Title: Plant growth prediction from time series data using ResNet-attention based spatial-channel network

Authors: S. Narayanan; D. Sridevi

Addresses: Department of Information Technology, SRM Valliammai Engineering College, SRM Nagar, Kattankulathur, Chengalpattu 603203, Tamilnadu, India ' Department of Computer Applications, SRM Valliammai Engineering College, SRM Nagar, Kattankulathur, Chengalpattu 603203, Tamilnadu, India

Abstract: The natural way of plant growth is a time consuming process and slows down the experimentation of phenotyping. Artificial intelligence (AI)-based models provide an automatic way and overcoming phenotyping challenges. This work introduces an enhanced deep learning (DL) model for predicting the growth of plants by making segmentation ground truth into the future. Initially, the entire annotations are automatic and generate images with XML-based annotations. Then, dual tree complex wavelet transform (DT-CWT) is provided to the original data for reducing the overfitting and reducing noise. The DL model ResNet-attention based spatial-channel module (R-At-SCM) is used for extracting features and forecasting future image frames. The experimentation is carried out on two benchmark datasets. The results depict better performance in terms of the error measures like MSE, RMSE, MAE and MAPE values. This model is well suited and trainable for various types of plants and mutations.

Keywords: phenotyping; plant growth; future image frames; deep learning; DL; time series.

DOI: 10.1504/IJMC.2026.150491

International Journal of Mobile Communications, 2026 Vol.27 No.1, pp.93 - 110

Received: 23 Feb 2023
Accepted: 12 Mar 2024
Published online: 15 Dec 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article