Abstract: Leaf recognition is an important step for plant computerized identification, especially for field living plants. Previous researches were mainly focused on leaf recognition by utilizing the peripheral contour of the leaf while ignoring the leaf venation that actually contains important genetic information.Conventional thresholding-based methods cannot extract the information accurately due to high diversity of leaf veins. In this paper, an approach based on artificial neural network learning is proposed to extract leaf venation. Ten features including edge gradients, local contrast and statistical features are extracted from a window centered at the image pixel and used to train a neural network classifier. Compared with conventional thresholding-based methods, the trained neural network is capable of extracting more accurate modality of leaf venation for subsequent leaf recognition.