Yes it is, though it is important to access the image service via a connection file that you create in the python code. Here is the essential snippet of code to use: arcpy.mapping.CreateGISServerConnectionFile("USE_GIS_SERVICES", arcpy.env.scratchFolder,
servername + ".ags", "http://" + servername + ".arcgis.com/arcgis/services",
"ARCGIS_SERVER", '', '', userName, passWord, "SAVE_USERNAME")
# set up image layer from the landscape server connection, and get its spatial reference
imageService = os.path.join(arcpy.env.scratchFolder, servername, imageservicename + ".ImageServer")
inputImageLayer = arcpy.MakeImageServerLayer_management(imageService,"imageLayer")Once you have that inputImageLayer, you can use it to get properties via the arcpy.describe function:rastSpatRef = arcpy.Describe(inputImageLayer).spatialReference
arcpy.env.cellSize = arcpy.Describe(inputImageLayer).meanCellWidth
theCellSize = arcpy.Describe(inputImageLayer).meanCellWidth
From there you can make a rasterLayer:arcpy.MakeRasterLayer_management(inputImageLayer, "temp_raster")
And that can be used in many Geoprocessing functions:arcpy.sa.Con("junk2g","temp_raster")
