Hi Fares. I have seen studies using GE to extract the landuse/landcover. The potential to use GE images is there, especially for pattern recognition. So you can do it yourself too and maybe apply object-based classification method.
But take note that GE images have limitations. One is its RGB renderings and no NIR band. Therefore, it may not be good for landuse classes that require high spectral signature, such as grassland and woodland.
Fares LAOUACHERIA Have you considered Google Earth Engine? There are existing landuse/landcover maps available, or you can create your own with a small bit of code. There are existing apps integrated with GEE to also download/visualize LULC across the globe.
Mr Fares LAOUACHERIA , The satellite images from USGS earth explorer are very useful and freely accessible, the images have a resolution cell size of 15 m
The treatment of satellite images can be done with the help of ARC GIS software from ESRI. The study area extracted from each satellite image and then classified according to the maximum likelihood supervised classification method. the google earth images help for assessing the accuracy of the classified images