Apple leaf disease segmentation and grading system based on attention and feature fusion

• Main Authors: CHENG JIXIANG, LI ZHIDAN, SONG ZUJIAN, WU YUAN, JIN JOO
• Format: Patent
• Language: Chinese; English
• Published: 12.07.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses an apple leaf disease segmentation and grading system based on attention and feature fusion, and the system comprises the following steps: collecting images of apple leaves infected with different disease types under natural conditions, forming an apple disease leaf source data set, and carrying out the data amplification and preprocessing, so as to obtain a disease leaf segmentation data set and a disease segmentation data set; respectively importing the diseased leaf segmentation data set and the disease segmentation data set into the constructed parallel boundary guidance network and the constructed parallel disease fusion network for training to obtain a diseased leaf segmentation model and a disease segmentation model; importing the to-be-segmented image into the diseased leaf segmentation model to obtain a diseased leaf segmentation result; importing the diseased leaf segmentation result into a disease segmentation model to obtain a disease segmentation result; the apple leaf dis