Efficient Diagnosis COVID-19 using Gabor and Transfer Learning

In this work, we propose a Deep Learning (DL) based approach to detect COVID-19 infection from chest CT Scans and X-rays images. The Keras-Tensorflow architecture is used with VGG16. Gabor Wavelet (GW) is used for feature extraction. The VGG16 Transfer Learning (TL) was used in the classification COVID-19 and no-COVID-19. Experiments are valued on three publicly available datasets. Our results show that the proposed approach achieves the very high performance of the VGG16 without Gabor on COVID19-CT dataset contains 3000 positive CT scans with clinical findings of COVID-19, and 3000 negative images without findings of COVID-19 with performance accuracy for the two cases X-rays et CT. It should be noted also that system using Gabor has a lot of performance regarding the speed of image processing time.