VegSeedsBD: A Comprehensive Image Dataset of Vegetable Seeds.

Published: 6 January 2025| Version 1 | DOI: 10.17632/dtpzbwwpm7.1
Contributors:
Md Hasanul Ferdaus,
,
,

Description

Type of data: 3456 x 4608 px images of vegetable seeds. Data format: JEPG. Dataset Contents: Original (raw) images of different varieties of vegetable seeds from single seed and bulk seeds perspectives. Number of classes: Fifteen (15) varieties of common vegetable seeds – (1) Bitter Gourd, (2) Bottle Gourd, (3) Corn, (4) Cucumber, (5) Hyacinth Bean, (6) Malabar Spinach, (7) Okra, (8) Papaya, (9) Pea, (10) Pumpkin, (11) Ribbed Gourd, (12) Snake Gourd, (13) Sponge Gourd, (14) String Bean, and (15) Wax Gourd. Total number of images: 4,500. Distribution of instances: Number of images in VegSeedsBD dataset – (1) Bitter Gourd: Single Seed = 200 and Bulk Seeds = 100. (2) Bottle Gourd: Single Seed = 200 and Bulk Seeds = 100. (3) Corn: Single Seed = 200 and Bulk Seeds = 100. (4) Cucumber: Single Seed = 200 and Bulk Seeds = 100. (5) Hyacinth Bean: Single Seed = 200 and Bulk Seeds = 100. (6) Malabar Spinach: Single Seed = 200 and Bulk Seeds = 100. (7) Okra: Single Seed = 200 and Bulk Seeds = 100. (8) Papaya: Single Seed = 200 and Bulk Seeds = 100. (9) Pea: Single Seed = 200 and Bulk Seeds = 100. (10) Pumpkin: Single Seed = 200 and Bulk Seeds = 100. (11) Ribbed Gourd: Single Seed = 200 and Bulk Seeds = 100. (12) Snake Gourd: Single Seed = 200 and Bulk Seeds = 100. (13) Sponge Gourd: Single Seed = 200 and Bulk Seeds = 100. (14) String Bean: Single Seed = 200 and Bulk Seeds = 100. (15) Wax Gourd: Single Seed = 200 and Bulk Seeds = 100. Dataset size: The total size of the dataset is 3.56 GB and the compressed ZIP file size is 3.36 GB. Data acquisition process: Images of vegetable seeds are captured using a high-definition smartphone camera from different angles and the two perspectives of single seed and bulk seeds. Data source location: Agro-industry grade seeds are collected from agricultural seed stores in Bangladesh. Where applicable: Training and evaluating machine learning and deep learning models to distinguish varieties of common vegetable seeds in Bangladesh to support automated identification systems of various vegetable seeds.

Files

Institutions

East West University, Bangladesh Ministry of Agriculture

Categories

Ecology, Agronomy, Crop Science, Computer Vision, Environmental Science, Environmental Monitoring, Plant Biology, Botany, Image Processing, Seed Biology, Object Detection, Machine Learning, Vegetable, Food Processing, Biodiversity, Plant Genetics, Image Classification, Vegetable Farming, Seed Bank, Conservation Agriculture, Agrochemical, Agriculture Industry, Seed, Seeders, Precision Agriculture, Sowing, Plant Breeding, Seed Production, Deep Learning, Agricultural Biotechnology, Agriculture

Licence