Plants are one of the long lasting species on Earth and are used for various purposes such as medicine, food and organics. Apex point and base point, coined here, are very important points of leaf facilitating extraction of various leaf shape based features. The main challenge in automatic plant recognition system is automatically identifying apex-base points and removing petiole of the leaf. Here, we propose a novel methodology, using straight line interpolation and bisection, for automatically identifying leaf apex point and base point. The coordinates of apex point and base point are successfully retrieved and the methodology is rotation and scale invariant. Finally, a logistic regression is used for classification. The proposed methodology is tested in several datasets like Leaflia, flavia and the results are satisfactory. © 2018, Springer Science+Business Media, LLC, part of Springer Nature.