An incalculable leaf classes are accessible all around towards oversee gigantic substance and the advancement of quick and compelling classification strategies has transformed into a region of effective look. Herbs and shrubs are essential for environment, precise recognition and characterization that winds up essential phenomena. Characterization methodology is conveyed out through number of sub systems. A recognizable proof on categorization subject is overseen by measuring information on one kind of classes. For this framework an initial database of petiole picture is made which includes pictures of a test petiole and proportionate herb data. Basic similar feature extractions are removed utilizing picture preparing strategies. The highlights must be steady so as to make the recognizable proof framework vigorous.Accordingly the herb/shrubs perceived utilizing deep learning techniques. Study of different kinds of leafs and their recognition is done in this paper. © 2020 IEEE.