A new approach for accurate classification of hyperspectral images using Virtual Sample Generation by Concurrent Self-Organizing Maps

Neagoe, Victor-Emil; Ciotec, Adrian-Dumitru

doi:10.1109/igarss.2013.6721339

Cited by 15 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dehghani and Zheng [25] constructed virtual samples by distribution patterns available in low-dimensional subspaces.Iqbal [26] trained using the minimum classification error (LGM-MCE) of geometric edges, and directly generated virtual samples by losing smoothness in the LGM-MCE framework. Neagoe et al [27] used Virtual Sample Generation by Concurrent Self-Organizing Maps (VSG-CSOM) to increase the size of the training set. Bishop [28] added a small amount of noise interference to the sample set, that is, regularize the model, which can improve the generalization ability of the model.…”

Section: Introductionmentioning

confidence: 99%

Non-contact inspection method for surface roughness on small samples

Yan

Zhao

et al. 2023

Surf. Topogr.: Metrol. Prop.

View full text Add to dashboard Cite

This study aims to improve the non-contact measurement accuracy of roughness of small samples. Therefore, machine vision and virtual sample generation technology are used to detect the roughness of small sample bearing steel (GCr15) in this study. The surface roughness of different specimens is tested with a contact roughness detector. Image acquisition is carried out on the specimen, histogram equalization image enhancement preprocessing is carried out on the image, and sample capacity expansion is carried out on the basis of data enhancement. On this basis, the gray-level co-occurrence matrix (GLCM) is used to extract the image texture features, and a linear regression model of roughness and Energy, Entropy, Homogeneity, Contrast and Correlation is established. Artificial neural networks (ANNs) is used to classify and predict their surface roughness, train the original samples and virtual samples respectively, and compare their prediction accuracy. The results show that the precision of the virtual samples based on resampling and singular value decomposition(SVD) are 10.13% and 21.21% higher than that of the original samples, respectively, and the average error between the predicted value and the measured value of visual roughness is 3.4%. Therefore, the machine vision and virtual sample generation technology are combined to achieve acceptable surface roughness detection of small samples, which provides a theoretical basis for online detection in gear processing.

show abstract

Section: Introductionmentioning

confidence: 99%

Non-contact inspection method for surface roughness on small samples

Yan

Zhao

et al. 2023

Surf. Topogr.: Metrol. Prop.

View full text Add to dashboard Cite

show abstract

“…In order to improve training set quality, we use the innovative idea to build a training set composed by virtual samples only that completely substitute the input original samples [12]. The structure of the VTDG-CSOM is shown in Fig.…”

Section: Support Vector Machine (Svm) Classifier Using An Improved Trmentioning

confidence: 99%

“…Training of each of the Self-Organizing Map modules SOM (k) (k=1,…, M). One uses the SOM unsupervised algorithm [11], [12], [13] to train each of the SOM modules. Namely, each SOM module has been trained only with the samples having the same label with the neural module label.…”

Section: Support Vector Machine (Svm) Classifier Using An Improved Trmentioning

confidence: 99%

“…Second processing step means object feature selection performed firstly using Gabor filtering [4], [5], [8] and refined by the second step based on the application of principal component analysis (PCA). The fourth processing stage of the method corresponds to support vector machine (SVM) classification using an improved training based on virtual training data generation (VTDG) by concurrent self-organizing maps (CSOM) [12], [14]. The proposed method is applied for military ground vehicle recognition in aerial images, being experimented for Moving and Stationary Target Acquisition and Recognition (MSTAR) public release database.…”

Section: Introductionmentioning

confidence: 99%

“…The area of Automatic Target Recognition (ATR) for SAR imagery is an ongoing research in many branches of the military and large research institutions [6], [7], [16], [17], [20]. On the other side, there has been an increasing interest in using artificial neural networks (ANN) for image processing and pattern recognition [1], [2], [7], [11], [12], [13], [14], [15], [17]. A typical target recognition system consists of a detection module (filtering and segmentation) and a recognition module (feature selection and classification) [1].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An Advanced Neural Network-Based Approach for Military Ground Vehicle Recognition in Sar Aerial Imagery

Neagoe¹,

Carata²,

Ciotec³

2016

AFASES 2016

Self Cite

View full text Add to dashboard Cite

The paper presents a novel neural network approach for automatic target recognition (ATR) in the synthetic aperture radar (SAR) aerial imagery; this is applied to identify military ground vehicles. The proposed ATR algorithm consists of a processing cascade with the following stages: (a) object detection using a pulse-coupled neural network (PCNN) segmentation module; (b) a first feature selection module using Gabor filtering (GF); (c) a second feature selection module using principal component analysis (PCA); (d) a support vector machine (SVM) classifier improved by using virtual training data generation (VTDG) with concurrent self-organization maps (CSOM). The proposed model has been applied for the recognition of three classes of military ground vehicles of the former Soviet

show abstract

CNN Hyperspectral Image Classification Using Training Sample Augmentation with Generative Adversarial Networks

Neagoe

Diaconescu

2020

2020 13th International Conference on Communications (COMM)

View full text Add to dashboard Cite

A new approach for accurate classification of hyperspectral images using Virtual Sample Generation by Concurrent Self-Organizing Maps

Cited by 15 publications

References 11 publications

Non-contact inspection method for surface roughness on small samples

Non-contact inspection method for surface roughness on small samples

An Advanced Neural Network-Based Approach for Military Ground Vehicle Recognition in Sar Aerial Imagery

CNN Hyperspectral Image Classification Using Training Sample Augmentation with Generative Adversarial Networks

Contact Info

Product

Resources

About