Open AccessCuticle segmentation based on visual saliency
Author(s) -
Wenxia Zhang,
Chunguang Wang,
Haichao Wang,
Xiaofei Yin
Publication year - 2021
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 - 10.1088/1755-1315/647/1/012183
Subject(s) - artificial intelligence , segmentation , cuticle (hair) , luminance , computer vision , computer science , pattern recognition (psychology) , image segmentation , rgb color model , mathematics , biology , anatomy
The cuticle of plant is an important tissue to keep water and resist the invasion of ultraviolet radiation and so on,it is also one of the most important indexes of plant drought resistance.In this paper,we introduce a method combined salient region detection and threshold segmentation for cuticle segmentation.The method used in the article has the following steps:Firstly,we transformed the slice image from the original RGB space into CIELAB space.Then according to the principle of visual significance,visually salient slice image regions were detected using color and luminance features.Finally,we obtained the segmentation results by threshold segmentation algorithm.We tested this method by 30 slice images of Cleistogenes Songorica.We got the average value of FPR(False Positive Rate),FNR(False Negative Rate)and GSA(Global Segmentation Accuracy) were 0.75%,7.19% and 98.43%.The method in this paper is more suitable for cuticle segmentation of Cleistogenes Songorica leaf slices and can also provide a reference for other plant leaf slice images.