Image-Based Field Monitoring of Sugar Beet Cercospora Leaf Spot Using Robust Template Matching and Pattern Recognition.

We present a novel image algorithm using template matching and pattern recognition frameworks for monitoring Cercospora leaf spot (CLS) development on the scale of a single sugar beet leaf under field-conditions. Due to the variety and complexity (i.e., illumination, non-rigid leaf movements and growth, soil background, etc.) under field conditions, it is a significant challenge to achieve continuous and robust foliar disease observation. We propose a novel and compact algorithm, composed of dual frameworks and a post-processing step. The algorithm has continuous and highly discriminative capabilities for observing the progress of disease in a single leaf from plant-level time sequence images. The first framework is based on a robust template matching that employs orientation code matching (OCM), which implements successive tracking of a single leaf from a beet plant against severe illumination changes and non-rigid plant movements. The second framework uses a pattern recognition method of a support vector machine (SVM) for achieving further disease classification from a clutter field background. Prior to SVM implementation, we propose a three feature combination of L*, a*, and Entropy×Density, which has strong discrimination power to classify CLS disease from the clutter image containing sandy soil, leaves, leaf stalks, and specular reflection. Additionally, post-processing is introduced to filter false positive noise to enhance the precision of the classification. Our field experiment results demonstrate the feasibility and applicability of the proposed algorithm for disease monitoring under real field conditions. Meanwhile, comparative results with other conventional matching methods and feature combinations show the effectiveness of our proposed algorithm in both foliage tracking and disease classification.