Optical spectroscopy applied in agriculture remains a challenge in developing countries for a better monitoring of agricultural productivity. The phytopathological approach by fluorescence spectroscopy makes it possible to detect viral and bacterial diseases in particular in real time in order to early follow the evolution of pathogens and consider preventing their spread. To illustrate this work, a multi-spectral microscope has been designed specifically for the representation of fluorescence images on a portion (leaf blade) of a plant without using a fluorescent molecule. Through the obtained microscopic images, the corresponding fluorescence spectrum has been reconstructed in order to distinguish a healthy leaf from a diseased leaf in terms of contrast. The main contribution is the development of the method of fluorescence spectroscopy by multispectral imaging at the microscopic scale for the early diagnosis of bacterial wilt due to Ralstonia solanacearum in tomato plants, two symptoms caused by the African cassava mosaic virus in a cassava field, and Cucumber mosaic virus following its inoculation to some zucchini plants without the use of a fluorescent marker. The results obtained show the effectiveness of the proposed technique, in order to facilitate the identification of these pathogens present in tomato, cassava and cucumber plants respectively by biological analysis laboratories

Keywords : Multispectral Imaging, Pathogens, Optical spectroscopy, Fluorescence