Getting to know the details of plants growing around us is of great importance medicinally and economically. Conventionally, plants are categorized mainly by taxonomists through investigation of various parts of the plant. However, most of the plants can be classified based on the leaf shape and associated features. This article describes how Artificial Neural Network is used to identify plant by inputting leaf image. Compared to earlier approaches, new input features and image processing approach that matter in efficient classification in Artificial Neural Network have been introduced. Image processing techniques are used to extract leaf shape features such as aspect ratio, width ratio, apex angle, apex ratio, base angle, centroid deviation ratio, moment ratio and circularity. These extracted features are used as inputs to neural network for classifying the plants. Under the current research, 534 leaves of 20 kinds of plants were collected. Out of these, 400 leaves were trained. The 134 testing samples were recognised with 92% accuracy; even without considering types of leaf margins, vein and removal of the petiole. Software has also been developed to identify leaf automatically except two mouse clicks by the user.

Pattern Recognition Neural Network Feature Extraction Leaf Recognition Plant Identification