Classification of apricot leafs via convolutional neural network

Plants are very significant for human life. They are used in various fields such as food sector, industry and medical. It is known that there are 310000 – 450.000 kinds of plants in the world. As the days pass, unknown species are coming to light. Nowadays traditional Methods are generally used while classifying the plants. Creating appropriate, practical and automatic system to introduce the plant is a practical study in terms of classifying, understanding and managing the plants. Introducing the plant on account of its leaf is one of the methods which are used for classifying plants. To protect and identify the plants, creating a database for each species is an important development for specialists. To create a digital plant classification system, designing automatic leaf recognition system which uses computer vision applications and image processing techniques will provide velocity and productivity the process. In this paper, 7 different apricot images were used. These species are apikoz şalak, Çataloğlu, Çekirge İz, Hacıkızı, Hırmanlı, Paviot ve Tokatoğlu Erzincan. These appricot species are classified by using deep learning methods. Appricot species are classified with %91,34 ±0,77 success by using convolutional neural network which is among deep learning methods. By using local receptive field extreme learning machine which is another deep learning method, apricot species are classified with %97,26±0,95 success.