LE–MSFE–DDNet: a defect detection network based on low-light enhancement and multi-scale feature extraction

Hu, Weihua; Wang, Tao; Wang, Yangsai; Chen, ZiYang; Huang, Guoheng

doi:10.1007/s00371-021-02210-6

Cited by 25 publications

(8 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…e segmentation accuracy rate was 98.29%, which was better than the segmentation method based on the gray level cooccurrence matrix. e authors of [17] combined the Gabor filter method with wavelet transform to identify the grayscale images of wood defects and introduced the multichannel Gabor filter, which could identify wood defects under the interference of complex background. e authors of [18] used principal component analysis to reduce the dimension of extracted features which can effectively identify and locate defects of complex inner holes.…”

Section: Research Status Of Wood Nondestructive Testingmentioning

confidence: 99%

Wood Quality Defect Detection Based on Deep Learning and Multicriteria Framework

Ping-an

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Traditional nondestructive testing technology for wood defects has a series of problems such as low identification accuracy, high cost, and cumbersome operation, and traditional testing methods cannot accurately show the specific location and size of wood internal defects; it is urgent to explore a new nondestructive testing scheme for wood defects. Aiming at this problem, this paper designs and develops an automatic detection method for wood surface defects based on deep learning algorithm and multicriteria framework. By comparing the performance of different deep learning detection methods on the data set, the advantages and disadvantages of the detection method in this paper are proved. After a series of works, such as the development and optimization of the experimental algorithm, the algorithm proposed meets the requirements in both the detection accuracy and training time.

show abstract

Section: Research Status Of Wood Nondestructive Testingmentioning

confidence: 99%

Wood Quality Defect Detection Based on Deep Learning and Multicriteria Framework

Ping-an

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…e classi cation methods based on probability statistics include hidden Markov models and SVM models. e SVM model is highly useable and generalizable in dealing with high-dimensional pattern recognition and nonlinear problems, and it is applicable to this paper for sports action classification [8]. us, SVM is selected as a classifier to classify sports actions through target detection and multifeature fusion, so as to achieve sports-assisted education.…”

Section: Related Workmentioning

confidence: 99%

Sports-Assisted Education Based on a Support Vector Machine and Genetic Algorithm

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

A sports-assisted education method based on a support vector machine (SVM) is proposed to address the problem of complex and variable sports actions leading to easy ghosting of target detection and high dimensionality of feature extraction, which reduces the low accuracy of sports action recognition. The ViBe target detection algorithm is improved by using Wronskian function and the “4-linked algorithm” seed filling algorithm, which effectively solves the ghosting problem and obtains clearer human sports targets. By using the genetic algorithm to fuse the eight-star model with sports action features extracted by the Zernike moment, redundant features are reduced and differentiability between different classes is ensured. Sports action classification was achieved by using a one-to-one construction of an SVM classifier. The results show that the proposed method can effectively recognize sports movements with an average recognition accuracy of more than 96%, which can assist physical education and has a certain practical application value.

show abstract

“…Under relatively ideal experimental conditions, the CNN-based metal defect detection algorithm is able to achieve the desired results. [1] . But the metal surface is easy to reflect, while the image captured by the camera will appear out of the low light, uneven light, these circumstances increase the difficulty of metal detection [2] .Hu [1] proposed a defect detection network (LE-MSFE-DDNet) based on low light enhancement and multiscale feature extraction to address the problem that low light and multiscale defects affect the detection rate of scratch defects.…”

Section: Introductionmentioning

confidence: 99%

“…[1] . But the metal surface is easy to reflect, while the image captured by the camera will appear out of the low light, uneven light, these circumstances increase the difficulty of metal detection [2] .Hu [1] proposed a defect detection network (LE-MSFE-DDNet) based on low light enhancement and multiscale feature extraction to address the problem that low light and multiscale defects affect the detection rate of scratch defects. An average accuracy of 94.3% was achieved on the homemade metal iron box dataset.…”

Section: Introductionmentioning

confidence: 99%

Research on AINDANE-faster R-CNN-based metal plate defect detection method

Yang¹,

Yan²,

An³

2023

Third International Conference on Digital Signal and Computer Communications (DSCC 2023)

View full text Add to dashboard Cite

To address the problem that metal reflection, ambient illumination and different defect size affect the defect detection accuracy in metal defect detection, an AINDANE-Faster R-CNN based metal flat detection method under complex illumination conditions is proposed, which firstly adopts a adaptive and integrated neighborhood dependent approach for nonlinear enhancement(AINDANE) to preprocess the defect images and improve the image brightness to highlight the detail features such as color,profile and texture of defects, and then ResNet50 network is utilized as the defect semantic feature extraction network for the Two-stage Faster R-CNN model. In addition, this paper also constructs a dataset of three defects of metal aluminum plate under low light and uneven light conditions, such as scratches, oil stains and pits, and the method achieves a mean average accuracy of 92.01% on the defect dataset. Compared to existing one-stage surface defect detection methods, the algorithm in this paper is optimal.

show abstract

LE–MSFE–DDNet: a defect detection network based on low-light enhancement and multi-scale feature extraction

Cited by 25 publications

References 42 publications

Wood Quality Defect Detection Based on Deep Learning and Multicriteria Framework

Wood Quality Defect Detection Based on Deep Learning and Multicriteria Framework

Sports-Assisted Education Based on a Support Vector Machine and Genetic Algorithm

Research on AINDANE-faster R-CNN-based metal plate defect detection method

Contact Info

Product

Resources

About