BananaSet: A dataset of banana varieties in Bangladesh

Abstract

This article introduces a primary dataset sourced from diverse marketplaces in Bangladesh, encompassing six distinct banana varieties predominantly consumed locally. The dataset comprises the following banana types: Shagor, Shabri, Champa, Anaji, Deshi, and Bichi. High-resolution images of bananas from each category were acquired using a smartphone camera. A total of 1166 images were captured but did not maintain a uniform distribution. Only the augmented data has 1000 images per category which is a total of 6000 images. The proposed dataset exhibits substantial potential for impact and utility, offering a range of attributes, including but not limited to the representation of six diverse banana varieties, each possessing unique flavors and holding promise for various applications within the agriculture and food manufacturing industries.

Keywords

• Banana Types
• Image Dataset
• Classification
• Image Processing
• Fruit Classification
• Deep Learning