Combining the use of CNN classification and strength-driven compression for the robust identification of bacterial species on hyperspectral culture plate images.
Alberto SignoroniMattia SavardiMario PezzoniFabrizio GuerriniSimone ArrigoniGiovanni TurraPublished in: IET Comput. Vis. (2018)
Keyphrases
- hyperspectral images
- hyperspectral
- hyperspectral data
- image data
- remotely sensed
- spectral bands
- remote sensing
- image classification
- hyperspectral imagery
- plant species
- multispectral
- hyperspectral image classification
- multi band
- satellite images
- hyperspectral imaging
- infrared
- change detection
- image analysis
- pixel classification
- data compression
- target detection
- spatial resolution
- image features
- image compression
- high spatial resolution
- spectral data
- pattern recognition
- image retrieval
- remote sensing images
- hyperspectral remote sensing
- random projections
- feature space
- license plate
- information content
- multispectral images
- high quality
- wavelet transform
- feature selection
- principal components
- normal form
- detection rate