Estimation of the amount of pear pollen based on flowering stage detection using deep learning

Keita Endo,Takefumi Hiraguri,Tomotaka Kimura,Hiroyuki Shimizu,Tomohito Shimada,Akane Shibasaki,Chisa Suzuki,Ryota Fujinuma,Yoshihiro Takemura

DOI: https://doi.org/10.1038/s41598-024-63611-w

IF: 4.6

2024-06-09

Scientific Reports

Abstract:Pear pollination is performed by artificial pollination because the pollination rate through insect pollination is not stable. Pollen must be collected to secure sufficient pollen for artificial pollination. However, recently, collecting sufficient amounts of pollen in Japan has become difficult, resulting in increased imports from overseas. To solve this problem, improving the efficiency of pollen collection and strengthening the domestic supply and demand system is necessary. In this study, we proposed an Artificial Intelligence (AI)-based method to estimate the amount of pear pollen. The proposed method used a deep learning-based object detection algorithm, You Only Look Once (YOLO), to classify and detect flower shapes in five stages, from bud to flowering, and to estimate the pollen amount. In this study, the performance of the proposed method was discussed by analyzing the accuracy and error of classification for multiple flower varieties. Although this study only discussed the performance of estimating the amount of pollen collected, in the future, we aim to establish a technique for estimating the time of maximum pollen collection using the method proposed in this study.

multidisciplinary sciences

What problem does this paper attempt to address?

The problem this paper attempts to address is improving the efficiency of pear flower pollen collection to ensure the amount of pollen needed for artificial pollination, thereby stabilizing fruit yield. In recent years, it has been difficult to collect enough pollen in Japan, leading to an increase in imported pollen, rising costs, and unstable supply. To solve this problem, the study proposes an artificial intelligence (AI)-based method that uses the object detection algorithm YOLO (You Only Look Once) in deep learning to classify and detect different flowering stages of pear flowers and estimate the amount of pollen accordingly. By analyzing the classification accuracy and errors of different varieties of pear flowers, the performance of the proposed method is discussed. In the future, the research team also plans to use this method to establish a technical system for predicting the optimal pollen collection time.