Aluminum Plate Surface Defects Classification Based on the BP Neural Network

2015 ◽  
Vol 734 ◽  
pp. 543-547 ◽  
Author(s):  
Qing Hua Li ◽  
Di Liu
Keyword(s):  
Neural Network ◽  
Bp Neural Network ◽  
Surface Defects ◽  
Image Feature ◽  
Classification Model ◽  
Aluminum Plate ◽  
Geometric Feature ◽  
Plate Surface ◽  
Target Image ◽  
Defects Classification

The aluminum plate surface defects recognition method of BP neural network is studied based on target detection .In order to detect the defects, the target image is binaried by adaptive threshold method. After binarizing the target image, three kinds of image feature, including geometric feature, grayscale feature and shape feature, are extracted from the target image and its corresponding binary image. The defects classification model based on back-propagation neural network utilizes three layers neural network structure model and the hyperbolic tangent function of S function as the activation function, the number of neurons in hidden layer is confirmed by experiments. The experimental results show that the classification accuracy of BP neural network classification model as high as 94%, this can meet our requirements.

scholarly journals Deep Learning Framework for Steel Surface Defects Classification

2019 ◽  
Vol 4 (1) ◽  
pp. 25-32
Author(s):  
Ranganath Singari ◽  
Karun Singla ◽  
Gangesh Chawla
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Network Architecture ◽  
Steel Surface ◽  
Surface Defects ◽  
Fixed Number ◽  
Quality Level ◽  
Acceptance Sampling ◽  
Quality Inspection ◽  
Defects Classification

Deep learning has offered new avenues in the field of industrial management. Traditional methods of quality inspection such as Acceptance Sampling relies on a probabilistic measure derived from inspecting a sample of finished products. Evaluating a fixed number of products to derive the quality level for the complete batch is not a robust approach. Visual inspection solutions based on deep learning can be employed in the large manufacturing units to improve the quality inspection units for steel surface defect detection. This leads to optimization of the human capital due to reduction in manual intervention and turnaround time in the overall supply chain of the industry. Consequently, the sample size in the Acceptance sampling can be increased with minimal effort vis-à-vis an increase in the overall accuracy of the inspection. The learning curve of this work is supported by Convolutional Neural Network which has been used to extract feature representations from grayscale images to classify theinputs into six types of surface defects. The neural network architecture is compiled in Keras framework using Tensorflow backend with state of the art Adam RMS Prop with Nesterov Momentum (NADAM) optimizer. The proposed classification algorithm holds the potential to identify the dominant flaws in the manufacturing system responsible for leaking costs.

Effective detection of metal surface defects based on double-line laser ultrasonic with convolutional neural networks

2021 ◽  
pp. 2150263
Author(s):  
Zixi Liu ◽  
Zhengliang Hu ◽  
Longxiang Wang ◽  
Tianshi Zhou ◽  
Jintao Chen ◽  
...  
Keyword(s):  
Signal Processing ◽  
Metal Surface ◽  
Surface Defects ◽  
Ultrasonic Signal ◽  
Classification Model ◽  
Frequency Resolution ◽  
Laser Ultrasonic ◽  
Double Line ◽  
Time Frequency ◽  
Defects Classification

The time–frequency analysis by smooth Pseudo-Wigner-Ville distribution (SPWVD) is utilized for the double-line laser ultrasonic signal processing, and the effective detection of the metal surface defect is achieved. The double-line source laser is adopted for achieving more defects information. The simulation model by using finite element method is established in a steel plate with three typical metal surface defects (i.e. crack, air hole and surface scratch) in detail. Besides, in order to improve the time resolution and frequency resolution of the signal, the SPWVD method is mainly used. In addition, the deep learning defect classification model based on VGG convolutional neural network (CNN) is set up, also, the data enhancement method is adopted to extend training data and improve the defects detection properties. The results show that, for different types of metal surface defects with sub-millimeter size, the classification accuracy of crack, air holes and scratch surface are 94.6%, 94% and 94.6%, respectively. The SPWVD and CNN algorithm for processing the laser ultrasonic signal and defects classification supplies a useful way to get the defect information, which is helpful for the ultrasonic signal processing and material evaluation.

scholarly journals An Intelligent Classification Model for Surface Defects on Cement Concrete Bridges

2020 ◽  
Vol 10 (3) ◽  
pp. 972 ◽  
Author(s):  
Jinsong Zhu ◽  
Jinbo Song
Keyword(s):  
Neural Network ◽  
Surface Defects ◽  
Concrete Bridges ◽  
Classification Model ◽  
Jiangxi Province ◽  
Support Vector ◽  
Detection Accuracy ◽  
Cement Concrete ◽  
Softmax Classifier ◽  
Road Inspection

This paper mainly improves the visual geometry group network-16 (VGG-16), which is a classic convolutional neural network (CNN), to classify the surface defects on cement concrete bridges in an accurate manner. Specifically, the number of fully connected layers was reduced by one, and the Softmax classifier was replaced with a Softmax classification layer with seven defect tags. The weight parameters of convolutional and pooling layers were shared in the pre-trained model, and the rectified linear unit (ReLU) function was taken as the activation function. The original images were collected by a road inspection vehicle driving across bridges on national and provincial highways in Jiangxi Province, China. The images on surface defects of cement concrete bridges were selected, and divided into a training set and a test set, and preprocessed through morphology-based weight adaptive denoising. To verify its performance, the improved VGG-16 was compared with traditional shallow neural networks (NNs) like the backpropagation neural network (BPNN), support vector machine (SVM), and deep CNNs like AlexNet, GoogLeNet, and ResNet on the same sample dataset of surface defects on cement concrete bridges. Judging by mean detection accuracy and top-5 accuracy, our model outperformed all the contrastive methods, and accurately differentiated between images with seven classes of defects such as normal, cracks, fracturing, plate fracturing, corner rupturing, edge/corner exfoliation, skeleton exposure, and repairs. The results indicate that our model can effectively extract the multi-layer features from surface defect images, which highlights the edges and textures. The research findings shed important new light on the detection of surface defects and classification of defect images.

An Extraction Method Based on Invariance Geometric Feature

2015 ◽  
Vol 713-715 ◽  
pp. 1570-1573
Author(s):  
Rong Fen Gong ◽  
Mao Xiang Chu ◽  
Yong Hui Yang
Keyword(s):  
Extraction Method ◽  
Steel Plate ◽  
The Other ◽  
Geometric Feature ◽  
Geometric Features ◽  
The Novel ◽  
Statistical Features ◽  
Plate Surface ◽  
Shape Features ◽  
Defects Classification

An extraction method based on invariance geometric feature is proposed in this paper. This method extracts two types of feature from the object in an image. One type is five invariance statistical features of edge distance. The other is two invariance shape features: rectangular similarity feature and circular similarity feature. Moreover, this proposed method is used to extract defect features for steel plate surface. Its performance is tested in scale and rotation invariance and defects classification. Experimental results show that the novel geometric features have the ability of invariance and can improve the accuracy of classification.

Analysis and Study of the Steel Plate Surface Defects Image Processing and the Fractal Dimension Characteristics

2012 ◽  
Vol 482-484 ◽  
pp. 1773-1776
Author(s):  
Xuan Wang ◽  
Wei Liu ◽  
Hui Cao ◽  
Dong Ping Ma
Keyword(s):  
Image Processing ◽  
Fractal Dimension ◽  
Surface Defect ◽  
Surface Defects ◽  
Steel Plate ◽  
Image Feature ◽  
Plate Surface ◽  
Surface Image ◽  
Surface Defect Detection ◽  
Image Pattern

Steel surface defect detection is the key point of this research. The paper mainly focuses on the image processing and image feature extraction of the steel plate surface. The paper also focuses on the calculating procedure and results of the fractal dimension in different defects images. It can be concluded from the results of the study, fractal dimension of the defect images becomes an important feature of the steel plate surface image pattern recognition.

A Calibration Method for Binocular Vision System Based on CPSO-BP Netural Network

2014 ◽  
Vol 490-491 ◽  
pp. 1686-1691
Author(s):  
Jun Zhou ◽  
Tao Xia ◽  
Ting Ting Wang ◽  
Hua Li Li ◽  
Yu Ping Fu
Keyword(s):  
Neural Network ◽  
Binocular Vision ◽  
Bp Neural Network ◽  
Vision System ◽  
Back Propagation ◽  
Calibration Method ◽  
Image Feature ◽  
Training Set ◽  
Swarm Optimization ◽  
Binocular Vision System

This paper presents a new calibration method for binocular vision system, based on CPSO-BP neural network. Firstly, the training set of the back propagation (BP) neural network is formed by the image feature point extracted from the binocular vision system. Then the cooperate particle swarm optimization (CPSO) algorithm is introduced to optimize the weights of the BP neural network, making the network with a stronger ability of the global optimization. Experimental results demonstrate that the proposed CPSO-BP-based algorithm has a higher calibration precision than the traditional BP-based calibration method.

Hailstone Classifier Based on Back Propagation Neural Network

2013 ◽  
Vol 433-435 ◽  
pp. 685-690
Author(s):  
Xiang Yang Liu ◽  
Hui Song Wan ◽  
Yuan Yuan Zhang ◽  
Shu Ming Jiang
Keyword(s):  
Neural Network ◽  
Image Processing ◽  
Bp Neural Network ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Image Feature ◽  
Radar Image ◽  
Data Support ◽  
Color Texture

The Back Propagation (BP) neural network was used for the construction of the hailstone classifier. Firstly, the database of the radar image feature was constructed. Through the image processing, the color, texture, shape and other dimensional features should be extracted and saved as the characteristic database to provide data support for the follow-up work. Secondly, Through the BP neural network, a machine for hail classifications can be built to achieve the hail samples auto-classification.

Recognition and Classification of Hot Strip Surface Defect Based on Binary Tree SVM

2012 ◽  
Vol 538-541 ◽  
pp. 427-430 ◽  
Author(s):  
An Na Wang ◽  
Chao Hu ◽  
Chang Liang Xue ◽  
Hong Rui Zhang
Keyword(s):  
Neural Network ◽  
Bp Neural Network ◽  
Binary Tree ◽  
Surface Defect ◽  
Surface Defects ◽  
Defect Classification ◽  
Strip Surface ◽  
Hot Strip ◽  
Classification Time

The paper presents a new method which uses Binary Tree SVM in the automatic classification of surface defects for hot strip. Two types of Binary Tree SVMs are applied in defect classification. Compared with BP neural network and one-against-one SVM, the algorithm adopted in the paper greatly improved the accuracy of classification and decreased the classification time.

Sign in / Sign up

Export Citation Format

Share Document