A LATE FUSION METHOD FOR MULTI-ORGAN PLANT IDENTIFICATION

Abstract

Plant identification that aims at determining the name of plant species from images of plant species’ observation. Previous studies have often focused on two organs. In this paper, a new late fusion method for multi-organ plant identification is proposed for combining two to six organs according to leaf, flower, fruit, stem, branch, entire. This method is based on combining the product rule and sum rule using weights assigned to plant organs. A deep learning method- a state of the art method- is applied for single organ identification. The experimental results have shown the effectiveness of the proposed method, it outperforms than max rule, sum rule, product rule. The results also indicate that the more organs are combined, the better the identification accuracy is. The proposed method achieves the highest accuracy of  98.8% when combining 6 organs.