Leaf classification in sunflower crops by computer vision and neural networks

• Published in: Computers and electronics in agriculture 2011-08, Vol.78 (1), p.9-18
• Main Authors: Arribas, Juan Ignacio, Sánchez-Ferrero, Gonzalo V., Ruiz-Ruiz, Gonzalo, Gómez-Gil, Jaime
• Format: Article
• Language: English
• Subjects: Agronomy. Soil science and plant productions; Algorithms; Biological and medical sciences; Classification; Computer vision; Crops; Electronics; Fundamental and applied biological sciences. Psychology; Learning machines; Leaves; Model selection; Neural networks; Sunflower; Sunflowers
• ISSN: 0168-1699; 1872-7107
• DOI: 10.1016/j.compag.2011.05.007

► An automatic machine learning classification of sunflower crops from weeds system is proposed. ► Five highly discriminative morphological leaf features optimally selected. ► Optimal features: perimeter, area, major ellipse axis, minor ellipse axis and logarithm of the height to width ratio. ► Results are promising in terms of both the Correct Classification Rate and the area under the receiver operation curve. ► Potential use in mechanical weeding and application of selective herbicide combined with a plant row algorithm. In this article, we present an automatic leaves image classification system for sunflower crops using neural networks, which could be used in selective herbicide applications. The system is comprised of four main stages. First, a segmentation based on rgb color space is performed. Second, many different features are detected and then extracted from the segmented image. Third, the most discriminable set of features are selected. Finally, the Generalized Softmax Perceptron (GSP) neural network architecture is used in conjunction with the recently proposed Posterior Probability Model Selection (PPMS) algorithm for complexity selection in order to select the leaves in an image and then classify them either as sunflower or non-sunflower. The experimental results show that the proposed system achieves a high level of accuracy with only five selected discriminative features obtaining an average Correct Classification Rate of 85% and an area under the receiver operation curve over 90%, for the test set.