Wheat Leaf Rust Detection Based on Multi-Scale DWT and LCS Color Thresholding Methods

Abstract

Health, Technology, education, and food production are the four main issues facing developing nations like Pakistan, and it is undeniable that agriculture is the most important factor behind economic growth. In addition, implementing a strategy for food production is crucial for citizens to ensure their survival, and it is assumed that these initiatives will result in sufficient farm productivity. One strategy to make a field productive is to take significant care of its components, which starts with cultivating healthy plants or crops. Wheat leaf rust is a fatal condition that attacks young seedlings. It is a significant fungi disease. Leaf rust has 25% effect on the productivity of wheat. To mitigate this issue, a Multi-Scale Discrete Wavelet Transform (MsclDWT) using hybrid fusion rules method is proposed to obtain the complementary information from multiple input images. In second phase, Lab color space followed by color thresholding method is applied to detect and segment wheat leaf rust disease in wheat crop. The proposed model also computes the rust-affected area of the wheat crop, which assists the farmers in the post-medication (anti rust spray) process. The empirical results show that the proposed model achieved 97% of accuracy in rusted pixels detection and classification and outperformed the existing comparative methods.