A Study of Vision-Based Measurement of Weld Joint Shape Incorporating the Neural Network

1994 ◽  
Vol 208 (1) ◽  
pp. 61-69 ◽  
Author(s):  
K-Y Bae ◽  
S-J Na
Keyword(s):  
Neural Network ◽  
Three Dimensional ◽  
Ccd Camera ◽  
Image Plane ◽  
Welding Distortion ◽  
Surface Geometry ◽  
Visual Sensing ◽  
Laser Stripe ◽  
The Neural Network ◽  
A Charge

Visual sensing of the surface geometry is often necessary to inspect and evaluate the quality of welded joints as well as to sense the transient distortion of a structure during welding for the feedback of its current geometry. This investigation presents a simple and non-contact digitization method of the vision-based system for measuring the three-dimensional surface geometry of the object distorted by welding. Its basic principles are based on the equation derived from the geometric optics, for which the illumination of the laser beam was controlled in the form of the projected plane. This method utilized a 10 mW He-Ne laser for the structured light and a charge coupled device (CCD) camera as the vision sensor. When the laser stripe is projected on to the weldment, a minute deviation from the perfect plane existing on the specimen surface causes a distortion of the stripe. The shape and amount of the weldment distortion can be then calculated by analysing the distorted laser stripe. In this study, a neural network was proposed and implemented for recognizing the laser stripe features from the image plane. A calibration scheme of corresponding an image to the world position was also adopted for determining the sectional features of the welding distortion. The feasibility of determining the welding distortion by the proposed vision-based system was demonstrated through the experiments with various types of specimen.

scholarly journals A convolutional neural network for defect classification in Bragg coherent X-ray diffraction

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Bruce Lim ◽  
Ewen Bellec ◽  
Maxime Dupraz ◽  
Steven Leake ◽  
Andrea Resta ◽  
...  
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Diffraction Pattern ◽  
Three Dimensional ◽  
X Ray Diffraction ◽  
Coherent Diffraction Imaging ◽  
The Neural Network ◽  
Diffraction Imaging ◽  
Diffraction Patterns ◽  
Defect Recognition

AbstractCoherent diffraction imaging enables the imaging of individual defects, such as dislocations or stacking faults, in materials. These defects and their surrounding elastic strain fields have a critical influence on the macroscopic properties and functionality of materials. However, their identification in Bragg coherent diffraction imaging remains a challenge and requires significant data mining. The ability to identify defects from the diffraction pattern alone would be a significant advantage when targeting specific defect types and accelerates experiment design and execution. Here, we exploit a computational tool based on a three-dimensional (3D) parametric atomistic model and a convolutional neural network to predict dislocations in a crystal from its 3D coherent diffraction pattern. Simulated diffraction patterns from several thousands of relaxed atomistic configurations of nanocrystals are used to train the neural network and to predict the presence or absence of dislocations as well as their type (screw or edge). Our study paves the way for defect-recognition in 3D coherent diffraction patterns for material science.

scholarly journals DEVELOPMENT OF ALGORITHMIC SUPPORT AND A DEDICATED DEVICE FOR IMAGING THE INTERNAL ORGANS OF THE PATIENT ACCORDING TO MRI AND CT SCANNERS

2019 ◽  
Vol 22 (6) ◽  
pp. 189-197
Author(s):  
E. S. Sirota ◽  
M. I. Truphanov
Keyword(s):  
Neural Network ◽  
Three Dimensional ◽  
Color Difference ◽  
Learning Criterion ◽  
Average Value ◽  
Advantages And Disadvantages ◽  
The Neural Network ◽  
Training Stage ◽  
Algorithmic Support

In work the algorithm of restoration of the images damaged as a result of influence of noise of various nature is considered. The advantages and disadvantages of the existing approaches, as well as the prospects of using artificial neural networks, are noted. A double-layer neural network is used as an image restoration tool, and it is assumed that the location of the damaged pixels is known. A neuron is represented as a 3x3 array, where each element of the array has a pixel color value that corresponds to the value of that color in the palette. The neural network is trained on intact images, while the color difference of pixels acts as a learning criterion. For a more accurate restoration, it is recommended at the training stage to select images similar in color to damaged ones. At the recovery stage, neurons (3x3) are formed around the damaged pixels, so that the damaged pixel is located in the middle of the neuron data array. The damaged pixel is assigned a neuron value depending on the average value of the weights matrix. An algorithm for the restoration of pixels, as well as its software implementation. The simulation was carried out in the RGB palette separately for each channel. To assess the quality of the recovery were selected groups of images with varying degrees of damage. Unlike existing solutions, the algorithm has the simplicity of implementation. The  research results show that regardless of the degree of damage (within 50%), about 70% of damaged pixels are restored. Further studies suggest a modification of the algorithm to restore images with enlarged areas of damage, as well as adapting it to restore three-dimensional images.

scholarly journals Reconstruction of 3D Object Shape Using Hybrid Modular Neural Network Architecture Trained on 3D Models from ShapeNetCore Dataset

Sensors ◽  
2019 ◽  
Vol 19 (7) ◽  
pp. 1553 ◽  
Author(s):  
Audrius Kulikajevas ◽  
Rytis Maskeliūnas ◽  
Robertas Damaševičius ◽  
Sanjay Misra
Keyword(s):  
Neural Network ◽  
Network Architecture ◽  
A Priori ◽  
Three Dimensional ◽  
Qualitative Evaluation ◽  
3D Models ◽  
Depth Sensors ◽  
The Neural Network ◽  
Artificial Neural Network Ann ◽  
Commercial Applications

Depth-based reconstruction of three-dimensional (3D) shape of objects is one of core problems in computer vision with a lot of commercial applications. However, the 3D scanning for point cloud-based video streaming is expensive and is generally unattainable to an average user due to required setup of multiple depth sensors. We propose a novel hybrid modular artificial neural network (ANN) architecture, which can reconstruct smooth polygonal meshes from a single depth frame, using a priori knowledge. The architecture of neural network consists of separate nodes for recognition of object type and reconstruction thus allowing for easy retraining and extension for new object types. We performed recognition of nine real-world objects using the neural network trained on the ShapeNetCore model dataset. The results evaluated quantitatively using the Intersection-over-Union (IoU), Completeness, Correctness and Quality metrics, and qualitative evaluation by visual inspection demonstrate the robustness of the proposed architecture with respect to different viewing angles and illumination conditions.

Structured-Light Based Rapid 3D Measurement of Plant Canopy Structure

2014 ◽  
Vol 701-702 ◽  
pp. 549-553 ◽  
Author(s):  
Zhe Chen ◽  
Xu Dong Li ◽  
Shao Guang Shi ◽  
Hong Zhi Jiang ◽  
Hui Jie Zhao
Keyword(s):  
Data Collection ◽  
Structure Data ◽  
Measurement System ◽  
Structured Light ◽  
Canopy Structure ◽  
Three Dimensional ◽  
Ccd Camera ◽  
Plant Canopy ◽  
Laser Stripe ◽  
3D Data

Density three dimensional plant canopy structure data has numerous applications in agriculture, but many existing 3D data collection approaches are time-consuming. In this paper, we present a measurement system based on structured-light for plant canopy structure data collection. The structured-light projector projects laser beam reflected by dual-oscillating mirror, arrives to the plant canopy, which is captured by a camera. We propose a new scanning mode, that is, during one exposure time of CCD camera, one mirror keeps moving in high frequency and small angle, while the other one maintains the same position, so that we can get a laser stripe rather than a spot in each image, from which about 100 sub-pixel centers of laser stripe can be extracted. Experiments show that the measurement system can rapid collect three dimensional information of the plant.

Flow Stress Prediction of High-Nb TiAl Alloys under High Temperature Deformation

2012 ◽  
Vol 510 ◽  
pp. 723-728 ◽  
Author(s):  
Liang Cheng ◽  
Hui Chang ◽  
Bin Tang ◽  
Hong Chao Kou ◽  
Jin Shan Li
Keyword(s):  
Neural Network ◽  
High Temperature ◽  
Flow Stress ◽  
Back Propagation ◽  
Three Dimensional ◽  
High Temperature Deformation ◽  
Temperature Deformation ◽  
Ann Model ◽  
The Neural Network ◽  
Hidden Layer

In this work, a back propagation artificial neural network (BP-ANN) model is conducted to predict the flow behaviors of high-Nb TiAl (TNB) alloys during high temperature deformation. The inputs of the neural network are deformation temperature, log strain rate and strain whereas flow stress is the output. There is a single hidden layer with 7 neutrons in the network, and the weights and bias of the network were optimized by Genetic Algorithm (GA). The comparison result suggests a very good correlation between experimental and predicted data. Besides, the non-experimental flow stress predicted by the network is shown to be in good agreement with the results calculated by three dimensional interpolation, which confirmed a good generalization capability of the proposed network.

An optimal design of axial-flow fan blades by the machining method and an artificial neural network

2002 ◽  
Vol 216 (3) ◽  
pp. 367-376 ◽  
Author(s):  
B-J Lin ◽  
C-I Hung ◽  
E-J Tang
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Tool Path ◽  
Three Dimensional ◽  
Axial Flow ◽  
Integrated Approach ◽  
Optimization Method ◽  
Surface Geometry ◽  
Axial Flow Fan ◽  
Artificial Neural

The geometry design and machining of blades for axial-flow fans are important issues because the twisted profile and flowfield of blades are complicated. The rapid design of a blade that performs well and satisfies machining requirements is one of the goals in designing fluid machinery blades. In this study, an integrated approach combining computational fluid dynamics (CFD), an artificial neural network, an optimization method and a machining method is proposed to design a three-dimensional blade for an axial-flow fan. From the machining point of view, the three-dimensional surface geometry of a fan blade can be defined as the swept surface of the tool path created by using the generated machining method. By taking advantage of its powerful learning capability, a back-propagation artificial neural network is used to set up the flowfield models and to forecast the flow performance of the axial-flow fan. The desired optimal blade geometry is obtained by using a complex optimization method.

scholarly journals Q QUANTITATIVE METHOD OF CALCULATING APPROACHES OF EXERCISES ON THE BASIS OF OUTPUT SIGNAL OF THE NEURAL NETWORK

2020 ◽  
pp. 65-69
Author(s):  
Juliy Broyda
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Computer Vision ◽  
Output Signal ◽  
Three Dimensional ◽  
Test Scheme ◽  
Human Posture ◽  
Professional Activities ◽  
The Neural Network ◽  
Human Joints

Today, artificial intelligence and computer vision systems are developing rapidly in the world; in particular, new architectures of neural networks that assess the three-dimensional human posture on video are produced. Such neural networks require analysis of their output signal in order to obtain useful data for the end user and their subsequent integration into user systems. The author proposes a new method of analysis of the output signal of the neural network, which estimates the position of a person in space that performs the calculation of repetitions of the exercise "squat". This method is based on the state machine, which adds one to the repetition counter at the end of the exercise cycle. The application of this method in the initial stages of the algorithm of exercise analysis will allow further development of systems that test squats and help athletes and coaches during training, as well as scientists in the field of biomechanics during their professional activities. A distinctive feature of this method is resistance to both input signal emissions, i.e. incorrect results of human posture recognition by the neural network, and to human movements that do not belong to the exercise directly. Also, the application of this method to the analysis of the neural network signal allows to combine the positive qualities inherent in neural networks used in computer vision (admissibility of high variability of clothing and background), and the positive qualities of analytical and algorithmic methods (easy interpretability of results, convenient adjustment, possibility to use the experts' subject experience for the selection of parameters). This method is not specific to any particular neural network and therefore can be used at the output of almost any system that determines the position of human joints in space. In addition to the description of the method, the article presents the results of its tests in different conditions. This test scheme can be used not only to apply this method to the exercise "squat", but to any other cyclic exercise.

scholarly journals Diagnosing Chronic Glaucoma Using Watershed and Convolutional Neural Network

2021 ◽  
Vol 9 (9) ◽  
pp. 272-283
Author(s):  
Naureen Fathima
Keyword(s):  
Neural Network ◽  
Image Processing ◽  
Convolutional Neural Network ◽  
Language Processing ◽  
Three Dimensional ◽  
Image Plane ◽  
Processing Technique ◽  
Image Processing Technique ◽  
Computational Technique ◽  
Fundus Imaging

Abstract: Glaucoma is a disease that relates to the vision of human eye,Glaucoma is a disease that affects the human eye's vision. This sickness is regarded as an irreversible condition that causes eyesight degeneration. One of the most common causes of lifelong blindness is glaucoma in persons over the age of 40. Because of its trade-off between portability, size, and cost, fundus imaging is the most often utilised screening tool for glaucoma detection. Fundus imaging is a two-dimensional (2D) depiction of the three-dimensional (3D), semitransparent retinal tissues projected on to the imaging plane using reflected light. The idea plane that depicts the physical display screen through which a user perceives a virtual 3D scene is referred to as the "image plane”. The bulk of current algorithms for autonomous glaucoma assessment using fundus images rely on handcrafted segmentation-based features, which are influenced by the segmentation method used and the retrieved features. Convolutional neural networks (CNNs) are known for, among other things, their ability to learn highly discriminative features from raw pixel intensities. This work describes a computational technique for detecting glaucoma automatically. The major goal is to use a "image processing technique" to diagnose glaucoma using a fundus image as input. It trains datasets using a convolutional neural network (CNN). The Watershed algorithm is used for segmentation and is the most widely used technique in image processing. The following image processing processes are performed: region of interest, morphological procedures, and segmentation. This technique can be used to determine whether or not a person has Glaucoma. Keywords: Recommender system, item-based collaborative filtering, Natural Language Processing, Deep learning.

Conductivity inversion of unidirectional CFRP laminate based on ECT using neural network

2020 ◽  
Vol 64 (1-4) ◽  
pp. 1431-1438
Author(s):  
Hongli Ji ◽  
Wei Shen ◽  
Chao Zhang ◽  
Xiaojuan Xu ◽  
Jinhao Qiu
Keyword(s):  
Neural Network ◽  
Eddy Current ◽  
Three Dimensional ◽  
Electrical Anisotropy ◽  
Détection Signal ◽  
Cfrp Laminate ◽  
Electromagnetic Model ◽  
The Neural Network ◽  
Probe Voltage ◽  
Detection Signal

For the electrical anisotropy of carbon fiber reinforced polymer (CFRP), conductivity of unidirectional CFRP laminate in three directions was inverted in this paper. The three-dimensional eddy current electromagnetic model of unidirectional composites was constructed by ANSYS software, and the influence of the electrical conductivity of the material on the detection signal of the probe in the longitudinal, transverse and thickness directions was studied. In order to improve the amplitude of the probe output signal induced by the change of conductivity, the optimal detection angle of the eddy current probe was determined. On this basis, the relationship between the conductivity and the detection signal was studied to estimate the initial values of the conductivity based on the experimental data obtained by the eddy current testing (ECT). According to the forward model, the theoretical probe voltage under the estimated conductivity were calculated. The database consisting of conductivity and corresponding theoretical results was built for the neural network to construct the mapping that can estimate conductivity by experimental results. Using neural network for iteration, the conductivity was inverted quickly and precisely.

Sign in / Sign up

Export Citation Format

Share Document