Mduma Neema, Mayo Flavia
Nelson Mandela African Institution of Science and Technology, Box 447 Tengeru, Arusha, Tanzania.
Data Brief. 2024 Mar 23;54:110359. doi: 10.1016/j.dib.2024.110359. eCollection 2024 Jun.
Maize Lethal Necrosis (MLN) and Maize Streak Virus (MSV) are among maize diseases which affect productivity in Tanzania and Africa at large. These diseases can be detected early for timely interventions and minimal losses. Machine learning (ML) has emerged as a powerful tool for automated diseases detection, offering several advantages over traditional methods. This article presents the updated dataset of 9356 imagery maize leaves to assist researchers in developing technological solutions for addressing crop diseases. The high-resolution imagery data presented in this dataset were captured using smartphone cameras in farm fields which were not selected in the previously published dataset. Also, data collection was taken in the range of three months from November 2022 to January 2023 to incorporate farming season not covered in the previously published dataset. The presented dataset can be used by researchers in the field of Artificial Intelligence (AI) to develop ML solutions and eliminate the need of manual inspection and reduce human bias. Developing ML solutions require large amount of data therefore, the updated and previously published datasets can be combined to accommodate diverse and wider applicability.
玉米致死坏死病（MLN）和玉米条纹病毒（MSV）是影响坦桑尼亚乃至整个非洲玉米产量的病害。这些病害可以早期检测出来，以便及时采取干预措施并减少损失。机器学习（ML）已成为自动病害检测的强大工具，与传统方法相比具有诸多优势。本文展示了包含9356张玉米叶片图像的更新数据集，以协助研究人员开发应对作物病害的技术解决方案。该数据集中呈现的高分辨率图像数据是使用智能手机相机在农田中采集的，这些农田未被选入之前发布的数据集中。此外，数据收集时间为2022年11月至2023年1月这三个月期间，以纳入之前发布的数据集中未涵盖的种植季节。该数据集可供人工智能（AI）领域的研究人员用于开发机器学习解决方案，从而无需人工检查并减少人为偏差。开发机器学习解决方案需要大量数据，因此，可以将更新后的数据集与之前发布的数据集相结合，以实现更广泛的适用性。
