The present application relates to the technical field of intelligent image recognition, and particularly relates to a fruit and vegetable
disease image recognition grading early warning method and
system, comprising: acquiring
time sequence images of the same target leaf, and registering the
time sequence images based on the contour features and main
vein endpoint features of the leaf; extracting the leaf
vein structure in the registered images, establishing a leaf
vein conduction network containing nodes and paths according to the connection relationship and thickness of the leaf vein, and dividing the leaf vein levels. The present application realizes early identification and grading early warning of fruit and vegetable diseases by continuously observing the same target leaf in
time sequence, combining the structural features of the leaf vein conduction network and the
color gradient variation law along the vein. The biological characteristics of
disease spreading along the leaf vein first are utilized, and the leaf vein
system is taken as the main carrier for
disease monitoring, so that the problem that the prior art does not distinguish the different propagation characteristics of diseases in the leaf vein and the leaf tissue is solved, and the
extremely light and slightly
colored areas formed around the leaf vein in the early stage of the disease can be effectively captured.