ColoredFlowersBD: A Comphrehensive Image Dataset of Colored Flowers in Bangladesh for Identification and Classification Using Machine Learning and Computer Vision

Type of data: 720 x 720 px images of colored flowers.
Data format: JEPG

Dataset contents: Original images of different varieties of colored flowers in Bangladesh from single flower and bulk-flower perspectives.

Number of classes: Thirteen colored flower varieties - (1) Chandramallika, (2) Cosmos Phul, (3) Gada, (4) Golap,
(5) Jaba, (6) Kagoj Phul, (7) Noyontara, (8) Radhachura, (9) Rangan, (10) Salvia, (11) Sandhyamani, (12) Surjomukhi, and (13) Zinnia.

Total number of images in the dataset: 7,993.

Distribution of instances:
(1) Chandramallika = 620 images in total. Single = 306, Bulk = 314.
(2) Cosmos Phul = 620 images in total. Single = 307, Bulk = 313.
(3) Gada = 617 images in total. Single = 304, Bulk = 313.
(4) Golap = 605 images in total. Single = 302, Bulk = 303.
(5) Jaba = 604 images in total. Single = 300, Bulk = 304.
(6) Kagoj Phul = 612 images in total. Single = 301, Bulk = 311.
(7) Noyontara = 609 images in total. Single = 303, Bulk = 306.
(8) Radhachura = 617 images in total. Single = 309, Bulk = 308.
(9) Rangan = 606 images in total. Single = 305, Bulk = 301.
(10) Salvia = 634 images in total. Single = 313, Bulk = 321.
(11) Sandhyamani = 615 images in total. Single = 305, Bulk = 310.
(12) Surjomukhi = 621 images in total. Single = 310, Bulk = 311.
(13) Zinnia = 613 images in total. Single = 307, Bulk = 306.

Dataset size: The total size of the dataset is 2.79 GB and the compressed ZIP file size is 2.71 GB.

Data acquisition process: Images of colored flowers are captured using a high-definition smartphone camera from different angles and two perspectives: single-flower and bulk-flower.

Data source location: Plant nurseries, local gardens, and flower shops located in different areas of Dhaka and Gazipur districts of Bangladesh.

Where applicable: Training and evaluating machine learning and deep learning models to distinguish colored flower varieties in Bangladesh to support automated identification and classification systems of various colored flowers which can be utilized in areas of computer vision, botanical research, floral biodiversity monitoring, agriculture and horticulture, environmental conservation, AI-based flower recognition, educational resources, food industry, pollination and ecology research, aesthetic and design applications.

数据类型：分辨率为720×720像素的彩色花卉图像。
数据格式：JPEG。
数据集内容：涵盖孟加拉国不同品种彩色花卉的原始图像，拍摄视角包含单朵花卉与批量花卉两种场景。
类别数量：共13个彩色花卉品种，分别为(1) Chandramallika、(2) Cosmos Phul、(3) Gada、(4) Golap、(5) Jaba、(6) Kagoj Phul、(7) Noyontara、(8) Radhachura、(9) Rangan、(10) Salvia、(11) Sandhyamani、(12) Surjomukhi、(13) Zinnia。
数据集总图像数：7993张。
样本分布：
(1) Chandramallika：总计620张，其中单视角样本306张，批量视角样本314张；
(2) Cosmos Phul：总计620张，单视角307张，批量视角313张；
(3) Gada：总计617张，单视角304张，批量视角313张；
(4) Golap：总计605张，单视角302张，批量视角303张；
(5) Jaba：总计604张，单视角300张，批量视角304张；
(6) Kagoj Phul：总计612张，单视角301张，批量视角311张；
(7) Noyontara：总计609张，单视角303张，批量视角306张；
(8) Radhachura：总计617张，单视角309张，批量视角308张；
(9) Rangan：总计606张，单视角305张，批量视角301张；
(10) Salvia：总计634张，单视角313张，批量视角321张；
(11) Sandhyamani：总计615张，单视角305张，批量视角310张；
(12) Surjomukhi：总计621张，单视角310张，批量视角311张；
(13) Zinnia：总计613张，单视角307张，批量视角306张。
数据集大小：数据集总容量为2.79 GB，压缩ZIP包容量为2.71 GB。
数据采集流程：采用高清智能手机摄像头，从不同角度拍摄单朵花卉与批量花卉两种视角的彩色花卉图像。
数据采集地点：孟加拉国达卡县与加济布尔县各区域的苗圃、本地花园及花店。
适用场景：可用于训练与评估机器学习、深度学习模型，以区分孟加拉国彩色花卉品种，支撑各类彩色花卉的自动识别与分类系统，可应用于计算机视觉、植物学研究、花卉生物多样性监测、农业与园艺、环境保护、基于人工智能的花卉识别、教育资源、食品工业、授粉与生态学研究、美学与设计应用等领域。