scholarly journals Study on 3D Clothing Color Application Based on Deep Learning-Enabled Macro-Micro Adversarial Network and Human Body Modeling

2021 ◽  
Vol 2021 ◽  
pp. 1-10
Author(s):  
Jingmiao Liu ◽  
Yu Ren ◽  
Xiaotong Qin
Keyword(s):  
Deep Learning ◽  
3D Model ◽  
Subjective Evaluation ◽  
Three Dimensional ◽  
Real Life ◽  
Color Difference ◽  
Original Image ◽  
Three Dimensional Model ◽  
Adversarial Network ◽  
Original Picture

In real life, people’s life gradually tends to be simple, so the convenience of online shopping makes more and more research begin to explore the convenience optimization of shopping, in which the fitting system is the research product. However, due to the immaturity of the virtual fitting system, there are a lot of problems, such as the expression of clothing color is not clear or deviation. In view of this, this paper proposes a 3D clothing color display model based on deep learning to support human modeling-driven. Firstly, the macro-micro adversarial network (MMAN) based on deep learning is used to analyze the original image, and then, the results are preprocessed. Finally, the 3D model with the original image color is constructed by using UV mapping. The experimental results show that the accuracy of the MMAN algorithm reaches 0.972, the established three-dimensional model is emotional enough, the expression of the clothing color is clear, and the difference between the color difference and the original image is within 0.01, and the subjective evaluation of volunteers is more than 90 points. The above results show that it is effective to use deep learning to build a 3D model with the original picture clothing color, which has great guiding significance for the research of character model modeling and simulation.

scholarly journals 3D Model Generation on Architectural Plan and Section Training through Machine Learning

Technologies ◽  
2019 ◽  
Vol 7 (4) ◽  
pp. 82
Author(s):  
Hang Zhang
Keyword(s):  
Machine Learning ◽  
3D Model ◽  
Three Dimensional ◽  
Image Data ◽  
Input Image ◽  
Model Generation ◽  
Three Dimensional Model ◽  
Generative Adversarial Network ◽  
Learning Group ◽  
Adversarial Network

Machine learning, especially the GAN (Generative Adversarial Network) model, has been developed tremendously in recent years. Since the NVIDIA Machine Learning group presented the StyleGAN in December 2018, it has become a new way for designers to make machines learn different or similar types of architectural photos, drawings, and renderings, then generate (a) similar fake images, (b) style-mixing images, and (c) truncation trick images. The author both collected and created input image data, and specially made architectural plan and section drawing inputs with a clear design purpose, then applied StyleGAN to train specific networks on these datasets. With the training process, we could look into the deep relationship between these input architectural plans or sections, then generate serialized transformation images (truncation trick images) to form the 3D (three-dimensional) model with a decent resolution (up to 1024 × 1024 × 1024 pixels). Though the results of the 3D model generation are difficult to use directly in 3D spatial modeling, these unexpected 3D forms still could inspire new design methods and greater possibilities of architectural plan and section design.

Intelligent Identification and Locating System of Steel Coil Based on Laser Technology

2014 ◽  
Vol 496-500 ◽  
pp. 1308-1312
Author(s):  
Yu Chi Wang ◽  
Si Yue Zhou ◽  
Zheng Peng Yuan
Keyword(s):  
3D Model ◽  
Three Dimensional ◽  
Least Square Method ◽  
Least Square ◽  
Circle Fitting ◽  
Three Dimensional Model ◽  
Laser Technology ◽  
Steel Coil ◽  
Main Emphasis ◽  
Locating System

In this paper, a solution based on laser technology is proposed for steel coil handling. The main emphasis is placed on the problem of steel coil identification and locating. In this paper, we first use laser data to construct three dimensional model of the worksite. And then we slice up the 3D model repeatedly to get the section images and make them to fusion. In order to wipe off the disturb of line, a line detection method based on probabilistic hough transform is proposed. Finally an improved least square method used in circle fitting is discussed.

scholarly journals Research on Application of Full Attribute 3D Design Model in Engineering

2021 ◽  
Vol 294 ◽  
pp. 02002
Author(s):  
Xinqian Xu ◽  
Rui Xia ◽  
Lai Xu ◽  
Xiwen Hu
Keyword(s):  
3D Model ◽  
Three Dimensional ◽  
Three Dimensional Model ◽  
Process Data ◽  
3D Design ◽  
Engineering Construction ◽  
Data Application ◽  
Construction Operation ◽  
Research On Application ◽  
Calculation Analysis

In order to take full advantages of the integration of 3D technology graphics and data, this paper proposes the concept of a full-attribute 3D model that satisfies the full-process data application of substation engineering construction. All stages of engineering construction extract the required data from the full-attribute 3D model for calculation, Analysis, statistics, management and other functions, to achieve all-round coverage of three-dimensional data information from design to construction and construction management, laying the foundation for data circulation and management of the entire life cycle of substation projects. A method of applying a full-attribute three- dimensional model for construction, operation and maintenance management is proposed, which lays a foundation for the realization of refined, digital and intelligent project management.

scholarly journals Segmentation and 3D Reconstruction of Cerebral Hematoma Based on Deep Learning

CONVERTER ◽  
2021 ◽  
pp. 482-488
Author(s):  
Taipeng Guo, Et al.
Keyword(s):  
Deep Learning ◽  
Cerebral Hemorrhage ◽  
Three Dimensional ◽  
Rapid Onset ◽  
Three Dimensional Model ◽  
Data Set ◽  
Disability Rate ◽  
Short Period ◽  
Dimensional Reconstruction ◽  
Cerebral Hematoma

Cerebral hemorrhage is a common clinical disease. Because of its rapid onset, high mortality and disability rate, in the treatment of cerebral hemorrhage, it is very important to accurately calculate the brain hematoma volume and feedback its location information in a short period of time. This paper proposes a method for precise segmentation and three-dimensional reconstruction of cerebral hematoma area based on deep learning. This method highlights the image information by expanding the CT image and eliminating the skull information, then accurately segments the cerebral hematoma areathrough the neural network model to build a three-dimensional model. We verify the experimental results based on the data set collected by the Affiliated Hospital of Xiangnan University, which proves the effectiveness of this method and its ability to significantly improve the speed incerebral hemorrhage area judgment and grasp information in clinical diagnosis.

scholarly journals Influence of Geometric Features Associativity of CAD Class Models on the Process of Their Updating – Comparative Analysis

2021 ◽  
Author(s):  
Grzegorz Świaczny
Keyword(s):  
3D Model ◽  
Three Dimensional ◽  
Optimal Solution ◽  
3D Models ◽  
Dimensional Model ◽  
Three Dimensional Model ◽  
Initial Design ◽  
The Individual ◽  
Class Models ◽  
Positive Effect

This article deals with the topic of one of the most important features of modern CAx class systems – associativity. The term refers to the ability to form relations (links) between two or more objects (in terms of their selected features), and with the consequence creating an associative (linked) three-dimensional model. The author pays special attention to the very process of creating relations between objects, as it has a key impact on the structural stability of CAD class models, and thus on their susceptibility to possible modifications. To show that not all associativity brings a positive effect, the author presents two examples of its implementation. In order to emphasize the influence of the method of linking individual elements, both examples are based on the same 3D model – a thin-walled part with a positioning pin. That means the geometric form of the default part is the same, whereas only relations of the individual objects of the 3D model change. In the first scenario, correctly defined relations between objects make that the positioning pin offset does not affect the initial design conditions. The second scenario shows an incorrect implementation of associativity, as a result of which the same operation of positioning pin offset gives non-compliance with the initial design conditions and with the consequence an undesirable change in its geometry. The article is an attempt to draw attention to the fact that the associative structure of 3D models is not always equal to the optimal solution. Only the well-thought-out nature of associativity allows to use all its advantages.

scholarly journals PHOTOGRAMMETRIC METHOD OF FORENSIC PHOTOGRAPHY Artem Kovalenk

2020 ◽  
Keyword(s):  
3D Model ◽  
Three Dimensional ◽  
Digital Camera ◽  
Criminal Proceedings ◽  
Three Dimensional Model ◽  
Forensic Photography ◽  
Computer Image Processing ◽  
Special Software ◽  
Photogrammetric Method ◽  
Special Rules

The article is devoted to defining the essence and directions of application of the photogrammetric method of forensic photography in criminal proceedings. It is determined that the photogrammetric method of forensic photography is to photograph objects according to special rules in order to further measure them on the obtained images. It is determined that the photogrammetric method of judicial photography consists in photographing objects according to special rules in order to perform their further measurements using the obtained images. The method of measuring objects using a single image is called monophotogrammetric and is known in the theory of forensic photography as measuring (scale) method; using a pair of images is called stereophotogrammetric (known as stereoscopic); measurement with more than two images is a polyphotogrammetric method. As a result of polyphotogrammetric fixation of an object with the help of a serial digital camera and computer image processing, it is possible to obtain its high-quality geometrically accurate textured three-dimensional model. Objects of different sizes, from micro-objects, to entire buildings or areas, interiors can be scanned with polyphotogrammetric method. The 3D model of the scanned object reconstructed as a result of polyphotogrammetric photography and software processing can be saved in a number of conventional formats in electronic form, viewed and researched using special software. The rules and algorithm of polyphotogrammetric scaning are given, the process of reconstruction of 3D model on the basis of received images is described. The advantages of the researched method of forensic photography are its non-invasiveness, clarity and accuracy of the obtained models, wide possibilities of their use and research.

scholarly journals Modeling in AUTOCAD for bachelors

2021 ◽  
pp. 25-28
Author(s):  
Л.В. Карпюк ◽  
Н.О. Давіденко
Keyword(s):  
3D Model ◽  
Three Dimensional ◽  
Model Space ◽  
Educational Process ◽  
Point Of View ◽  
Two Dimensional ◽  
Dimensional Model ◽  
Three Dimensional Model ◽  
Three Dimensional Objects ◽  
Graphic Editor

The article discusses the methods of using the AutoCad graphic editor for creating three-dimensional objects. The possibilities of three-dimensional modeling in the AutoCad graphic editor for optimizing the educational process of bachelors of technical specialties are also considered. The article analyzes the best ways to create mechanical engineering drawings.The most developed software tool for the production of design documentation is AutoCAD - a universal graphic design system. Creating models of any complexity in space by using this graphic editor, the user will be able to see their relative position, estimate the distance between them. The model can be freely moved in space, viewing many options. The ability to control the point of view allows to conveniently select the view of the 3D model that is being developed. Zooming, panning in real time with the ability to freely rotate the camera around the model provide the ability to quickly view objects from any point of view. The article provides examples of choosing the most optimal option for creating a three-dimensional model. The traditional way to create a 3D model drawing is to make 2D views of the model. When creating a flat drawing, there is a possibility of error when making projections, since they are created independently from each other and consist of several images. It is rather difficult to represent an object in space from a flat drawing. At present, modern software graphic editors are aimed at creating three-dimensional models that allow to create realistic models and, on their basis, get two-dimensional projections. Graphic editor AutoCad allows to create three-dimensional objects based on standard commands, in the form of a cylinder, cone, box, torus, etc., when editing which you can get the desired shapes. After creating a three-dimensional model, the user can get its two-dimensional projections not only on the main planes, but also on any plane at will. The 3D modeling method allows you to create a complex drawing with any number of images based on a 3D model. There are ways to create 2D plane drawings from a 3D model and the ability to edit ready-made designs that can be inserted from model space into paper space. Editing takes place by changing the parameters of a 3D object in model space, and these changes are automatically reflected in paper space. This method allows us to use the tools to quickly create a system of 3-4 linked views for a 3D AutoCad model.

scholarly journals Three-Dimensional Model Extraction and Parameter Measurement of Cerebral Hemangioma Based on Mimics

2022 ◽  
Vol 2160 (1) ◽  
pp. 012012
Author(s):  
An Qu ◽  
Tianmin Guan ◽  
Tianxiang Gan ◽  
Yuanyuan Li ◽  
Zhuang Lin ◽  
...  
Keyword(s):  
Therapeutic Effect ◽  
Cerebral Angiography ◽  
Structural Design ◽  
3D Model ◽  
Three Dimensional ◽  
The Body ◽  
Model Reconstruction ◽  
Dimensional Model ◽  
Three Dimensional Model ◽  
3D Model Reconstruction

Abstract In this paper, Mimics is using to reconstruct the 3D model of hemangioma from 2D cerebral angiography images. The process of 3D model reconstruction is formulated. The hemangioma model is extracted and the parameters of the hemangioma are measured, which provided a model basis for the structural design of the spring coil. The shape and structure of the coil are very important for the therapeutic effect. During the treatment, the coil is implanted into the hemangioma from outside the body through a catheter.

scholarly journals Reconstruction of an architectural object. Case study: Citadel fort No. 33 “Krakus” in Krakow

2018 ◽  
Vol 63 ◽  
pp. 00010
Author(s):  
Izabela Piech ◽  
Boguslawa Kwoczynska ◽  
Artur Ciszewski
Keyword(s):  
Point Cloud ◽  
3D Model ◽  
Three Dimensional ◽  
Laser Scanner ◽  
Field Work ◽  
Dimensional Model ◽  
Three Dimensional Model ◽  
Laser Measurements ◽  
The University

The aim of the study was to recreate, in the form of a 3D model, the Citadel fort No. 33 “Krakus” in Krakow. The data on the basis of which the three-dimensional model was made were obtained using a Leica ScanStation P40 terrestrial laser scanner, which is owned by the Faculty of Environmental Engineering and Geodesy of the University of Agriculture Hugona Kollataj in Krakow. The scope of field work included performing laser measurements, and then processing the point cloud in the Leica Cyclone 3D program and creating a full architectural model in SketchUp 2016.

A three-dimensional model of the human airway tree

1999 ◽  
Vol 87 (6) ◽  
pp. 2207-2217 ◽  
Author(s):  
Hiroko Kitaoka ◽  
Ryuji Takaki ◽  
Béla Suki
Keyword(s):  
3D Model ◽  
Human Lung ◽  
Three Dimensional ◽  
Deterministic Algorithm ◽  
Three Dimensional Model ◽  
Fluid Delivery ◽  
Human Airway ◽  
Airway Tree ◽  
Function Relationship ◽  
Good Agreement

A three-dimensional (3D) model of the human airway tree is proposed using a deterministic algorithm that can generate a branching duct system in an organ. The algorithm is based on two principles: 1) the amount of fluid delivery through a branch is proportional to the volume of the region it supplies; and 2) the terminal branches are arranged homogeneously within the organ. These principles define the basic process of branching: generation of the dimensions and directionality of two daughter branches is governed by the properties of the parent branch and the region the parent supplies. The algorithm is composed of nine basic rules and four complementary rules. When the contour of an organ and the position of the trunk are specified, branches are successively generated by the algorithm. Applied to the human lung, the algorithm generates an airway tree that consists of ∼54,000 branches. Its morphometric characteristics are in good agreement with those reported in the literature. The algorithm and the 3D airway model are useful for studying the structure-function relationship in the lung.

Sign in / Sign up

Export Citation Format

Share Document