Paper Title
Real-Time Huanglongbing Detection System using Pattern Recognition Techniques over Resource Constrained Devices

Abstract
Huanglongbing (citrus greening) is a disease caused due to bacteria in citrus tree leaves, causing considerable damage to citrus fruits worldwide. Efficient and fast methods for this disease detection must be carried out to minimize the losses occurring due to Huanglongbing disease. In this paper, we propose pattern recognition based system to detect Hua-nglongbing from one of the citrus fruit leaves i.e. orange leaves. Our complete proposed system contains five vital steps: image acquisition, image segmentation, feature extraction, classification and real-time implementation on resource con-strained devices. We use three key features based on shape, color and texture which includes, asymmetry index of shape features, Cb component of YCbCr color space and local binary patterns of leaf texture. We evaluate our method on two dif-ferent classifier named, Support Vector Machine (SVM) and K-Nearest Neighbor (KNN). By using our method we achieved an accuracy of 90.5% by correctly classifying normal and diseased leaf images. Keywords - Citrus Greening, Pattern Recognition, Classification, Segmentation.