Open AccessEnvironmental Greening And Tree Health Maintenance Based on Color Histogram
Author(s) -
Junlin Xie,
Ruiping Wang,
Shanpeng Ma,
Xu Han,
Qirui Yu,
Huijun Yang,
Xin Wang
Publication year - 2019
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/300/3/032112
Subject(s) - histogram , forest health , artificial intelligence , tree (set theory) , segmentation , matching (statistics) , computer science , decision tree , contrast (vision) , pattern recognition (psychology) , mathematics , statistics , image (mathematics) , biology , agroforestry , mathematical analysis
Due to man-made damage, plant diseases and insect pests, environmental greening, ecological environment and other reasons, ancient and famous trees have been severely damaged. The protection of ancient trees requires diagnosis of their health and biological and chemical treatment measures. Aiming at the inconvenience, high cost and low degree of automation of manual methods for health diagnosis of ancient trees, the solution and overall architecture of a tree health diagnosis system based on OpenCV and genetic algorithm are analyzed and designed. Firstly, the leaf image mode is trained, and the color histogram is used to distinguish the leaf density and color difference between different individuals, and the fitting index is used to obtain more accurate diagnostic classification standards. Then, the matching value is calculated by genetic algorithm training through the H average value in HSVHSV and the H average value of healthy ancient trees in the histogram to obtain the matching grade. Then, the tree image is read, artificial segmentation is carried out to obtain the quasi-healthy crown score, weighted summation is carried out to obtain the leaf score, and the health grade is separated according to the ratio. Finally, the two are weighted to obtain the final health grade. The experimental results show that the system can accurately and effectively distinguish different leaf health grades according to the difference of leaf color and density, and provide quantitative data for health diagnosis research of ancient and famous trees.