Type of Data: 256x256 px Banana images. Data Format: JPEG Contents of the Dataset: Banana varieties and ripeness stages.

Number of Classes: (1) Four most popular banana varieties in Bangladesh - Bangla Kola, Chompa Kola, Sabri Kola, and Sagor Kola, and (2) Four ripeness stages - Green, Semi-ripe, Ripe, and Overripe.

Number of Images: (1) Total Original (Raw) images of banana varieties = 2,471, Augmented to 7,413 images, and (2) Total Original (Raw) images of ripeness stages = 820, Augmented to 2,457 images.

Distribution of Instances: (1) Original (Raw) images in each class of banana varieties: Bangla Kola = 444, Champa Kola = 994, Sabri Kola = 509, and Sagor Kola = 524; (2) Augmented images in each class of banana varieties: Bangla Kola = 1332, Champa Kola = 2,982, Sabri Kola = 1,527, Sagor Kola = 1,572; (3) Original (Raw) images in each class of Ripeness stages: Green = 212, Semi-ripe = 204, Ripe = 201, and Overripe = 203; (4) Augmented images in each class of Ripeness stages: Green = 636, Semi-ripe = 609, Ripe = 603, and Overripe = 609.

Dataset Size: (1) Total size of the Original (Raw) banana varieties dataset = 17.2 MB; (2) Total size of the Augmented banana varieties dataset = 78.5 MB; (3) Total size of the Original (Raw) ripeness stages dataset = 5.55 MB; and (4) Total size of the Augmented ripeness stages dataset = 25.2 MB.

Data Acquisition Process: Images of bananas are captured using high-quality smartphone cameras. Data Source Location: Local banana wholesale markets and retail fruit shops located in different places in Bangladesh. Where Applicable: The dataset presents considerable potential for fostering innovation and developing automated, efficient processes across various industries, such as precision agriculture, food processing, and supply chain management. By training Machine Learning (ML) and Deep Learning (DL) models on this dataset, it becomes possible to accurately classify banana varieties and evaluate their ripeness stages. These models can be utilized to design automated systems for determining ideal harvest times, establishing banana quality control standards, analyzing consumer preferences to guide product development and marketing strategies, and streamlining the supply chain through enhanced harvesting, sorting, packaging, and inventory management. Additionally, researchers focused on advancing Computer Vision technologies in food and agricultural sciences will find the dataset valuable for improving precision farming and food processing methods. As a result, the dataset offers substantial potential for automating banana production and processing, cutting labor costs, and boosting overall operational efficiency.

Note: This dataset is an updated version of "BananaImageBD: An Extensive Image Dataset of Common Bangladeshi Banana Varieties with Different Ripeness Levels", DOI: 10.17632/ptfscwtnyz.1

Dataset Description:

The Banana Ripeness Classification Dataset contains 13,478 images of bananas at various stages of ripeness. This dataset is designed to help in developing models that can classify bananas based on their ripeness, providing insights for agricultural applications.

The dataset is divided into three sets:

Training Set: 11,793 images (87% of the total dataset) used for model training.

Validation Set: 1,123 images (8% of the total dataset) used for model validation and hyperparameter tuning.

Test Set: 562 images (4% of the total dataset) used for final model evaluation to assess generalization performance.

Preprocessing:

Auto-Orient: Applied Resize: Stretch to 416x416 Modify Classes: 2 remapped, 0 dropped

Augmentation:

Outputs per training example: 3 Flip: Horizontal, Vertical 90° Rotate: Clockwise, Counter-Clockwise, Upside Down Crop: 0% Minimum Zoom, 20% Maximum Zoom Rotation: Between -15° and +15° Hue: Between -10° and +10° Saturation: Between -10% and +10% Brightness: Between -10% and +10% Exposure: Between -10% and +10% Blur: Up to 1px

Labeled Banana images suitable for training and evaluating computer vision and deep learning models.

Real Dataset Banana Images The real dataset developed consists of 3,495 images of Cavendish bananas.

Synthetic Dataset Banana Images The synthetic dataset developed consists of 161,280 images of Cavendish bananas.

Nano-Banana Generated Images

9,457 high-quality images generated using the Nano-Banana model (Google Gemini 2.5 Flash Image Preview).

  Dataset Overview

Total Images: 9,457 images Generation Method: Nano-Banana (Google Gemini 2.5 Flash Image Preview) Storage Format: Optimized binary (Hugging Face Image type) File Organization: Normal large parquet files (not chunked) License: MIT

  Schema

Column Type Description

id int Unique identifier

image Image… See the full description on the dataset page: https://huggingface.co/datasets/bitmind/nano-banana.

Here are a few use cases for this project:

1. Grocery Inventory Management: Retailers could use the "Banana" computer vision model to more effectively manage their fruit inventory. The ability to identify Unripe and Ripe Bananas automatically could help optimize sell-through rates and minimize waste.

2. Agricultural Quality Control: Banana farmers and distributors could use this model to analyze the ripeness of their crops. By identifying unripe and ripe bananas, they can better plan their distribution and minimize potential losses.

3. Health and Nutrition Apps: Developers of food tracking or health apps could implement this model to help users identify their banana's ripeness levels. This could be useful for people needing to control their sugar intake as ripeness influences sugar content.

4. Educational Tools: This model could be integrated into educational software or applications teaching about nutrition, agriculture, or biology. The categorization can help students understand various stages of fruit maturation.

5. Augmented Reality Games and Apps: This model could be used in AR games or learning apps where users interact with everyday objects. For example, it could be used to identify bananas and trigger specific interactions or learning experiences.

1. Bananas are not only nutritious but also delicious. Both raw and ripe bananas are beneficial for health. It is a good source of potassium, vitamin C, vitamin B6, and dietary fiber, according to nutritional guidelines.
2. Potassium in bananas helps regulate blood pressure and supports overall heart health. Its' dietary fiber aids digestion and helps in maintaining gastrointestinal regularity. Natural sugars like glucose, fructose, and sucrose in bananas provide quick and sustained energy.
3. With a low glycemic index, bananas assist in keeping blood sugar levels stable. The fiber content promotes a feeling of fullness, aiding in weight management. It contains serotonin, a neurotransmitter that contributes to a good mood and stress reduction.
4. Vitamin A in bananas supports healthy vision, and they may contribute to age-related macular degeneration prevention. The presence of magnesium and vitamin B6 in bananas helps maintain strong bones. It also contains prebiotic fiber that supports the growth of beneficial gut bacteria, contributing to a healthy digestive system.
5. The prevalence of various diseases associated with bananas highlights the need for proper measures to be taken to mitigate their impact, which may lead to a reduction in banana production on a large scale. Therefore, adopting preventive measures as soon as symptoms of diseases are observed is crucial.
6. Diseases in crops pose a significant challenge to agricultural production, impacting the quality and productivity of the crops. Due to environmental factors, diseases in crops can have adverse effects on both yield and quality. For instance, banana diseases can negatively impact the yield and quality of bananas, leading to significant economic losses for farmers. Traditional methods of identifying and managing crop diseases are often time-consuming, labor-intensive, inefficient, and subjective.
7. Such classification and identification tasks have been a promising area for computer vision in recent years.
8. A large dataset of seven different banana classes—Healthy Leaf, Bract Mosaic Virus Disease, Black Sigatoka, Insect Pest Diseases, Moko Disease, Panama Disease, and Yellow Sigatoka—is shown in order to create machine vision-based algorithms.
9. There are 408 images of bananas in all, taken in actual fields. Then, in order to expand the number of data points, shifting, flipping, zooming, shearing, brightness enhancement, and rotation techniques are used to create a total of 2856 augmented images from these original images.

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F16850153%2Ff1f7dd7feebca89cefc3f0b9d74fd982%2Fmovies-minions-bananas-wallpaper-preview.jpg?generation=1732031695984694&alt=media" alt=""> This comprehensive banana dataset captures important information about banana samples from different regions and varieties. The key attributes are: ample_id: A unique identifier assigned to each banana sample in the dataset. This allows the samples to be tracked and referenced uniquely.

variety: The cultivar or breed of banana, such as Cavendish, Red Dacca, or Lady Finger. Knowing the specific banana variety provides context about the sample's physical characteristics and growing conditions.

region: The geographic origin of the banana, such as Ecuador, Philippines, or Costa Rica. The region can influence factors like climate, soil, and growing practices that impact the banana's qualities.

quality_score: A numerical score, likely on a scale of 1-4 that rates the overall quality of the banana sample. This could encompass factors like appearance, texture, and lack of defects.

quality_category: A text label that categorizes the quality score into broader groupings like "Excellent" etc This provides an easier-to-understand quality assessment.

ripeness_index: A numerical index representing the ripeness level of the banana, potentially ranging from 1 (green/unripe) to 10 (overripe). This quantifies the maturity of the fruit.

ripeness_category: A text label like "Green", "Yellow", "Ripe", or "Overripe" that corresponds to the ripeness index. This gives a clear, qualitative ripeness classification.

sugar_content_brix: The sugar content of the banana measured in degrees Brix. This is a common way to assess the sweetness and quality of the fruit.

firmness_kgf: The firmness of the banana measured in kilograms-force. This indicates the texture and maturity of the sample.

length_cm: The physical length of the banana in centimeters. This size metric can vary by variety and growing conditions.

Dataset merupakan dataset langsung dari sumbernya asli, terdiri dari tingkat kematangan pisang jenis Cavendish dan Ambon. Pisang Cavendish berkontribusi pada tingkat kematangan setengah matang, matang, dan terlalu matang. Pisang Ambon digunakan sebagai pengganti untuk tingkat kematangan mentah dan juga untuk menambahkan warna yang sesuai pada tingkat kematangan terlalu matang. Dataset terdiri dari 491 citra mentah, 606 citra setengah matang, 294 citra matang, dan 474 citra terlalu matang. Semua citra pada dataset memiliki ukuran (resolusi) 224 x 224 piksel. Dataset yang totalnya terdiri dari 1.865 citra dibagi menjadi dua bagian, yaitu 1.490 citra data latih dan 375 citra data test, masing-masing sesuai dengan 80% dan 20% dari total dataset.

Each folder inside the banana image dataset folder contains a banana tier with its six side images. The number corresponds to its sampleID. There are 194 banana tier subjects with six images corresponding to its side views: front, left, right, back, top and bottom. A total of 1,164 banana images.

The banana_features.csv shows the extracted features: RGB (red, green, blue) color values, the image side view, class and finger size. There are four class values with its respective numerical value (1: ‘extra class’, 2: ‘class II’, 3: ‘class I’, 4: ‘reject’) and there are 65, 49, 30 and 50 samples per class, respectively.

Note that the finger size feature was taken by manually measuring the length of the top middle finger of a banana tier ("hand") in millimeter (mm). This specific finger was assumed to have a regular size compared to other fingers.

This dataset can be used for machine learning classification of banana based on its features. In addition, the available images can be used for automating the manual measurement of the top middle finger size.

Banana Dataset

## Overview

Banana Dataset is a dataset for object detection tasks - it contains Fruits annotations for 550 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

Banana Disease Recognition Dataset Introduction: The Banana Disease Recognition Dataset is a collection of images aimed at facilitating research and development in the field of banana disease detection and classification. This dataset contains a total of 777 images, with 700 images designated for training and 77 images for testing. The dataset encompasses six classes of banana diseases and one class for healthy banana leaves. Each class consists of 100 training images and 11 testing images… See the full description on the dataset page: https://huggingface.co/datasets/as-cle-bert/banana-disease-classification.

AtharavJadhav/Banana dataset hosted on Hugging Face and contributed by the HF Datasets community

Banana cultivation is frequently challenged by various diseases that severely impact yield. These diseases detrimentally affect banana plants, causing growth inhibition, diminished fruit production, and even plant fatality. The consequences are far-reaching, as afflicted plants struggle to yield marketable fruit, leading to financial setbacks for banana growers and the potential to disrupt the global banana supply.

Banana Leaf Spot Diseases in the Dataset

The dataset comprises a diverse collection of images showcasing three prominent banana leaf spot diseases, namely: 1. Sigatoka 2. Cordana 3. Pestalotiopsis Additionally, images depicting healthy banana leaves are incorporated for comprehensive analysis.

How the Images Were Captured?

The images were captured using smartphone cameras in the banana fields of Bangabandhu Sheikh Mujibur Rahman Agricultural University, Bangladesh, and nearby banana fields in June 2021. All images were labelled by an expert plant pathologist.

Dataset Description and Organization

The dataset is constituted of two subsets. a) Original Set: This comprises 937 RGB images, divided into 4 classes and provided in JPG format. b) Augmented Set: This set supplements the original collection with 400 images per class, culminating in a total of 1600 images. Employing augmentation techniques, such as Gaussian blur, horizontal flip, cropping, linear contrast adjustment, shear, translation, and rotational shear, we enhanced the dataset's diversity. All images have a standard resolution of 224 x 224 pixels.

Publications Related to this Dataset

Please consider reading the following research articles based on this dataset: 1. BananaSqueezeNet: A very fast, lightweight convolutional neural network for the diagnosis of three prominent banana leaf diseases 2. Bananalsd: A Banana Leaf Images Dataset for Classification of Banana Leaf Diseases Using Machine Learning

Citation

If you're using this dataset for your work, please cite the following articles: @article{bhuiyan2023bananasqueezenet, title={BananaSqueezeNet: A very fast, lightweight convolutional neural network for the diagnosis of three prominent banana leaf diseases}, author={Bhuiyan, Md Abdullahil Baki and Abdullah, Hasan Muhammad and Arman, Shifat E and Rahman, Sayed Saminur and Al Mahmud, Kaies}, journal={Smart Agricultural Technology}, volume={4}, pages={100214}, year={2023}, publisher={Elsevier} } @article{arman2023bananalsd, title={BananaLSD: A banana leaf images dataset for classification of banana leaf diseases using machine learning}, author={Arman, Shifat E and Bhuiyan, Md Abdullahil Baki and Abdullah, Hasan Muhammad and Islam, Shariful and Chowdhury, Tahsin Tanha and Hossain, Md Arban}, journal={Data in Brief}, pages={109608}, year={2023}, publisher={Elsevier} }

The dataset is a collection of images representing various conditions of bananas, specifically aimed at training machine learning models for image classification or augmentation tasks. The dataset is organized into multiple subfolders, each representing a different condition or class of bananas. These classes include:

Healthy Bananas Bananas with Fusarium Wilt Bananas with Natural Leaf Death Bananas with Rhizome Root Issues Each image in the dataset is initially stored in its respective class folder and typically contains a banana or bananas under different conditions, viewed from different angles, and possibly with varying levels of resolution or lighting.

The dataset is then processed for various machine learning tasks like classification, detection, or augmentation. Specifically, this dataset is aimed at providing a variety of augmented images to ensure a more robust training set, which is critical for improving the generalization performance of machine learning models.

Note for Researchers Using the dataset

This dataset was created by Shuvo Kumar Basak. If you use this dataset for your research or academic purposes, please ensure to cite this dataset appropriately. If you have published your research using this dataset, please share a link to your paper. Good Luck.

🍌 Banalyzer - Banana Ripeness Classification Dataset A deep learning image dataset for classifying bananas into 4 ripeness stages: Unripe, Ripe, Overripe, and Rotten. Built using transfer learning with MobileNetV2 for efficient training and deployment.

📦 What's Included Image Dataset: Organized training and test sets for all 4 ripeness classes Training Script (train.py): MobileNetV2 transfer learning implementation with data augmentation Prediction Script (predict.py): Command-line tool for single image classification Web Interface (streamlitapp.py): Interactive Streamlit app with camera support Complete Documentation: README with setup and usage instructions

🎯 Use Cases Food quality control and automated sorting Reducing food waste through optimal timing Learning computer vision and transfer learning Building production-ready classification systems

Made with ❤️ using TensorFlow & Streamlit ⭐ If you find this dataset useful, please upvote and share your results! 📖 Full documentation in README.md | 🐛 Report issues in discussions | 💡 Share your projects!

Here are a few use cases for this project:

1. Grocery Store Inventory Management: Stores can use the Banana Ripeness Classification model to automatically monitor the ripeness of their banana stock, allowing them to more efficiently manage inventory by replacing overripe and rotten bananas while prioritizing the sale of ripe ones.

2. Produce Quality Control in Supply Chain: Producers and distributors can implement the model to assess the quality and ripeness of bananas during the shipping process, helping to reduce food waste by identifying and addressing ripeness issues before the produce reaches the stores.

3. Automated Crop Harvesting: Farmers can integrate the Banana Ripeness Classification model into robotic harvesting systems, ensuring that only bananas at optimal ripeness stages are picked. This would streamline the harvesting process and potentially lead to higher market value for the produce.

4. Smart Home Kitchen Management: Homeowners can use the model with a smartphone app or smart appliances to monitor the ripeness of bananas and other produce in their kitchen, alerting them to consume or utilize the bananas before they become overripe, promoting healthier eating habits and reducing food waste.

5. Food Industry and Recipe Recommendations: Recipe and meal planning apps can leverage the Banana Ripeness Classification model to suggest tailored recipes based on the user's available banana ripeness level. For example, suggesting banana bread recipes for overripe bananas, or salads and smoothies for ripe ones.

Banana Disease Detection

## Overview

Banana Disease Detection is a dataset for object detection tasks - it contains Fusariumwilt annotations for 615 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Our general objective is that the dataset proposed here be useful for the development of Machine Learning and Computer Vision algorithms whose central object of analysis is the banana. The images contained here are of bananas from the Prata Catarina cultivar, with labeling of eight classes that represent levels of control of the fruit. For the labeling process, they were labeled via Bounding box, demarcating the banana in the image and assigning it a degree of maturation following the norms proposed in CEAGESP (2006). All 1000 images of bananas were taken using only smartphones. These images were collected on February 4th, 13th and 17th, 2023, with variations of the background on a smooth surface (white marble), on clayey soil or on foliage. Lastly, all data was uploaded to the roboflow platform and labeled using the bounding boxes method.

This dataset contains colour images of top fruit types with two quality labels: Good and Bad. It includes six classes: Apple_Good, Apple_Bad, Banana_Good, Banana_Bad, Lime_Good, Lime_Bad. The approximate sample counts per class are:

• Apple_Bad: 1141

• Apple_Good: 1134

• Banana_Bad: 1087

• Banana_Good: 1113

• Lime_Bad: 1085

• Lime_Good: 1094

The images were captured under varying lighting conditions, backgrounds and viewpoints, using high-resolution mobile phone cameras, both indoor and outdoor. The dataset is derived from the FruitNet dataset (Meshram et al., 2022) which originally included six fruits (apple, banana, guava, lime, orange, pomegranate) and three quality labels (Good/Bad/Mixed) with ~19 500 images.

In this version we have selected the six classes (three fruits × two quality levels) to provide a balanced corpus of ~ 6,700 images. Each image is stored at 256×256 resolution (or rescaled to this size) and labelled with the class name.

Intended Use:
This dataset is suitable for research in computer vision, particularly fruit quality classification, defect detection, visual sorting in agriculture, and related machine learning tasks.

Source / Reference:
Meshram V. A., Patil K., Ramteke S. D. “MNet: A Framework to Reduce Fruit Image Misclassification” (2021) IIETA. DOI: 10.18280/isi.260203. Dataset details: top Indian fruits, Good/Bad labels, ~12,000 images. (https://www.iieta.org/journals/isi/paper/10.18280/isi.260203?utm_source=chatgpt.com">IIETA)

Dataset structuring:

• Directory per class (e.g. Apple_Bad/)

• JPEG or PNG images

• Recommended preprocessing: resizing to 256×256, normalising pixel values, optional data augmentation