Multispectral Image under Tissue Classification Algorithm in Screening of Cervical Cancer

Wang, Pei; Wang, Shuwei; Zhang, Yuan; Xiao-yan, Duan

doi:10.1155/2022/9048123

Cited by 2 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MSX technology addresses the limitations of standard thermal imaging by embedding high-contrast, high-resolution details from an onboard digital camera onto thermal images in real-time. This integration allows for a more intuitive understanding of the thermal data, as the thermal image is overlaid with clear visual details [68,69]. The key advantage of MSX over standard thermal imaging is the enhanced feature recognition and clarity it provides.…”

Section: Msx Technologymentioning

confidence: 99%

Data Fusion and Ensemble Learning for Advanced Anomaly Detection Using Multi-Spectral RGB and Thermal Imaging of Small Wind Turbine Blades

Memari,

Shekaramiz,

Masoum

et al. 2024

Energies

View full text Add to dashboard Cite

This paper introduces an innovative approach to Wind Turbine Blade (WTB) inspection through the synergistic use of thermal and RGB imaging, coupled with advanced deep learning techniques. We curated a unique dataset of 1000 thermal images of healthy and faulty blades using a FLIR C5 Compact Thermal Camera, which is equipped with Multi-Spectral Dynamic Imaging technology for enhanced imaging. This paper focuses on evaluating 35 deep learning classifiers, with a standout ensemble model combining Vision Transformer (ViT) and DenseNet161, achieving a remarkable 100% accuracy on the dataset. This model demonstrates the exceptional potential of deep learning in thermal diagnostic applications, particularly in predictive maintenance within the renewable energy sector. Our findings underscore the synergistic combination of ViT’s global feature analysis and DenseNet161’s dense connectivity, highlighting the importance of controlled environments and sophisticated preprocessing for accurate thermal image capture. This research contributes significantly to the field by providing a comprehensive dataset and demonstrating the efficacy of several deep learning models in ensuring the operational efficiency and reliability of wind turbines.

show abstract

Section: Msx Technologymentioning

confidence: 99%

Data Fusion and Ensemble Learning for Advanced Anomaly Detection Using Multi-Spectral RGB and Thermal Imaging of Small Wind Turbine Blades

Memari,

Shekaramiz,

Masoum

et al. 2024

Energies

View full text Add to dashboard Cite

show abstract

“…Multispectral and hyperspectral images are helpful in many fields, e.g. in geology (Ninomiya & Fu, 2019), agriculture (Hassan-Esfahani et al, 2014), food industry (Benouis et al, 2021), healthcare (Wang et al, 2022) and cultural heritage (Colantonio et al, 2018). Research in the field of multi-and hyperspectral image capture is useful and current, as well as, due to the affordability of high-resolution commercial cameras, also research that uses various techniques of converting RGB to multidimensional spectral image space.…”

mentioning

confidence: 99%

Improved reconstruction of the reflectance spectra from RGB readings using two instead of one digital camera

Lazar¹,

Hladnik²

2022

Proceedings - The Eleventh International Symposium GRID 2022

View full text Add to dashboard Cite

The colour of an observed object can be described in many different manners, and the description by its reflectance provides the unambiguous colour representation. The reflectance description can be acquired by expensive multispectral cameras or, e.g., with time-sequential multispectral illumination. In our experiment, we propose that under the condition of constant and uniform illumination, the reflectance can be deduced from the object's RGB camera readouts, captured alongside the set of colour patches with known spectral characteristics. Translation from a colour description in RGB space into reflectance spectra, independent of illuminant and camera sensor characteristics, was performed with the help of an artificial neural network (ANN). In our study, the hypothesis was proposed that the ANN's performance of reflectance reconstruction can be enhanced by employing richer learning datasets using RGB input sets of two cameras instead of just one. Additional second camera information would be adequate only if the equivalent channels of cameras used are linearly independent. A quantitative measure of nonlinearity (QMoN), which is the metric primarily developed for use in chemistry, was employed to estimate the degree of nonlinearity. Additional attention was paid to ANN training, structure and learning set sizes. Two ANN training algorithms have been utilised, a faster GPU executed standard backpropagation and an order of magnitude slower CPU based, but with significantly better convergence Levenberg-Marquardt training algorithm. The number of neurons in the hidden ANN layer varied from the size of the input layer to a number greater than the number in the output layer. The complete set of colour samples was divided into five learning sets of different sizes, with the smaller sets being subsets of the larger ones. To assess performances of the resulting ANNs, mean squared error, the goodness of fit and colour differences calculated from original and reconstructed reflectances assuming several standard illuminations have been compared. A noticeable reflectance performance improvement has been found by using two cameras, even though the cameras' equivalent channels exerted only small degrees of nonlinearity.

show abstract

Multispectral Image under Tissue Classification Algorithm in Screening of Cervical Cancer

Cited by 2 publications

References 21 publications

Data Fusion and Ensemble Learning for Advanced Anomaly Detection Using Multi-Spectral RGB and Thermal Imaging of Small Wind Turbine Blades

Data Fusion and Ensemble Learning for Advanced Anomaly Detection Using Multi-Spectral RGB and Thermal Imaging of Small Wind Turbine Blades

Improved reconstruction of the reflectance spectra from RGB readings using two instead of one digital camera

Contact Info

Product

Resources

About