Neural network classifier of hyperspectral images of skin pathologies

The paper presents results of using a neural network classifier to analyze images malignant skin lesions obtained hyper-spectral camera. Using three-block VGG architecture, we conducted the classification set two-dimensional melanoma, papilloma and basal cell carcinoma, in range 530 – 570 600 606 nm, characterized by highest absorption melanin hemoglobin. sufficiency inclusion training limited spectral is analyzed. show significant prospects algorithms for processing hyperspectral data pathologies. With relatively small used study, accuracy three types neoplasms was as high 96 %.