Indoor Plant Varieties for Computer Vision Applications: A Diverse Image Dataset

The Indoor Plant Varieties Dataset is a valuable asset for researchers, plant care specialists, and AI developers, offering essential tools for the accurate identification and classification of common indoor plant species. This dataset supports the development of computer vision models tailored to indoor plant recognition, enabling users to create applications that simplify plant identification and care.

Collected between October 28 and November 20, 2023, the dataset consists of 1172 high-resolution images (3000x3000 pixels, JPG format) across seven plant classes: Aglaonema, Cryptanthus, Devil’s Ivy, Heartleaf Philodendron, N-Joy Pothos, Rhaphidophora, and ZZ Plant. Each class represents unique indoor plant species frequently found in homes and offices, collected from BADC (Bangladesh Agricultural Development Corporation) in Kashimpur and nearby nurseries. Images were captured using a Xiaomi M2101K61 smartphone, ensuring consistency in quality. Domain experts (agronomists from BADC) confirmed the class labels, adding reliability to the dataset. To enhance the dataset further, Data Augmentation techniques were applied, including flipping, rotation, zoom, shear, brightness adjustments, and noise reduction. This augmentation expanded the dataset to 7000 images (1000 per class), enhancing the effectiveness of deep learning models for accurate plant classification.

1. Original Data:
Number of datasets: 1172
Data format: .jpg

2. Augmented Data:
Number of datasets: 7000
Data format: .jpg

室内植物品种数据集是一笔宝贵的资源，对于研究人员、植物养护专家及AI开发者而言，该数据集提供了准确识别和分类常见室内植物种类的基本工具。本数据集支持针对室内植物识别的计算机视觉模型开发，使用户能够创建简化植物识别与养护的应用程序。数据集的收集时间介于2023年10月28日至11月20日之间，包含1172张高分辨率图像（3000x3000像素，JPG格式），跨越七个植物类别：龟背竹、 Cryptanthus、常春藤、心叶蔓绿绒、N-Joy 铁线莲、麒麟叶及 ZZ 植物。每个类别代表了一种独特的室内植物种类，常见于家庭与办公室，数据来源于卡西姆普尔的孟加拉国农业发展公司（BADC）及其周边苗圃。图像均采用小米 M2101K61 智能手机拍摄，以确保图像质量的一致性。领域专家（来自BADC的农学家）确认了类别标签，为数据集增添了可靠性。为进一步丰富数据集，采用了数据增强技术，包括翻转、旋转、缩放、剪切、亮度调整和噪声减少等，使数据集扩展至7000张图像（每类1000张），从而增强了深度学习模型在准确植物分类方面的有效性。

1. 原始数据：
数据集数量：1172
数据格式：.jpg

2. 增强数据：
数据集数量：7000
数据格式：.jpg