Image processing methods for quantitatively detecting soybean rust from multispectral images

摘要: Soybean rust is one of the most destructive foliar diseases soybean, and can cause significant yield loss. Timely application fungicide in early stage infection, which critically important for effective control disease, heavily dependent upon capability to quantitatively detect infection. This paper reports research outcomes from developing image processing methods detecting severity multi-spectral images. A fast manual threshold-setting method was originally developed based on HSI (Hue Saturation Intensity) colour model segmenting infected areas plant leaves. Two disease diagnostic parameters, ratio area (RIA) index (RCI), were extracted used as symptom indicators quantifying severity. To achieve automatic detection, an alternative analysing centroid leaf distribution polar coordinate system investigated. Leaf images with various levels collected analyzed. Validation results showed that capable soybean under laboratory conditions, whereas centroid-locating had potential be applied field.