PaddyVarietyBD: Classifying Paddy Variations of Bangladesh with a Novel Image Dataset

Paddy Varieties Image Dataset Description:
This dataset provides a comprehensive collection of 35 paddy variety images, compiled from the Bangladesh Rice Research Institute (BRRI) and the Bangladesh Institute of Nuclear Agriculture (BINA). The purpose of this dataset is to aid in the identification and classification of distinct paddy varieties based on their unique features, supporting both agricultural research and machine learning applications.
Paddy Varieties Included:
The dataset covers a total of 35 paddy varieties, as follows:
• BD Varieties: BD30, BD33, BD34, BD49, BD51, BD52, BD56, BD57, BD70, BD72, BD75, BD76, BD79, BD85, BD87, BD91, BD93, BD95
• Binadhan Varieties: Binadhan7, Binadhan8, Binadhan10, Binadhan11, Binadhan12, Binadhan14, Binadhan16, Binadhan17, Binadhan19, Binadhan20, Binadhan21, Binadhan23, Binadhan24, Binadhan25, Binadhan26
• BR Varieties: BR22, BR23
Dataset Breakdown:
The dataset consists of two main components: Original Images and Augmented Images. These are divided into individual folders for each of the 35 paddy varieties, providing easy access for further analysis and use.
• Original Image Dataset:
o Total number of images: 14,000
o Image format: JPG
o Image resolution: 640x480 pixels
• Augmented Image Dataset:
o Total number of images: 56,000
o Image format: JPG
o Image resolution: 640x480 pixels
o Augmentation techniques: Rotation (90° left, 90° right, 180°), flipping, and shearing
Image Capture and Methodology
Images for this dataset were captured using high-resolution digital microscope cameras, with magnifications of 1600x and 1000x. These images were taken in September 2023, ensuring high-quality close-up views of the paddy varieties.
Dataset Storage and Access:
The dataset is stored in two separate zip files:
• Original.zip contains the 14,000 original images.
• Augmented.zip contains the 56,000 augmented images.
Each zip file includes 35 individual folders, with each folder corresponding to a specific paddy variety.

稻米品种图像数据集描述：本数据集汇聚了来自孟加拉国水稻研究所（BRRI）和孟加拉国核农业研究所（BINA）的35种稻米品种图像，旨在辅助稻米品种的识别与分类，其独特特征为农业研究与机器学习应用提供支持。数据集包含以下稻米品种：

BD品种：BD30、BD33、BD34、BD49、BD51、BD52、BD56、BD57、BD70、BD72、BD75、BD76、BD79、BD85、BD87、BD91、BD93、BD95
Binadhan品种：Binadhan7、Binadhan8、Binadhan10、Binadhan11、Binadhan12、Binadhan14、Binadhan16、Binadhan17、Binadhan19、Binadhan20、Binadhan21、Binadhan23、Binadhan24、Binadhan25、Binadhan26
BR品种：BR22、BR23

数据集组成：该数据集主要由原始图像和增强图像两部分构成，分别针对35种稻米品种分别建立独立文件夹，便于进一步的分析与应用。

原始图像数据集：
总数：14,000张
图像格式：JPG
图像分辨率：640x480像素

增强图像数据集：
总数：56,000张
图像格式：JPG
图像分辨率：640x480像素
增强技术：旋转（左90°、右90°、180°）、翻转和剪切

图像捕捉与方法：本数据集图像采用高分辨率数字显微镜相机捕捉，具有1600x和1000x的放大倍数，拍摄时间为2023年9月，确保了稻米品种的高质量近距离观察。

数据集存储与访问：数据集存储于两个独立的ZIP文件中。
Original.zip包含14,000张原始图像。
Augmented.zip包含56,000张增强图像。每个ZIP文件包含35个独立文件夹，每个文件夹对应一种特定的稻米品种。