Abstract:Wheat stripe rust caused by Puccinia striiformis f. sp. tritici, is one of the most important and devastating diseases in wheat production. Identification and classification of wheat stripe rust plays an important role in high-quality production of wheat, which helps to quantitatively assess the level of wheat stripe rust severity in the field to make strategies to achieve effective control for wheat stripe rust in early. Currently, estimation disease severity of wheat stripe rust is mainly relied on naked-eye observation according to the manual field investigation. However, this method is labour-intensive, time-consuming, besides requiring workers with high professional knowledge. In order to quickly and accurately evaluate the disease level of wheat stripe rust, a novel grading method of disease severity of wheat stripe rust based on hyperspectral imaging technology was proposed. Firstly, hyperspectral images of 320 infected at different levels and 40 healthy wheat leaf samples were captured by a HyperSIS hyperspectral system covering the visible and near-infrared region (400~1000Nm). Secondly, via the analysis of spectral reflectance of leaf and background regions, there were obvious differences in spectral reflectance at the 555Nm wavelength. Therefore, the image of the 555Nm wavelength was named the feature image, which was manipulated by threshold segmentation to obtain a mask image. The logical and operation was conducted by using the original hyperspectral image and mask image to remove the background information. Thirdly, the principal component analysis (PCA) method was used for the dimension reduction of hyperspectral images. The operation results showed that the second principal component image (PC2) can significantly identify the stripe rust spot area and healthy area. On this basis, stripe rust spots area was efficiently segmented by using an Otsu method. Finally, the degree of the disease severity of wheat stripe rust was graded according to the proportion of stripe rust spots area on a whole leaf. To verify the effectiveness of the proposed method, a total of 270 leaf samples were collected for the performance evaluation. Experimental results showed that 265 samples could be accurately classified at different disease severities of wheat stripe rust and the overall classification accuracy was 9815%. In conclusion, the experimental results indicated that the method using hyperspectral imaging technology proposed is able to satisfy the precision demand of quantitative calculation and provide a foundation for evaluating the field disease level of wheat stripe rust and a new idea for resistance identification method of wheat stripe rust.
