Deep learning-based strawberry ripeness prediction and growth analysis in vertical farming from time-series image data

Fecha de publicación: 01/09/2023
Fuente: ISHS (International Society for Horticultural Science)
Post date: Friday 1 September 2023
ISHS Secretariat

Zhixian Lin is a Ph.D. candidate in the College of Biosystems Engineering and Food Science at Zhejiang University, Hangzhou, China. He works under the supervision of Prof. Dr. Tao Lin. His research focuses on data-driven based modeling for crop growth in plant factories. Strawberries have high consumer demand due to their palatability and nutritional benefits. However, strawberry farming can be challenging as it is susceptible to pests, diseases, and environmental fluctuations. Growing strawberries in vertical farming with LED lighting and precise environmental control can provide higher yields and better-quality fruit. Predicting strawberry ripeness and monitoring growth is critical to the precision management of strawberry cultivation. Detailed information on the strawberry ripeness stage, such as the optimal harvest date, can help farmers to plan and organize the timing and sequence of production activities to optimize the use of space, labor and resources, while meeting the market demand and maximizing profitability. In this study, an automatic approach based on deep learning was developed to predict strawberry ripeness and monitor strawberry growth in vertical farming. A time-series image dataset was constructed by experimenting with monitoring the strawberry fruit development after the de-greening fruit stage. The study proposed a dual-branch attention fusion (DBAF) model for strawberry fruit ripeness prediction using time-series image data, which comprises two parallel branches for feature extraction and incorporates an attention mechanism to enhance the presentation of features from in-field images. A segmentation network was trained to extract the boundary and geometric traits of fruits, enabling the generation of growth curves based on fruit size. The automatic approach based on deep learning showed superior performance and stability in the in-field scenario. The research highlights the potential of deep learning-based approaches for the precision management of strawberry cultivation in vertical farming.
Zhixian Lin won the ISHS Young Minds Award for the best poster presentation at the VertiFarm2023: II International Workshop on Vertical Farming in China in May 2023.
Zhixian Lin, College of Biosystems Engineering and Food Science, Zhejiang University, 866 Yuhangtang Road, Hangzhou, Zhejiang, China, e-mail:
The article is available in Chronica HorticulturaeTags: strawberryCategories: Young Minds Award Winners