Open AccessObject Detection under Natural Illumination Conditions using Superpixels and Local Binary Pattern Feature
Author(s) -
Chuanyuan Zhao,
Xiangjuan Li,
Yue Zhao,
Xiaotao Shi
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1237/3/032027
Subject(s) - artificial intelligence , local binary patterns , pattern recognition (psychology) , computer vision , computer science , classifier (uml) , object detection , feature (linguistics) , texture (cosmology) , feature extraction , object (grammar) , binary number , mathematics , image (mathematics) , histogram , linguistics , philosophy , arithmetic
This paper proposes a novel object detection method under natural illumination conditions in which a set of features, texture features and Local Binary Pattern (LBP) features, are extracted from the images acquired by a colour camera. The goal of this study is to develop a robust and fast algorithm to detect immature green citrus fruit in individual trees from colour images acquired under natural outdoor conditions. Colour and shape features are used to remove the complex background as much as possible. Since leaves share much more similarities with green citrus fruit in colour and to some extent in the shape, texture features are used for citrus detection. Statistical features, Tamura features and LBP features are used to build the K NN classifier. Experimental results show that the proposed approach provides fairly good object detection performance and confirms an efficient way for outdoor green citrus detection.