IDDMSLD: An image dataset for detecting Malabar spinach leaf diseases.

Abstract

Agriculture has always played a vital role in the economic development of Bangladesh. In Agriculture, leaf diseases have become an issue because they can lead to a major drop in both quality and quantity of crops. Therefore, leveraging technology to automatically detect diseases on leaves plays an important role in farming. Malabar Spinach (Basella alba) is a well-known, widely grown leafy vegetable, which is valued for its nutritional benefits. However, there is almost no dataset that can aid in identifying diseases affecting this important crop, which often leads to decreased quality as well as financial drawback. This lack of resources makes it difficult for farmers to recognize and manage common diseases. Our purpose is to solve this problem by creating a unique dataset of Bangladesh's Malabar Spinach leaves that will ease agricultural management and disease detection. Our dataset contains both healthy and diseased samples, categorised into four common ailments: Anthracnose, Bacterial Spot, Downy Mildew, and Pest Damage. We collected 3,006 original images in total. Images were collected from various locations in Bangladesh, including Mirpur, Savar, Sirajganj and Gazipur, with photographs taken under natural lighting conditions at different times of the day. This dataset will help the researchers for further research on Malabar Spinach disease detection implementing various efficient computational models and applying advanced machine learning techniques.

Code and data availability

The paper is a Data in Brief article describing a public Mendeley Data repository of 3,006 Malabar spinach leaf images (healthy plus four disease classes) used for plant disease phenotyping. The dataset is paper-specific, publicly deposited, and directly actionable via the stated direct URL.