lines in the three-dimensional curved surface is faithfully reserved in the unfolded two-dimensional plane;

the physics-based energy model, particularly the spring-mass point model, allows for proper stretching or scaling of the critical boundaries and regions inside the grid lines during deployment;

in view of the fact that key points of patterns of planar works or products are concentrated on outlines of words or pictures, the outlines and the key points can be found in a complex curved surface through an image processing or deep learning method, and the geometric distance of the outlines and the key points is ensured to be unchanged in an unfolded two-dimensional plane. While for filled areas of non-textured features in the contour, appropriate geometric warping is allowed;

and marking the corrected pattern part, distinguishing the corrected pattern part from the pattern part which does not need correction after flattening, and respectively expressing the corrected pattern part and the pattern part by using different weighting functions to reflect different weights in future research, tracing and other processes. Meanwhile, the related information is stored together with the digital image information and the size information of the planar work or the product in an absolute two-dimensional space state.

If the full-color three-dimensional model of the planar work or product obtained by three-dimensional color scanning or three-dimensional scanning+photographing and texture mapping loses uniformity and relevance in terms of parameters of color brightness, hue and purity due to the influence of scanning or photographing light rays or environment and the like, so that each small area on the planar work or product obtained after flattening in absolute two-dimensional space lacks internal connection, the connection is very hard, so that the work or product lacks a feeling of one-step care in the whole, namely the integral charm characteristic of the work or product, the planar work or product obtained after flattening in absolute two-dimensional space needs to be subjected to integral fine adjustment, and organic fusion in terms of color brightness, hue and purity is realized on each curved surface after flattening of the three-dimensional model according to reasonable parameters, so that the original expression state of the work or product in the creation or manufacturing stage is restored to the greatest extent, and the method adopted comprises the following steps:

an image restoration method (deep image inpaiting) based on deep learning is characterized in that image features around an area needing restoration are utilized to train a deep neural network, and a local restoration patch is generated to be naturally fused with an overall image;

image transfer (style transfer) to automatically apply color, lines, art styles on one photograph to another photograph;

super resolution image generation (super resolution), which uses deep learning to build a relation between a low-definition photo and a high-definition photo of the same object, so as to help to clear the blurred photo;

creating a classifier (classifier) between the generated image and the actual photographed picture by using a generated countermeasure network (Generative Adversarial Network), wherein when the classifier cannot distinguish whether the image is generated or photographed, the generated image patch can be integrated with the subject picture;

an image smoothing method (image smoothing) for adjusting uneven light, color, hue, and reducing image noise, comprising:

an interpolation method, a linear smoothing method and a convolution method;

storing the obtained image information, size information and other auxiliary information related to the image information and the size information into a related data information base;

and displaying the image information in the information base through a computer vision technology. The user can zoom, rotate, and move the image by touching the screen, mouse, or keyboard with a finger.

In summary, the method for collecting the digitized data of the planar work or the product which is actually in a three-dimensional state due to the deformation of the substrate material can not harm the planar work or the product, can restore the original performance of the planar work or the product in the creation or the manufacturing stage with high precision, and is a stable, effective and reliable method for collecting the digitized data of the planar work or the product.

The data acquisition and digital protection method of the planar work or product applying the inventive concept of the present invention can be implemented in various ways, 2 of which are specifically described below:

example 1

S1, a precious cultural relic is newly collected in a museum, and the novel cultural relic is a poster drawn on a large iron sheet by a red army fighter in a long period, and although the iron sheet is bent and the crease is tired, the poster is clearly visible. In order to better promote the peace spirit of the march, the museum decides to digitally rescue and protect the cultural relics, digitally copy the poster and display it on line. At the same time, the poster from which the data was collected may also be used by researchers to conduct further research. However, since the iron sheet is subject to the years of deterioration, the texture is already very fragile, and forced to level it and take a picture or scan it in a plane, it is certainly possible to cause great damage to it. Moreover, in order to strike flat the recessed and raised portions of the poster, the poster must be damaged to some extent. Because of the limitation, the traditional photographing and plane scanning cannot certainly finish the tasks of digital rescuing and protecting, so that the image data acquisition of the cultural relics can be carried out only by adopting the non-physical and digital flattening technology without changing the material state of the article;

s2, naturally expanding and fixing the iron sheet to the greatest extent by a means of not changing the material state of the iron sheet and not damaging the material of the iron sheet;

s3, acquiring a full-color high-reduction digital three-dimensional model of an iron sheet which is actually in a three-dimensional state due to material deformation and a poster attached to the iron sheet by using a three-dimensional color scanning or three-dimensional scanning+photographing and texture fusion mode;

s4, flattening a full-color high-reduction digital three-dimensional model of the poster in a three-dimensional state due to the deformation of the iron sheet by using three-dimensional curved surface flattening software or other related software, so that the poster is in an absolute two-dimensional space representation state, namely an original representation state in a poster manufacturing stage;

due to the long-term and severe preservation environment, the iron sheet is covered in a folding way at a local fine part, wherein the folding way cannot flatten 100% of curved surfaces. In this regard, researchers apply the correction program provided by the invention to intelligently correct the correction program, so that the digital restoration with high fidelity is realized to the greatest extent. At the same time, the corrected region is marked by the researcher to distinguish it from the uncorrected region. In future research, professionals can have a deeper understanding of the true appearance of the poster;

s5, storing the digital image, the size information and other relevant information of the poster in the absolute two-dimensional space restored by the technology of the invention into a database of the museum.

Example 2

S1, a museum newly receives a bright-end and bright-end large brocade found in a minority area, and an expert considers that the large brocade has protection and collection values and can be identified as a secondary cultural relic. In order to better protect the ethnic culture of the minority and better publicize the contribution of each minority to the whole ethnic culture of the Chinese, the museum decides to digitally rescue and protect the brocade, digitally copies the brocade and displays the brocade on line. Meanwhile, the brocade with the acquired data is further researched by researchers. However, because brocade is fragile, it may be injured by ironing it with a tool such as an iron and shooting or scanning it. Moreover, due to the special ductility of the brocade, the brocade is hung on a wall to shoot, so that the brocade is not only stretched and uneven, but also the deformation of the brocade can be increased when the brocade is hung; even if the brocade is placed under the glass plate for shooting or scanning, the soft brocade is deformed greatly due to the weight of the glass. Finally, the method comprises the following steps. The museum decides to acquire image data by adopting the digital flattening technology without changing the brocade material state;

s2, naturally unfolding the brocade to the greatest extent and fixing the brocade by a means of not changing the material state of the brocade and not damaging the material of the brocade;

s3, acquiring a full-color high-reduction digital three-dimensional model of the brocade in a three-dimensional state, which is deformed due to natural arrangement, by using a three-dimensional color scanning or three-dimensional scanning+photographing and texture fusion mode;

s4, flattening the full-color high-reduction digital three-dimensional model of the brocade which is actually in a three-dimensional state due to deformation by using three-dimensional curved surface flattening software or other related software, so that the model is in an absolute two-dimensional space representation state, namely an original representation state in a brocade manufacturing stage;

s5, storing the digitized image of the brocade in the absolute two-dimensional space restored by the technology of the invention into a database of the museum.

It should be noted that:

the required structure for the construction of such devices is apparent from the description above. In addition, the present invention is not directed to any particular programming language. It will be appreciated that the teachings of the present invention described herein may be implemented in a variety of programming languages, and the above description of specific languages is provided for disclosure of enablement and best mode of the present invention.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

Similarly, it should be appreciated that in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be construed as reflecting the intention that: i.e., the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.

Those skilled in the art will appreciate that modules in the apparatus of the embodiments may be adaptively changed and disposed in one or more apparatuses different from the embodiments. The modules or units or components of the embodiments may be combined into one module or unit or component and, furthermore, they may be divided into a plurality of sub-modules or sub-units or sub-components. Any combination of all features disclosed in this specification, and all processes or units of any method or apparatus so disclosed, may be employed, except that at least some of such features and/or processes or units are mutually exclusive. Each feature disclosed in this specification may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features but not others included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.

Various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that some or all of the functions of some or all of the components in the creation means of a virtual machine according to an embodiment of the present invention may be implemented in practice using a microprocessor or Digital Signal Processor (DSP). The present invention can also be implemented as an apparatus or device program (e.g., a computer program and a computer program product) for performing a portion or all of the methods described herein. Such a program embodying the present invention may be stored on a computer readable medium, or may have the form of one or more signals. Such signals may be downloaded from an internet website, provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware.

The present invention is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present invention are intended to be included in the scope of the present invention.

Claims (5)

1. A method of image data acquisition of a planar work or article attached to a deformed substrate, comprising:

the method for acquiring information data of a planar work or product which is actually in a three-dimensional state due to deformation of a base material by using an absolute two-dimensional space standard, digitally recording the image and size information of the planar work or product with high reduction degree comprises the following steps:

the planar work or the product is unfolded and fixed to the maximum extent by means of not changing the state of the substrate material and not damaging the substrate material;

acquiring a full-color high-reduction digital three-dimensional model of the planar work or the product which is actually in a three-dimensional state due to the deformation of the substrate material in a three-dimensional color scanning or three-dimensional scanning+photographing and texture fusion mode;

flattening the full-color high-reduction digital three-dimensional model of the planar work or product in a three-dimensional state due to deformation of the base material by using three-dimensional curved surface flattening software or other related software to enable the planar work or product to be in an absolute two-dimensional space representation state, namely an original representation state of the planar work or product in an creation or manufacturing stage, wherein the method comprises the following steps:

if the folds of the surface of a planar work or product are shielded in the three-dimensional scanning process, so that errors occur in the numerical calculation process of using a 3dmesh triangular grid to represent a curved surface, and the three-dimensional model curved surface cannot realize 100% Gaussian expansion, firstly calculating the Gaussian curvature of each part of the three-dimensional model curved surface, and flattening the curved surface with the Gaussian curvature of 0 and the surface texture of the curved surface into the same two-dimensional plane;

for very few complex curved surfaces which cannot be flattened absolutely, flattening is performed by other methods, including:

controlling the unfolding process through the boundary and the grid lines, and ensuring that the geometric relationship of the boundary and the grid lines in the three-dimensional curved surface is faithfully reserved in the unfolded two-dimensional plane;

based on the physical energy model, including the spring-mass point model, the region inside the critical boundaries and grid lines allows for proper stretching or scaling during deployment;

according to the characteristic that key points of patterns of planar works or products are concentrated on outlines of words or pictures, the outlines and the key points are found in a complex curved surface through an image processing or deep learning method, and the geometric distance of the key points is ensured to be unchanged in an unfolded two-dimensional plane; while for filled areas of non-textured features in the contour, appropriate geometric warping is allowed;

and storing the digitized image with high reduction degree of the planar work or the product in the absolute two-dimensional space state after flattening and the size information for later use.

2. The method of claim 1, wherein the three-dimensional scanning and three-dimensional color scanning methods comprise:

(1) Scanning the planar work or the product which is actually in a three-dimensional state by using a three-dimensional scanner;

(2) Photographing the planar work or the product which is actually in a three-dimensional state by using a binocular depth camera, and generating a digital three-dimensional model by processing distance information;

(3) And continuously photographing or shooting a video of the planar work or the product which is actually in a stereoscopic state by using a common monocular camera, and reconstructing a digital three-dimensional model by using a photogrammetry algorithm according to the corresponding relation of the overlapped parts of the plurality of photos and the camera projection geometric relation of each photo.

3. The method according to claim 1, characterized in that it comprises:

by writing specific software, the full-color high-reduction digital three-dimensional model of the planar work or product in a three-dimensional state due to the deformation of the base material is obtained, and meanwhile, the high-reduction digital image and the size information of the planar work or product in an absolute two-dimensional space state are directly calculated and obtained;

relevant software is implanted in the intelligent camera or the camera is connected with the relevant software through a computer or a network, so that the purposes of directly calculating and obtaining the high-reduction digital image and the size information of the planar work or the product in an absolute two-dimensional space state in the photographing process are realized.

4. The method of claim 1, wherein the base material comprises paper, board, sheet metal, plastic, rubber, leather, cloth, satin, silk, slate, tile, glass, ceramic tile, bone chips, and any other material that can be spread to form a two-dimensional space and onto which a planar work or article can be created or fabricated in a manner that is depicted, written, printed, developed, rendered, and the like.

5. The method of claim 1, wherein for permanent changes to the original pattern due to deformation of the substrate, reasonable corrections are made to the substrate material based on its characteristics, logic of the pattern creation or design, and other relevant information to make it more nearly the original representation of the planar work or article at the creation and production stages, wherein the techniques that need to be applied include:

deep learning, namely obtaining key points, external contours and internal lines on a three-dimensional model;

computer geometry, ensuring that the geometric features on the three-dimensional model are consistent with the feature topology of the two-dimensional image;

an image analysis algorithm, which obtains surface texture features;

and (3) expanding the complex curved surface, and controlling the distortion deformation of the three-dimensional model in the process of expanding the three-dimensional model to the two-dimensional plane to the minimum extent.