CN114998527A - High-accuracy three-dimensional human body surface reconstruction system - Google Patents
High-accuracy three-dimensional human body surface reconstruction system Download PDFInfo
- Publication number
- CN114998527A CN114998527A CN202210736632.9A CN202210736632A CN114998527A CN 114998527 A CN114998527 A CN 114998527A CN 202210736632 A CN202210736632 A CN 202210736632A CN 114998527 A CN114998527 A CN 114998527A
- Authority
- CN
- China
- Prior art keywords
- model
- module
- human body
- data
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000010276 construction Methods 0.000 claims abstract description 50
- 238000000926 separation method Methods 0.000 claims abstract description 16
- 238000013507 mapping Methods 0.000 claims abstract description 14
- 238000005457 optimization Methods 0.000 claims abstract description 14
- 241000544061 Cuculus canorus Species 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims description 74
- 235000005770 birds nest Nutrition 0.000 claims description 51
- 235000005765 wild carrot Nutrition 0.000 claims description 51
- 238000000605 extraction Methods 0.000 claims description 31
- 238000007781 pre-processing Methods 0.000 claims description 20
- 238000000034 method Methods 0.000 claims description 11
- 238000013479 data entry Methods 0.000 claims description 7
- 239000000463 material Substances 0.000 claims description 5
- 241000288673 Chiroptera Species 0.000 claims description 3
- 108091026890 Coding region Proteins 0.000 claims description 3
- 230000002457 bidirectional effect Effects 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000009827 uniform distribution Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 238000013075 data extraction Methods 0.000 abstract description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004381 surface treatment Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/04—Texture mapping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Library & Information Science (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Computer Graphics (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a high-accuracy three-dimensional human body surface reconstruction system in the technical field of computer vision, which obtains human body image related data information through direct scanning, carries out denoising, feature point data extraction and feature data separation after classification, further directly provides human body data features through pictures, can automatically match a proper model according to the prior human body related feature model, does not need to carry out model reconstruction again, further can effectively reduce reconstruction calculated amount, can save time, simultaneously carries out texture mapping construction by combining the human body surface features, thereby completing rich construction of model surface details, fully utilizes the advantages of a bat algorithm and a cuckoo optimization algorithm by adopting a mixed optimization algorithm combining the bat algorithm and the cuckoo optimization algorithm, obtains a human body database query plan which is better than the traditional algorithm, the needed human body model can be quickly inquired in the human body database.
Description
Technical Field
The invention relates to the technical field of computer vision, in particular to a high-accuracy three-dimensional human body surface reconstruction system.
Background
In the field of computer vision, three-dimensional surface reconstruction of a target object is a very classic problem in the field of computer vision, and at present, there are many types of three-dimensional model construction methods for a human body, for example, a method of using a memory-efficient multi-layer GPU data structure to reconstruct a three-dimensional model of a human body, or a KinectFusion method using a movable volume to reconstruct a three-dimensional model of a human body in the case of an outdoor large scene, or three-dimensional surface reconstruction of a target object by using two-dimensional images at a plurality of different angles, or reconstruction in a laser three-dimensional scanning manner, and the like.
In the process of reconstructing the human body model by adopting the reconstruction methods, more devices are generally required to be used for matching, and a special laser three-dimensional scanning instrument is required by a laser three-dimensional scanning method, but the instrument is very expensive, is difficult to bear by a personal user, is not suitable for common people, has large reconstruction calculation amount, wastes time and is difficult to rapidly and accurately reconstruct the surface of the three-dimensional human body.
Based on the above, the invention designs a three-dimensional human body surface reconstruction system with high accuracy to solve the above problems.
Disclosure of Invention
The invention aims to provide a three-dimensional human body surface reconstruction system with high accuracy so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: a three-dimensional human body surface reconstruction system with high accuracy comprises an information acquisition unit, wherein the information acquisition unit acquires and records acquired three-dimensional human body data information into a preprocessing module for primary processing, the preprocessing module is used for carrying out denoising preprocessing on surface characteristic data to obtain depth information without noise smoothness, the preprocessing module transfers the processed data to a characteristic point extraction module, the characteristic point extraction module is used for screening and extracting main characteristic points of the acquired human body data information, the characteristic point extraction module sends the extracted human body characteristic point data to a data separation module, the data separation module is used for dispersing the extracted and classified human body characteristic data information, the data separation module inputs the separated data to a central processing unit, the central processing unit carries out data integration, and the model query extraction module can call the extracted data to the central processing unit to obtain data, the human body model building module can obtain model data from the model inquiring and extracting module, the human body model building module is used for combining models built according to human body characteristic data information, the model inquiring and extracting module is used for searching and searching models built by human bodies in a human body database, the model inquiring and extracting module and the human body database can carry out bidirectional information obtaining, the human body database is used for storing human body model data information, the human body database is connected with the information updating module to update the human body data in real time, the central processing unit is connected with the head model rebuilding module, the head model rebuilding module is used for modeling classified human body head data information, and the head model rebuilding module is connected with the human body model building module to carry out information obtaining.
The human body model building module is connected with the connection processing module to form data intercommunication, the connection processing module is used for performing connection processing on the human body model, and the connection processing module is sequentially distributed and connected with the surface processing module, the feature adding module, the model map processing module and the model display module to perform display building on the human body feature model;
the surface processing module is used for carrying out surface finishing processing on the reconstructed human body model, and the characteristic adding module is used for adding main characteristic points contained in a human body; the model map processing module is used for re-projecting the obtained human body model on each image and carrying out calibration on the position and the direction, and the model display module is used for displaying and viewing the reconstructed model in an image mode.
The central processing unit is respectively connected with the first model building module, the second model building module and the third model building module and is used for building models for classifying data information of all parts of a human body; the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module are respectively connected with the information updating module, and model data are obtained from the human body database through the information updating module.
Preferably, the model query extraction module adopts a hybrid optimization algorithm combining a bat algorithm and a cuckoo optimization algorithm, and the specific steps are as follows:
s1, obtaining a coding sequence L by adopting a subsequent traversal of a junction tree, initializing each parameter of a hybrid optimization algorithm, calculating the fitness value of a bird nest according to an objective function, determining the current optimal position and optimal value of the bird nest, reserving the position of the bird nest with the previous generation of the optimal position, updating the position of the bird nest by utilizing an algorithm of a Levy Flight mechanism to obtain a new position of the bird nest, and performing border-crossing processing;
the Levy Flight mechanism comprises the following algorithms:
in the formula (I), the compound is shown in the specification,andrespectively representing the position vectors of the ith bird nest in the t generation and the t +1 generation, alpha is a step length adjusting factor and can take different values according to different conditions,for point-to-point multiplication, Le' vy (lambda) is a random search path;
s2, calculating the fitness value of the new bird nest position by using the objective function, comparing the fitness value with the previous generation, covering the fitness value of the previous generation when the fitness value is superior to the fitness value of the previous generation, updating the corresponding bird nest position, determining the current optimal bird nest position and the optimal value, and using the fitness value of the foreign eggProbability of discovery p α And random number R epsilon [0,1 ] subjected to uniform distribution]Making a comparison, e.g. R > p α Randomly changing the positions of the bird nests, thereby obtaining a group of new bird nest positions;
s3, taking the new bird nest position obtained in the previous step as an initial point of a bat algorithm, updating the position of the bird nest by using the bat algorithm to obtain a group of new bird nest positions, then evaluating the fitness value of the bird nest positions, determining the current optimal bird nest position and the optimal value after comparison, and when a termination condition is reached, inquiring a plan of a database corresponding to the optimal bird nest position;
wherein, the bat algorithm comprises the following steps:
initial population number NP, maximum pulse volume A 0 Maximum pulse rate R 0 Search pulse frequency range [ f min ,f max ]Attenuation coefficient alpha of sound volume, enhancement coefficient gamma of search frequency, search precision epsilon or maximum iteration number iter _ max, and then randomly initializing the bat position x i And according to the quality of the fitness value, searching the current optimal solution x * In the evolution process of the population, the search pulse frequency, the speed and the position of each generation of individuals are changed according to the following formula:
f i =f min +(f max -f min )×β
in the formula, beta is belonged to [0,1 ]]Is a uniformly distributed random number, f i Is the search pulse frequency of bat i,respectively representing the velocity of the bat i at times t and t-1,respectively representing the position, x, of the bat i at times t and t-1 * Representing the optimal solution of all current bats, and then generating a uniformly distributed random number rand if rand > r i Randomly disturbing the current optimal solution to generate a new solution, and performing border crossing treatment on the new solution if rand is less than A i And f (x) i )<f(x * ) Then receive the new solution generated in the previous step and then pair r according to the following formula i And A i Updating:
and then outputting the global optimal value and the optimal solution.
Preferably, the information acquisition unit comprises a scanning acquisition module and a data entry module, and the scanning acquisition module and the data entry module can respectively perform information entry and data direct entry of the three-dimensional scanning device.
Preferably, the scanning acquisition module scans a human body image by using an infrared camera to obtain image data with clear outline; the data entry module is used for entering the human body related data information.
Preferably, the human body database includes first model data, second model data, and third model data, and is configured to perform classified storage and management on multiple pieces of model data information of a human body, where data stored in the first model data, the second model data, and the third model data correspond to data to be constructed by the first model construction module, the second model construction module, and the third model construction module, respectively.
Preferably, the model mapping processing module is configured to re-project the obtained human body model onto each image, perform calibration in position and direction, and then project color and texture information in the image onto a triangulated vertex of the model by using a texture mapping function in OpenGL to complete texture mapping and material restoration, so as to obtain a surface-processed three-dimensional model.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the invention, the human body image related data information is obtained through direct scanning, denoising and feature point data extraction are carried out, and the classified feature data are separated, so that the human body data features can be directly provided through a photo, a proper model can be automatically matched according to the conventional human body related feature model, model reconstruction is not required again, the reconstruction calculated amount can be effectively reduced, time can be saved, meanwhile, texture mapping construction is carried out by combining with the human body surface features, so that the abundant construction of the model surface details is completed, and the reconstruction of a three-dimensional human body can be rapidly and accurately realized;
2. the invention obtains the relevant data information of the human body image through the information acquisition unit, then carries out denoising processing through the preprocessing module to obtain smooth data information, carries out human body characteristic data information disassembly and classification through the characteristic point extraction module, simultaneously separates the relevant data after each classification, then carries out reconstruction of the head of the human body through the head reconstruction module, then matches the corresponding first model data, second model data and third model data according to the first model construction module, the second model construction module and the third model construction module, can quickly obtain a reconstructed model, can reconstruct any one or more items through the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module, and can selectively match and be suitable for a historical model, thereby leading the model to have reasonable selectivity and diversity processing, the method is simple and easy to implement, and the operation speed is increased;
3. according to the invention, after the human body model is reconstructed through the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module, the reconstructed characteristic data of each part of the human body model can be directly sent to the information updating module, and the reconstructed model information can be sent to the human body database for storage through the information updating module, so that new human body model data can be obtained, further, the real-time update of the human body database can be completed in the process of reconstructing the model, and the diversification of the human body model is increased.
The bat algorithm and the cuckoo optimization algorithm are combined to form a mixed optimization algorithm, so that the advantages of the bat algorithm and the cuckoo algorithm can be fully utilized, a better human body database query plan than the traditional algorithm is obtained, a needed human body model can be rapidly queried in the human body database, and a user can conveniently and rapidly build the model.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic diagram of the system of the present invention;
FIG. 2 is a schematic flow chart of example 1 of the present invention;
FIG. 3 is a schematic flow chart of example 2 of the present invention;
fig. 4 is a schematic flow structure diagram according to embodiment 3 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-4, the present invention provides a technical solution: a three-dimensional human body surface reconstruction system with high accuracy comprises an information acquisition unit, wherein the information acquisition unit acquires and records acquired three-dimensional human body data information into a preprocessing module for primary processing, the information acquisition unit comprises a scanning acquisition module and a data recording module, and the scanning acquisition module and the data recording module can respectively perform information recording and data direct recording of a three-dimensional scanning device; the scanning acquisition module scans a human body image by adopting an infrared camera to obtain image data with clear outline; the data entry module is used for entering the human body related data information.
The preprocessing module is used for carrying out denoising preprocessing on surface characteristic data to obtain depth information without noise smoothness, the preprocessing module transfers the processed data to the characteristic point extraction module, the characteristic point extraction module is used for screening and extracting main characteristic points of the obtained human body data information, the characteristic point extraction module sends the human body characteristic point extraction data to the data separation module, the data separation module is used for dispersing the extracted and classified human body characteristic data information, the data separation module inputs the separated data to the central processing unit and carries out data integration by the central processing unit, the model query extraction module can call the extraction data to the central processing unit to extract the data, the human body model construction module can obtain model data from the model query extraction module and is used for combining models constructed according to the human body characteristic data information, the model query extraction module is used for searching and searching models which are constructed by human bodies in a human body database, the model query extraction module and the human body database can carry out bidirectional information acquisition, the human body database is used for storing human body model data information, the human body database comprises first model data, second model data and third model data and is used for respectively carrying out classified storage management on a plurality of pieces of model data information of the human bodies, and the data stored in the first model data, the second model data and the third model data respectively correspond to the data required to be constructed by the first model construction module, the second model construction module and the third model construction module.
The human body database is connected with the information updating module to update human body data in real time, the central processing unit is connected with the head model rebuilding module, the head model rebuilding module is used for modeling classified human body head data information, and the head model rebuilding module is connected with the human body model building module to acquire information.
The human body model building module is connected with the connection processing module to form data intercommunication, the connection processing module is used for performing connection processing on the human body model, and the connection processing module is sequentially distributed and connected with the surface processing module, the feature adding module, the model map processing module and the model display module to perform display building on the human body feature model;
the surface treatment module is used for carrying out surface finishing treatment on the reconstructed human body model, and the characteristic adding module is used for adding main characteristic points contained in the human body; the model map processing module is used for re-projecting the obtained human body model on each image and carrying out calibration on the position and the direction, and the model display module is used for displaying and viewing the reconstructed model in an image mode.
The model mapping processing module is used for re-projecting the obtained human body model onto each image, calibrating the position and the direction, and then projecting color and texture information in the image onto a model triangularization vertex by using a texture mapping function in OpenGL to complete texture mapping and material restoration so as to obtain a three-dimensional model after surface processing.
The central processing unit is respectively connected with the first model building module, the second model building module and the third model building module and is used for building models for classifying data information of all parts of a human body; the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module are respectively connected with the information updating module, and model data are obtained from the human body database through the information updating module.
The model query extraction module adopts a mixed optimization algorithm combining a bat algorithm and a cuckoo optimization algorithm, and comprises the following specific steps:
s1, obtaining a coding sequence L by adopting a subsequent traversal connecting tree, initializing each parameter of a hybrid optimization algorithm, calculating the adaptability value of the bird nest according to an objective function, determining the current optimal position and optimal value of the bird nest, reserving the position of the bird nest with the previous generation optimal position, updating the position of the bird nest by utilizing an algorithm of a Levy Flight mechanism to obtain a new position of the bird nest, and performing border crossing processing;
the Levy Flight mechanism has the algorithm as follows:
in the formula (I), the compound is shown in the specification,andrespectively representing the position vectors of the ith bird nest in the t generation and the t +1 generation, alpha is a step length adjusting factor and can take different values according to different conditions,for point-to-point multiplication, Le' vy (λ) is a random search path;
s2, calculating the fitness value of the new bird nest position by using the objective function, comparing the fitness value with the previous generation, covering the fitness value of the previous generation when the fitness value is superior to the fitness value of the previous generation, updating the corresponding bird nest position, determining the current optimal bird nest position and the optimal value, and then using the discovery probability p of the foreign egg α And random number R epsilon [0,1 ] subjected to uniform distribution]Making a comparison, e.g. R > p α Randomly changing the positions of the bird nests to obtain a group of new bird nest positions;
s3, taking the new bird nest position obtained in the previous step as an initial point of a bat algorithm, updating the position of the bird nest by using the bat algorithm to obtain a group of new bird nest positions, then evaluating the fitness value of the bird nest positions, determining the current optimal bird nest position and the optimal value after comparison, and when a termination condition is reached, inquiring a plan of a database corresponding to the optimal bird nest position;
wherein, the bat algorithm comprises the following steps:
initial population number of individuals NP, maximum pulse volume A 0 Maximum pulse rate R 0 Search pulse frequency range [ f min ,f max ]Attenuation coefficient alpha of sound volume, enhancement coefficient gamma of search frequency, search precision epsilon or maximum iteration number iter _ max, and then randomly initializing the bat position x i And according to the quality of the fitness value, searching the current optimal solution x * In the evolution process of the population, the search pulse frequency, the speed and the position of each generation of individuals are changed according to the following formula:
f i =f min +(f max -f min )×β
wherein beta is epsilon [0,1 ]]Is a uniformly distributed random number, f i Is the search pulse frequency of bat i,respectively representing the velocity of the bat i at times t and t-1,respectively representing the position, x, of the bat i at times t and t-1 * Representing the optimal solution of all current bats, and then generating a uniformly distributed random number rand if rand > r i Randomly disturbing the current optimal solution to generate a new solution, and performing border crossing treatment on the new solution if rand is less than A i And f (x) i )<f(x * ) Then receive the new solution generated in the previous step and then pair r according to the following formula i And A i Updating:
and then outputting the global optimal value and the optimal solution.
The three-dimensional human body surface reconstruction system improves the accuracy and adopts the steps as follows: directly acquiring relevant data information of a human body image through an information acquisition unit, transmitting the relevant data information to a preprocessing module, performing denoising processing through the preprocessing module, extracting the data information of the human body characteristic points through a characteristic point extraction module, separating the classified characteristic data through a data separation module, transmitting the data through a central processing unit, directly transmitting the data to a model query extraction module, directly matching a corresponding human body model from a human body database through the model query extraction module, transmitting the data to a human body model construction module for direct construction, performing model reconstruction through a connection processing module, performing surface processing on the model through a surface processing module, adding relevant characteristics of the human body through a characteristic adding module, and finally performing rich processing on model information mapping through a model mapping processing module, displaying the model through a model display module;
the technical effects achieved by adopting the steps are as follows: acquiring human body image related data information through direct scanning, denoising, extracting feature point data, and separating classified feature data, so that human body data features can be directly provided through a photo, a proper model can be automatically matched according to a conventional human body related feature model, model reconstruction is not required again, reconstruction calculation amount can be effectively reduced, time can be saved, texture mapping construction is performed by combining human body surface features, abundant construction of model surface details is completed, multi-part reconstruction of the model is not required, and three-dimensional human body reconstruction can be rapidly and accurately realized;
the steps of selectively matching the corresponding stage models are: firstly, the information acquisition unit directly acquires the relevant data information of the human body image, then the relevant data information is transmitted to the preprocessing module, then the denoising processing is carried out through the preprocessing module, the characteristic point data information of the human body is extracted through the characteristic point extraction module, then the classified characteristic data is separated through the data separation module and transmitted through the central processing unit, then the head model is reconstructed through the head model reconstruction module, the model can be constructed through the first model construction module, the second model construction module and the third model construction module, meanwhile, the model query extraction module can be used for matching the model in the corresponding human body database, then the model is directly transmitted to the human body model construction module for construction processing, the model reconstruction is carried out through the connection processing module, and then the surface processing module is used for carrying out the surface processing on the model, and the characteristic adding module is used for adding relevant characteristics of the human body, and finally the model information map is subjected to rich processing through the model map processing module, and then the model display module is used for displaying the model.
The technical effect achieved by adopting the steps is as follows: acquiring relevant data information of a human body image through an information acquisition unit, performing denoising processing through a preprocessing module to obtain smooth data information, performing human body characteristic data information disassembly and classification through a characteristic point extraction module, separating the classified relevant data, reconstructing the head of the human body through a head reconstruction module, matching corresponding first model data, second model data and third model data according to a first model construction module, a second model construction module and a third model construction module, rapidly obtaining a reconstructed model, reconstructing any one or more than one item through the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module, and selectively matching and adapting to a historical model, so that the model has reasonable selectivity and diversity processing, simple and easy to operate, and the operation speed is improved.
The method comprises the following steps of directly acquiring relevant data information of a human body image through an information acquisition unit, transmitting the relevant data information to a preprocessing module, performing denoising processing through the preprocessing module, extracting data information of human body feature points through a feature point extraction module, separating classified feature data through a data separation module, reconstructing a head according to head features to obtain a separation model, and performing construction connection processing on the separation model.
The second embodiment is different from the first embodiment in that the reconstruction processing is performed on the head according to the characteristics of the head, and meanwhile, the reconstruction processing is performed on the other three parts of data according to the characteristics, and then the construction connection processing is performed on the separation model.
The third embodiment is different from the second embodiment in that it is no longer necessary to determine whether there is human body model data with corresponding features in the human body database, and after the head is reconstructed according to the head features and the other three parts of data are respectively reconstructed according to the features, the reconstructed human body information is uploaded to the database.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims (6)
1. A three-dimensional human body surface reconstruction system with high accuracy comprises an information acquisition unit, wherein the information acquisition unit acquires and records acquired three-dimensional human body data information into a preprocessing module for primary processing, the preprocessing module is used for carrying out denoising preprocessing on surface characteristic data to obtain depth information without noise smoothness, the preprocessing module transfers the processed data to a characteristic point extraction module, the characteristic point extraction module is used for screening and extracting main characteristic points of the acquired human body data information, the characteristic point extraction module sends the extracted human body characteristic point data to a data separation module, the data separation module is used for dispersing the extracted and classified human body characteristic data information, the data separation module inputs the separated data to a central processing unit, the central processing unit carries out data integration, and the model query extraction module can call the extracted data to the central processing unit to obtain data, the human body model building module can obtain model data from the model query extraction module, the human body model building module is used for combining models built according to human body characteristic data information, the model query extraction module is used for searching and searching models built by human bodies in a human body database, the model query extraction module and the human body database can carry out bidirectional information acquisition, the human body database is used for storing human body model data information, the human body database is connected with an information updating module to update human body data in real time, the central processing unit is connected with the head model reconstruction module, the head model reconstruction module is used for modeling classified human body head data information, and the head model reconstruction module is connected with the human body model building module to carry out information acquisition;
the human body model building module is connected with the connection processing module to form data intercommunication, the connection processing module is used for performing connection processing on the human body model, and the connection processing module is sequentially distributed and connected with the surface processing module, the feature adding module, the model map processing module and the model display module to perform display building on the human body feature model;
the surface processing module is used for carrying out surface finishing processing on the reconstructed human body model, and the characteristic adding module is used for adding main characteristic points contained in a human body; the model map processing module is used for re-projecting the obtained human body model on each image and carrying out calibration in position and direction, and the model display module is used for displaying and viewing the reconstructed model in an image mode;
the central processing unit is respectively connected with the first model building module, the second model building module and the third model building module and is used for building models for classifying data information of all parts of a human body; the head model reconstruction module, the first model construction module, the second model construction module and the third model construction module are respectively connected with the information updating module, and model data are obtained from the human body database through the information updating module.
2. The system for reconstructing a three-dimensional human body surface with high accuracy as claimed in claim 1, wherein the model query extraction module employs a hybrid optimization algorithm combining a bat algorithm and a cuckoo optimization algorithm, and the specific steps are as follows:
s1, obtaining a coding sequence L by adopting a subsequent traversal connecting tree, initializing each parameter of a hybrid optimization algorithm, calculating the adaptability value of the bird nest according to an objective function, determining the current optimal position and optimal value of the bird nest, reserving the position of the bird nest with the previous generation optimal position, updating the position of the bird nest by utilizing an algorithm of a Levy Flight mechanism to obtain a new position of the bird nest, and performing border crossing processing;
the Levy Flight mechanism has the algorithm as follows:
in the formula (I), the compound is shown in the specification,andrespectively representing the position vectors of the ith bird nest in the t generation and the t +1 generation, alpha is a step length adjusting factor and can take different values according to different conditions,for point-to-point multiplication, Le' vy (lambda) is a random search path;
s2, calculating the fitness value of the new bird nest position by using the objective function, comparing the fitness value with the previous generation, covering the fitness value of the previous generation when the fitness value is superior to the fitness value of the previous generation, updating the corresponding bird nest position, determining the current optimal bird nest position and the optimal value, and then using the discovery probability p of the foreign egg α And random number R epsilon [0,1 ] subjected to uniform distribution]Making a comparison, e.g. R > p α Randomly changing the positions of the bird nests to obtain a group of new bird nest positions;
s3, taking the new bird nest position obtained in the last step as the initial point of a bat algorithm, updating the positions of the bird nests by the bat algorithm to obtain a group of new bird nest positions, evaluating the fitness value of the bird nest positions, determining the current optimal bird nest position and the optimal value after comparison, and when the termination condition is reached, inquiring a plan in a database corresponding to the optimal bird nest position;
wherein, the bat algorithm comprises the following steps:
initial population number of individuals NP, maximum pulse volume A 0 Maximum pulse rate R 0 Search pulse frequency range [ f min ,f max ]Attenuation coefficient alpha of sound volume, enhancement coefficient gamma of search frequency, search precision epsilon or maximum iteration number iter _ max, and then randomly initializing the bat position x i And according to the quality of the fitness value, searching the current optimal solution x * In the evolution process of the population, the search pulse frequency, the speed and the position of each generation of individuals are changed according to the following formula:
f i =f min +(f max -f min )×β
In the formula, beta is belonged to [0,1 ]]Is a uniformly distributed random number, f i Is the search pulse frequency of bat i,respectively representing the velocity of the bat i at the time t and t-1,respectively representing the position, x, of the bat i at times t and t-1 * Representing the optimal solution of all current bats, and then generating a uniformly distributed random number rand if rand > r i Then the random disturbance is carried out to the current optimum solution to generate a new solution, and the new solution is processed by the border crossing, if rand is less than A i And f (x) i )<f(x * ) Then receive the new solution generated in the previous step and then apply r according to the following formula i And A i Updating:
r i t+1 =R 0 [1-exp(-λt)]
and then outputting the global optimal value and the optimal solution.
3. The system for reconstructing a three-dimensional human body surface with high accuracy according to claim 1, wherein the information acquisition unit comprises a scanning acquisition module and a data entry module, and the scanning acquisition module and the data entry module can perform information entry and data direct entry of a three-dimensional scanning device respectively.
4. The system for reconstructing a three-dimensional human body surface with high accuracy according to claim 3, wherein the scanning acquisition module scans the human body image by using an infrared camera to obtain image data with clear outline; the data entry module is used for entering the relevant data information of the human body.
5. The system for reconstructing a three-dimensional human body surface with high accuracy according to claim 1, wherein the human body database comprises a first model data, a second model data and a third model data, and is configured to perform classified storage and management on a plurality of pieces of model data of the human body, and the data stored in the first model data, the second model data and the third model data respectively correspond to data to be constructed by the first model construction module, the second model construction module and the third model construction module.
6. The system of claim 1, wherein the model mapping processing module is configured to re-project the obtained human body model onto each image, perform calibration in position and direction, and then project color and texture information in the image onto a triangulated vertex of the model by using a texture mapping function in OpenGL to complete texture mapping and material restoration, thereby obtaining the surface-processed three-dimensional model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210736632.9A CN114998527B (en) | 2022-06-27 | 2022-06-27 | High-accuracy three-dimensional human body surface reconstruction system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210736632.9A CN114998527B (en) | 2022-06-27 | 2022-06-27 | High-accuracy three-dimensional human body surface reconstruction system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114998527A true CN114998527A (en) | 2022-09-02 |
CN114998527B CN114998527B (en) | 2024-05-03 |
Family
ID=83037620
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210736632.9A Active CN114998527B (en) | 2022-06-27 | 2022-06-27 | High-accuracy three-dimensional human body surface reconstruction system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114998527B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115329615A (en) * | 2022-10-18 | 2022-11-11 | 北京名洋数字科技股份有限公司 | Automatic projection model system and method for pre-storing exhibit based on database |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101650834A (en) * | 2009-07-16 | 2010-02-17 | 上海交通大学 | Three dimensional reconstruction method of human body surface under complex scene |
US20100158354A1 (en) * | 2008-12-22 | 2010-06-24 | Electronics And Telecommunications Research Institute | Method of creating animatable digital clone from multi-view images |
CN103955963A (en) * | 2014-04-30 | 2014-07-30 | 崔岩 | Digital human body three-dimensional reconstruction method and system based on Kinect device |
CN109087340A (en) * | 2018-06-04 | 2018-12-25 | 成都通甲优博科技有限责任公司 | A kind of face three-dimensional rebuilding method and system comprising dimensional information |
CN109800653A (en) * | 2018-12-24 | 2019-05-24 | 重庆第二师范学院 | A kind of characteristics of human body's parameter extracting method and system based on image analysis |
CN110047131A (en) * | 2019-04-02 | 2019-07-23 | 中科恒运股份有限公司 | A kind of point cloud fusion method that human body three-dimensional is rebuild |
US20190295250A1 (en) * | 2016-07-25 | 2019-09-26 | Nuctech Company Limited | Method, apparatus and system for reconstructing images of 3d surface |
US20200126295A1 (en) * | 2018-10-22 | 2020-04-23 | The Hong Kong Polytechnic University | Method and/or system for reconstructing from images a personalized 3d human body model and thereof |
WO2021184933A1 (en) * | 2020-03-20 | 2021-09-23 | 华为技术有限公司 | Three-dimensional human body model reconstruction method |
CN113936090A (en) * | 2021-06-18 | 2022-01-14 | 中国科学院自动化研究所 | Three-dimensional human body reconstruction method and device, electronic equipment and storage medium |
-
2022
- 2022-06-27 CN CN202210736632.9A patent/CN114998527B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100158354A1 (en) * | 2008-12-22 | 2010-06-24 | Electronics And Telecommunications Research Institute | Method of creating animatable digital clone from multi-view images |
CN101650834A (en) * | 2009-07-16 | 2010-02-17 | 上海交通大学 | Three dimensional reconstruction method of human body surface under complex scene |
CN103955963A (en) * | 2014-04-30 | 2014-07-30 | 崔岩 | Digital human body three-dimensional reconstruction method and system based on Kinect device |
US20190295250A1 (en) * | 2016-07-25 | 2019-09-26 | Nuctech Company Limited | Method, apparatus and system for reconstructing images of 3d surface |
CN109087340A (en) * | 2018-06-04 | 2018-12-25 | 成都通甲优博科技有限责任公司 | A kind of face three-dimensional rebuilding method and system comprising dimensional information |
US20200126295A1 (en) * | 2018-10-22 | 2020-04-23 | The Hong Kong Polytechnic University | Method and/or system for reconstructing from images a personalized 3d human body model and thereof |
CN109800653A (en) * | 2018-12-24 | 2019-05-24 | 重庆第二师范学院 | A kind of characteristics of human body's parameter extracting method and system based on image analysis |
CN110047131A (en) * | 2019-04-02 | 2019-07-23 | 中科恒运股份有限公司 | A kind of point cloud fusion method that human body three-dimensional is rebuild |
WO2021184933A1 (en) * | 2020-03-20 | 2021-09-23 | 华为技术有限公司 | Three-dimensional human body model reconstruction method |
CN113936090A (en) * | 2021-06-18 | 2022-01-14 | 中国科学院自动化研究所 | Three-dimensional human body reconstruction method and device, electronic equipment and storage medium |
Non-Patent Citations (1)
Title |
---|
陈佳宇;钟跃崎;余志才;: "基于二值图像的三维人体模型重建", 毛纺科技, no. 09, 16 September 2020 (2020-09-16) * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115329615A (en) * | 2022-10-18 | 2022-11-11 | 北京名洋数字科技股份有限公司 | Automatic projection model system and method for pre-storing exhibit based on database |
CN115329615B (en) * | 2022-10-18 | 2023-02-10 | 北京名洋数字科技股份有限公司 | Automatic projection model system and method for pre-storing exhibits based on database |
Also Published As
Publication number | Publication date |
---|---|
CN114998527B (en) | 2024-05-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107945282B (en) | Rapid multi-view three-dimensional synthesis and display method and device based on countermeasure network | |
CN109255831B (en) | Single-view face three-dimensional reconstruction and texture generation method based on multi-task learning | |
CN109344855B (en) | Depth model face beauty evaluation method based on sequencing guided regression | |
CN105513114B (en) | The method and apparatus of three-dimensional animation generation | |
CN110428493B (en) | Single-image human body three-dimensional reconstruction method and system based on grid deformation | |
CN109598754A (en) | A kind of binocular depth estimation method based on depth convolutional network | |
CN105931283B (en) | A kind of 3-dimensional digital content intelligence production cloud platform based on motion capture big data | |
CN108648161A (en) | The binocular vision obstacle detection system and method for asymmetric nuclear convolutional neural networks | |
WO2015149302A1 (en) | Method for rebuilding tree model on the basis of point cloud and data driving | |
CN112529999A (en) | Parameter estimation model training method, device, equipment and storage medium | |
CN108154104A (en) | A kind of estimation method of human posture based on depth image super-pixel union feature | |
CN112633191A (en) | Method, device and equipment for reconstructing three-dimensional face and storage medium | |
CN112288645A (en) | Skull face restoration model construction method, restoration method and restoration system | |
CN114998527B (en) | High-accuracy three-dimensional human body surface reconstruction system | |
WO2024103890A1 (en) | Model construction method and apparatus, reconstruction method and apparatus, and electronic device and non-volatile readable storage medium | |
CN111354077A (en) | Three-dimensional face reconstruction method based on binocular vision | |
CN107610219A (en) | The thick densification method of Pixel-level point cloud that geometry clue perceives in a kind of three-dimensional scenic reconstruct | |
CN108615229A (en) | Collision detection optimization method based on curvature points cluster and decision tree | |
CN113362422A (en) | Shadow robust makeup transfer system and method based on decoupling representation | |
US11138785B2 (en) | Method and system for generating 3D image of character | |
CN113255514B (en) | Behavior identification method based on local scene perception graph convolutional network | |
CN109887023A (en) | A kind of binocular fusion stereo image quality evaluation method based on weighted gradient amplitude | |
CN115346091A (en) | Method and device for generating Mura defect image data set | |
CN114972584A (en) | Method, system, equipment and product for constructing multi-dimensional model of three-dimensional animation | |
Hu et al. | Self-perceptual generative adversarial network for synthetic aperture sonar image generation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |