The invention discloses an
image processing-based intelligent detection method and
system for the degree of a grape air burning
disease, and the method comprises the steps: automatically collecting a visible light image of a grape
plant, segmenting and recognizing a leaf and fruit region, extracting the color and texture features, and comparing the color and texture features with a pre-stored air burning
disease symptom feature
library, thereby recognizing an early damaged abnormal region. By calculating the proportion and morphological characteristics of affected areas, a single leaf affected index and a fruit affected grade are generated, the quantitative
disease degree grade of the whole
plant is obtained through comprehensive evaluation, and a disease development trend obtained through historical
image sequence analysis and real-time temperature and
humidity data collected by an environment sensor are fused. The early warning of the disease
diffusion risk is realized, and the targeted grading farming operation suggestions are generated according to the disease grade, the
spatial distribution characteristics and the prediction result. According to the invention, diagnosis and early warning of the grape air cautery are realized, the problems of lagging and strong subjectivity of traditional manual inspection are effectively overcome, and decision support is provided for accurate prevention and control.