Transductive hyperspectral image classification: toward integrating spectral and relational features via an iterative ensemble system

Appice, Annalisa; Guccione, Pietro; Malerba, Donato

doi:10.1007/s10994-016-5559-7

Cited by 16 publications

(25 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, we intend to investigate different HS feature engineering algorithms to feed the classification stage of the learning methodology proposed in this study. In particular, we plan to explore the performance of Gabor features (Jia et al, 2015), autocorrelation features (Appice & Malerba, 2019), morphological features (Appice et al, 2016; and frequency features (Guccione et al, 2015) in the investigated HS saliency detection scenario. Finally, we plan to extend the investigation of the feasibility of a parallel strategy for implementing the proposed algorithm.…”

Section: Discussionmentioning

confidence: 99%

Leveraging colour-based pseudo-labels to supervise saliency detection in hyperspectral image datasets

et al. 2021

Self Cite

View full text Add to dashboard Cite

Saliency detection mimics the natural visual attention mechanism that identifies an imagery region to be salient when it attracts visual attention more than the background. This image analysis task covers many important applications in several fields such as military science, ocean research, resources exploration, disaster and land-use monitoring tasks. Despite hundreds of models have been proposed for saliency detection in colour images, there is still a large room for improving saliency detection performances in hyperspectral imaging analysis. In the present study, an ensemble learning methodology for saliency detection in hyperspectral imagery datasets is presented. It enhances saliency assignments yielded through a robust colour-based technique with new saliency information extracted by taking advantage of the abundance of spectral information on multiple hyperspectral images. The experiments performed with the proposed methodology provide encouraging results, also compared to several competitors.

show abstract

Section: Discussionmentioning

confidence: 99%

Leveraging colour-based pseudo-labels to supervise saliency detection in hyperspectral image datasets

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…The purpose of the logarithmic transformation (log (I/R)) of the reciprocal of spectral reflectance is to enhance the spectral difference value of the lower reflectivity (Brzska and Tomczyk, 2015). Comparing the original spectral reflectance curve, aronia melanocarpa leaves generally has low spectral reflectance in the visible band, and the band is the reflection and absorption of leaves with pigment (Appice and Guccione, 2016). Therefore, the reciprocal logarithmic transformation of the original spectral reflectance can effectively enhance the spectral difference of these bands, and at the same time, it can also weaken the random error caused by different illumination (Goh and Youn, 2016).…”

Section: Imaging Spectrum Acquisition and Reflectance Conversionmentioning

confidence: 99%

Hyper Spectral Image Characteristics of Aronia Melanocarpa Leaves under Saline Alkali Stress

2020

Global NEST: The International Journal

View full text Add to dashboard Cite

<p>Objective: To obtain the characteristics analysis results of aronia melanocarpa leaves under saline alkali stress state and improve yield and quality of aronia melanocarpa. Methods: Using hyper spectral imaging system to obtain hyper spectral images of aronia melanocarpa leaves under saline alkali stress. The clear hyper spectral image is obtained by the conversion of reflectance, spectral envelope removal, and spectral denoising, and final hyper spectral image is obtained by the normalization of a clearer hyper spectral image. Results: Spectral information of aronia melanocarpa leaves under slight saline alkali stress: in the visible band, leaf reflectance is lower than leaf nutrient rich stress condition; in the near infrared band, the nutrient rich leaves was significantly higher than that of stress leaves. Spectral information of leaves under moderate saline alkali stress: spectral reflectance of different lesion spots for a same leaf in 550-680nm band is: severe > moderate > slight > normal, in the near infrared band is on the contrary; spectral reflectance for different lesion grades in 550-680 nm, severe lesion leaves have highest reflectance, and normal leaves have the lowest reflectance. Under severe saline alkali stress, the leaf spectral information is: there is no significant difference in leaf spectral reflectance between the attachment aphid and the damaged leaf, but the difference is obvious for normal leaves in the band of 450-500 nm, 560-680 nm and 750-900 nm. Comparative results analysis for the three of saline alkali stress degree is: the near infrared band of 560-680 nm and the visible band of 780-900 nm is the sensitive band for the diagnosis of three kinds of stress; slight saline alkali stress has the most significant differences at 550 nm, and 780-900 nm; severe saline alkali stress has at 680 nm and 780-900 nm. Conclusion: The proposed method can analyze hyper spectral image characteristics of aronia melanocarpa leaves under different saline alkali stress the condition of is a kind of plant leaves, which is an efficient method for the analysis of characteristics of plant leaves under saline alkali stress.</p>

show abstract

“…The discriminative method [24] uses the maximum interval algorithm to train the learning decision boundary of the labeled sample and unlabeled sample. The purpose of learning is to make the classification hyperplane through the low density data region, and to make the distance maximum between the classification hyperplane and the nearest sample.…”

Section: Related Work 21 Semi-supervised Learningmentioning

confidence: 99%

Semi-supervised Classification Based Mixed Sampling for Imbalanced Data

Zhao

Liu

2019

Open Physics

View full text Add to dashboard Cite

In practical application, there are a large amount of imbalanced data containing only a small number of labeled data. In order to improve the classification performance of this kind of problem, this paper proposes a semi-supervised learning algorithm based on mixed sampling for imbalanced data classification (S2MAID), which combines semi-supervised learning, over sampling, under sampling and ensemble learning. Firstly, a kind of under sampling algorithm UD-density is provided to select samples with high information content from majority class set for semi-supervised learning. Secondly, a safe supervised-learning method is used to mark unlabeled sample and expand the labeled sample. Thirdly, a kind of over sampling algorithm SMOTE-density is provided to make the imbalanced data set become balance set. Fourthly, an ensemble technology is used to generate a strong classifier. Finally, the experiment is carried out on imbalanced data with containing only a few labeled samples, and semi-supervised learning process is simulated. The proposed S2MAID is verified and the experimental result shows that the proposed S2MAID has a better classification performance.

show abstract

Transductive hyperspectral image classification: toward integrating spectral and relational features via an iterative ensemble system

Cited by 16 publications

References 73 publications

Leveraging colour-based pseudo-labels to supervise saliency detection in hyperspectral image datasets

Leveraging colour-based pseudo-labels to supervise saliency detection in hyperspectral image datasets

Hyper Spectral Image Characteristics of Aronia Melanocarpa Leaves under Saline Alkali Stress

Semi-supervised Classification Based Mixed Sampling for Imbalanced Data

Contact Info

Product

Resources

About